100 General Information

Abstract
This section provides an overview of typical heat exchanger applications, available
industry technology and standards, and the Company’s normal degree of involve-
ment in heat exchanger design and selection.

Contents Page
110 Types of Exchangers and Applications 100-2
120 Industry Codes and Practices 100-2
130 Company Practice 100-3

December 1989  1989 Chevron USA Inc. All rights reserved. 100-1
100 General Information Heat Exchanger and Cooling Tower Manual

110 Types of Exchangers and Applications

The type of exchanger used for a given service depends on process, safety and oper-
ating requirements, and economics. Selection of the most economical heat
exchanger in a given service includes the initial cost of the exchanger, installation
costs, maintenance costs, incremental pumping and compression costs, and where
applicable, utility system investment and energy requirements.
As an example, the initial cost of a water-cooled shell and tube heat exchanger will
be much less than an air-cooled heat exchanger for the same service. However, the
cost of the cooling water supply, including makeup water, chemical treatment, blow-
down disposal, the cooling tower and fans, water circulation pumps, and piping to
and from the water-cooled exchanger, often more than offset the lower first cost of
the water-cooled exchanger. Operating and maintenance costs for water coolers are
also more than for air coolers.
Air cooling is usually more economical than water cooling when the required
process fluid outlet temperature is 20°F or more over the design ambient air temper-
ature. Specific air cooler types, and their applications and limitations, are discussed
in Section 600.
In most cases, a custom designed shell and tube heat exchanger will be the econom-
ical choice. Specific varieties of shell and tube exchangers, and their applications
and limitations, are discussed in Section 400.
Pre-engineered double pipe or small multitube hairpin heat exchangers compete
with custom designed shell and tube exchangers and are usually economical when
the required heat transfer surface area is less than about 1000 square feet. Informa-
tion on small pre-engineered exchangers is given in Section 400.
Plate and frame heat exchangers may be economical where weight and space are
highly valued, alloy material is required, pressures and temperatures are moderate
(less than 200 psig and 250°F), and fluids are compatible with rubber gaskets. Tita-
nium plate and frame exchangers on offshore platforms in sea water/fresh water
service are appropriate applications.

120 Industry Codes and Practices

The Company uses Heat Transfer Research Inc. (HTRI) technology, supplemented
by Company experience, for design and evaluation of thermal and hydraulic perfor-
mance of heat exchangers. HTRI conducts large scale tests of heat exchangers and
incorporates the results into reports and computer programs. HTRI has tested exten-
sively shell and tube exchangers in boiling, condensing and single phase service,
and completed limited testing of air coolers and plate and frame exchangers. HTRI
information is contract confidential and should not be disclosed outside of the
Company.
Mechanical design of exchangers is based on standards developed by the American
Society of Mechanical Engineers (ASME), Tubular Exchanger Manufacturers Asso-
ciation (TEMA), American Petroleum Institute (API), and Chevron.

100-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 100 General Information

ASME covers mechanical design of pressure vessel components (shells, channels,

tubesheets, headers and some aspects of large flanges), and imposes restrictions on
piping, instrumentation and inspection of steam boilers.
TEMA addresses shell and tube heat exchangers only, and covers nomenclature,
fabrication tolerances, standard tube sizes, standard clearances, minimum plate
thickness, and minimum tie rod requirements. TEMA also provides tubesheet
design rules that differ from ASME tubesheet design rules.
TEMA tubesheet rules are used for most shell and tube exchangers. ASME
tubesheet rules are used for stayed fixed tubesheet steam boilers, and air cooler
headers. API standards address mechanical design criteria of shell and tube
exchangers, and air-cooled exchangers.
There are no industry standards for double pipe, multitube hairpin or plate and
frame heat exchangers.
Fouling, corrosion, tube vibration, leak tightness of flanges, and safe design for tube
rupture pressure transients are also discussed in this manual because the subjects are
beyond the scope of industry standards. (See Sections 200, 300 and 500.)

130 Company Practice

Company practices vary. Some locations normally specify all geometric parameters
that affect thermal and hydraulic performance of shell and tube exchangers, with
due consideration for fouling, corrosion, tube vibration, and maintainability.
Mechanical design is then left to the manufacturer, subject to the Company’s speci-
fied criteria and standards. Other locations only specify the fluid streams, flow rates,
and desired thermal performance. They rely on Vendors to do the mechanical and
thermal design, and have reported satisfactory results.
Experience with shell and tube exchanger manufacturers also varies. Some facili-
ties have experienced problems with manufacturers’ designs for controlling tube
vibration, fouling, corrosion, and thermal and hydraulic performance, whereas other
locations routinely leave these design considerations to the manufacturer and expe-
rience no problems.
Thermal, hydraulic, and mechanical design of air coolers are usually completed by
the manufacturer, subject to the Company’s specified criteria and standards. Air
coolers are used to condense and cool low viscosity fluids. Vendors have adequate
expertise in these services, and in air side design, including fan design, hub seals, tip
seals, auto-variable blade control, etc. If fouling is a potential problem, the
Company must take responsibility for it and advise the Vendor how to control it.
Tube vibration is not a problem in air coolers.
Double pipe and small multitube hairpin exchangers are pre-engineered and have
proprietary components. They are preferred for small duty, low viscosity, single
phase services. The Company normally selects specific models from catalog infor-
mation.

December 1989  1989 Chevron USA Inc. All rights reserved. 100-3
100 General Information Heat Exchanger and Cooling Tower Manual

100-4  1989 Chevron USA Inc. All rights reserved. December 1989
200 Design Background

Abstract
This section presents introductory information and general design guidelines. Heat
transfer and fluid flow correlations and rigorous analytical methods are referenced.

Contents Page
210 Heat Transfer 200-2
211 Mean Temperature Difference
212 Overall Heat Transfer Coefficient
213 Single-phase
214 Two-Phase Liquid/Gas Heat Transfer
215 Boiling
216 Condensing
220 Economic Pressure Drop and Velocity 200-13
230 Flow Splitting 200-14
240 Fouling 200-15
250 Overdesign 200-16
260 Tube Vibration 200-17
270 Enhanced Surfaces 200-19
280 Computer Program Abstracts 200-19

December 1989  1989 Chevron USA Inc. All rights reserved. 200-1
200 Design Background Heat Exchanger and Cooling Tower Manual

210 Heat Transfer

The rate equation for heat transfer is:

Q = U  A  MTD
(Eq. 200-1)
where:
Q = Heat duty, Btu/hr
U = Overall heat transfer coefficient, Btu/(hr °Fft2)
A = Heat transfer area, ft2
MTD = Mean temperature difference, °F
The area, A, and the reference area in the overall coefficient, U, must be the same.
Outside bare tube surface area is the usual reference area for tubular equipment.
Estimation methods for mean temperature differences and overall heat transfer coef-
ficients are discussed in the following sections. Flow in tubes and flow across bare
tube bundles for single-phase and some multiphase cases are covered.

211 Mean Temperature Difference

Mean temperature difference, MTD, is defined as the area weighted average temper-
ature difference between hot and cold streams in the heat exchanger:

MTD  (1/A) A (Th-tc) dA

(Eq. 200-2)
where:
Th = Local hot fluid temperature, °F
tc = Local cold fluid temperature, °F
MTD depends on flow arrangement and has been calculated in dimensionless form
for commonly used flow arrangements, with the assumptions that U is constant and
the heat content of each stream varies linearly with temperature.
Several different forms of MTD charts are available in the literature. The most
useful form involves cross plots of four dimensionless parameters (R, N, P and )
defined as follows.
R = Relative Heat Capacity Ratio
= Cc/Ch = (Ti-To)/(to-ti)
N = Number of Transfer Units = UA/Cc
P = Thermal Effectiveness = (to-ti)/(Ti-ti)
 = Dimensionless MTD = MTD/(Ti-ti)

200-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

where:
Ch = (MCp)h = Hot stream heat capacity rate, Btu/hr°F
Cc = (M Cp)c = Cold stream heat capacity rate, Btu/hr °F
M = Mass flow Rate, lb/hr
Cp = Specific heat, Btu/lb °F
Ti = Hot stream inlet temperature, °F
To = Hot stream outlet temperature, °F
ti = Cold stream inlet temperature, °F
to = Cold stream outlet temperature, °F
Graphs relating R, N, P, and  are presented in Appendix A for commonly used flow
arrangements. These graphs are useful for design of individual exchangers, evalua-
tion of exchanger performance at nondesign conditions, evaluation of heat
exchanger network performance, and evaluation of field performance data, as
described below.
For design, terminal temperatures (and therefore P and R) are known, and required
area can be calculated from N. For evaluation of networks or alternative design
conditions, N, R and inlet temperatures are known, and outlet temperatures can be
calculated from P. For evaluation of field performance data, terminal temperatures
(R and P) are known, and U can be calculated from N.
A more common presentation of MTD information is F-factor graphs, where F is
defined as the actual mean temperature difference (MTD) divided by the “log mean
temperature difference” (LMTD). LMTD is the actual MTD for counterflow (F=1),
and is calculated as follows:

LMTD = [(Ti-to)-(To-ti)] / ln [(Ti-to)/(To-ti)]

(Eq. 200-3)
Using the definition of F, Equation 200-1 becomes:

Q = U A F  (LMTD)

(Eq. 200-4)
F-graphs are plots of F against P with R as a parameter. They provide the same
information as  on the more general graphs. Both the general graphs and F-graphs
are provided in Appendix A for each flow arrangement. Only the general graph is
given for counterflow (because F1 for counterflow). F=1 for pure component
(isothermal) boiling and condensing regardless of flow arrangement.
MTD graphs described above are based on the assumption that the heat content of
each stream varies linearly with temperature. This may not be correct where phase
change is involved. For example, cooling a superheated vapor may involve a vari-
able temperature desuperheating zone, an isothermal condensing zone, and a vari-

December 1989  1989 Chevron USA Inc. All rights reserved. 200-3
200 Design Background Heat Exchanger and Cooling Tower Manual

able temperature subcooling zone. Such cases should be analyzed in segments, as

indicated in Figure 200-1, so that the linear assumption is valid for each segment.

Fig. 200-1 Nonlinear Heat Release

212 Overall Heat Transfer Coefficient

Heat transfer coefficients can also be considered as reciprocals to heat transfer resis-
tances. The overall resistance is the sum of the individual resistances illustrated in
Figure 200-2.
Individual inside and outside film and fouling resistances are customarily referred to
the tube surface at which they occur, whereas the overall resistance and overall heat
transfer coefficient are customarily referred to the outside tube surface. Individual
resistances are therefore added as follows:

1/U = (1/hi + Rfi)(Ao/Ai) + Rw + 1/ho + Rfo

(Eq. 200-5)
where:
hi = Inside film coefficient, Btu/hr °F ft2
Rfi = Inside fouling resistance, hr °F ft2/Btu
Ao = Outside area, ft2
Ai = Inside area, ft2
Rw = Tube wall resistance, hr °F ft2/Btu

200-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

ho = Outside film coefficient, Btu/hr °F  ft2

Rfo = Outside fouling resistance, hr °F ft2/Btu
The tube wall resistance is given by:

 d o  12   d o
R w = -------------------- ln  -----
2k w  di 
(Eq. 200-6)
where:
do = Tube O.D., inches
di = Tube I.D., inches
kw = Tube wall thermal conductivity, Btu/hr °F ft
Estimation of film coefficients and the factors that control them are discussed
below.

Fig. 200-2 Heat Transfer Through Circular Tubes

213 Single-phase
This section discusses single-phase convective heat transfer inside tubes and outside
bare tube bundles. Externally finned tubes are discussed in Section 270.
Analysis of heat transfer in commercial heat exchangers can be extremely complex.
For in-tube flow, local tube side heat transfer coefficients vary along the length of
the tube as the flow structure develops, may include both natural and force convec-

December 1989  1989 Chevron USA Inc. All rights reserved. 200-5
200 Design Background Heat Exchanger and Cooling Tower Manual

tion components, and may involve significant fluid property variation in both radial
and axial directions.
Shell side flow involves similar effects, but is much more complicated. The various
flow paths that exist on the shell side of an exchanger are illustrated in
Figure 200-3.

Fig. 200-3 Shell Side Flow Streams

Each flow fraction has a different heat transfer effect and may vary along the length
of the exchanger. “B-stream” flow dominates overall shell side heat transfer. It is
most effective in the cross flow region between baffle tips and somewhat less effec-
tive in the window region around the end of the baffle. “A-stream” is effective, but
applies only to the short section of tube in the baffle hole. “C,” “E,” and “F” streams
are relatively ineffective and are usually minimized by the use of seal bars, dummy
tubes, and small clearances. “C,” “E,” and “F” should each be less than 0.1.
Rigorous computer simulation is required to analyze all of the complexities that
exist on shell and tube sides of a heat exchanger. Good designs, however, fall within
narrow limits where design and evaluation procedures can be greatly simplified. For
example, a typical shell and tube heat exchanger for pumped liquids is a TEMA
“AEU” with 3/4-inch, 14 BWG (minimum wall) carbon steel tubes on 1-inch square
pitch, 45 degree tube layout angle, 20% cut segmental baffles, TEMA standard
clearances, and about 0.25 psi/ft tube side pressure gradient and 0.5 psi/ft (axial)
shell side pressure gradient. This is near optimum for most pumped liquid cases as
discussed in Section 220. Rigorous computer generated heat transfer coefficients for
this case are shown in Figure 200-4. These curves may be used for initial estimates
and scoping studies. Final designs should be checked using the HTRI ST computer
program.

200-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

Figure 200-4 applies to water and to hydrocarbons. The difference between the
water and hydrocarbon curves reflects the difference in fluid properties. Hydro-
carbon fluid properties also vary widely; however, they vary with each other and
with temperature in such a way that heat transfer can be correlated to viscosity and
density. Figure 200-4 applies to hydrocarbons extending from naphtha to residuum
and for temperatures from ambient to 650°F. Overall accuracy of the curves is about
15%.

Fig. 200-4 Heat Transfer Coefficients for Liquid Phase Shell and Tube Heat Exchangers (See Section 213 for instruc-
tions and limitations.)

Heat transfer film coefficients in Figure 200-4 are presented in the conventional
way. Tube side coefficients, hi, are referred to the inside tube surface area. Shell side
coefficients, ho, are referred to the outside tube surface area. The overall heat
transfer coefficient, referred to the outside surface, is:
U = 1/[1/ho + Rw + (1/hi)/(Ai/Ao)]
where the wall resistance, Rw, is 0.0003 hr°F  ft2/Btu and the area ratio, Ai/Ao, is
3/4.

December 1989  1989 Chevron USA Inc. All rights reserved. 200-7
200 Design Background Heat Exchanger and Cooling Tower Manual

To compare shell and tube side film coefficients, both coefficients should be
referred to the same surface area, usually the outside surface. That is, the shell side
coefficient should be compared to 0.75 times the tube side coefficient in
Figure 200-4.
From Figure 200-4, 75% of the tube side coefficient is approximately equal to the
shell side coefficient for both water and hydrocarbon in the turbulent regime
(viscosity < 2 centipoise). This is as it should be. Equal expenditure of pumping
power per unit of heat transfer surface and equally efficient conversion of pressure
drop to heat transfer result in equal heat transfer coefficients in the turbulent regime.
In the laminar regime (viscosities > 10 centipoise), the shell side coefficient is about
seven times the tube side coefficient when referred to the same surface area.
In-tube laminar flow heat exchange is never economical, sometimes leads to
“viscosity plugging” (see Section 610), and should be avoided.
The pumping power expended per unit of heat transfer surface is the controlling
factor influencing heat transfer in turbulent flow. Where expensive alloys are
needed, it is usually economical to spend more on pumping power to save on
exchanger costs. Film coefficients vary at about the 0.4 power of the pressure
gradient.
General heat transfer correlations for liquids and gases are given in Appendix B.
Most of this information is taken from the HTRI Design Manual and is proprietary.
Equation B-1 in Appendix B applies to turbulent flow of liquids or gases in tubes
and is accurate within about 15%. This is a Dittus-Boelter type correlation and is
similar to those found in general heat transfer text books. Corresponding text book
correlations are based on open literature data with accuracy in the 20% to 30%
range.
Appendix B also gives approximate methods to estimate laminar and transition flow
heat transfer in tubes. This information may be off by a factor of two. When accu-
rate laminar flow heat transfer information is needed, the HTRI ST simulation
programs should be used. Simple shell side heat transfer correlations for liquids and
gases are included in Appendix B. They apply to well proportioned shell and tube
exchangers with turbulent flow only. Extreme geometries or low shell side veloci-
ties require computer analysis.
The best conversion of pressure drop to heat transfer occurs with rotated square tube
layout (45 degrees) in liquid service. Inline square tube layout (90 degrees) is
slightly better in gas service where Reynolds numbers are typically very high.

214 Two-Phase Liquid/Gas Heat Transfer

This section describes sensible heat transfer in two-phase hydroprocessing feed
effluent heat exchangers, where phase change (boiling or condensing) is not control-
ling.
Two-phase heat exchangers are usually designed to operate in the annular or churn
flow regimes on both shell and tube sides. For these flow regimes, liquid coats the
heat transfer surfaces and a continuous gas phase flows in the core. Liquid droplets

200-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

are entrained in the core flow. Heat transfer resistance between the gas and the
liquid is negligible and can be ignored. The governing resistance is convection and
conduction in the liquid film covering the heat transfer surface. This type of flow
can be modeled as a pseudo single-phase fluid using “no-slip” mixture density and
velocity, mixture heat capacity (or enthalpy), liquid viscosity and liquid thermal
conductivity. “No-slip” means that the liquid and gas are assumed to flow at the
same velocity. These pseudo single-phase fluid properties can be input to the HTRI
single-phase simulation program or used with the simpler methods described in the
previous section. This approach has been validated against Company in-plant test
data for well mixed flow only.
When horizontal two-phase heat exchangers operate in stratified flow regimes,
liquid accumulates in the bottom of the exchanger. If the outlet nozzle is on top, a
stagnant liquid level rises in the exchanger as needed to force the upper part of the
exchanger into a mixed phase flow regime where net liquid transport is possible.
For this situation, heat transfer varies as flow rate to the 1.6 power, and pressure
drop is nearly independent of flow rate. This is simply the effect of varying level of
stagnant liquid.
If the outlet nozzle is on the bottom, vapor flows in the upper part and liquid flows
in the lower part. The fraction of the exchanger associated with each phase is more
or less in proportion to their relative volume flow rates.
Stratified two-phase flow in exchangers results in very poor thermal performance
and should be avoided.
Appendix C defines appropriate flow regime boundaries and gives the equations
used to calculate pseudo single-phase properties.

215 Boiling
Pool boiling data are the foundation for correlating the performance of commercial
boiling equipment. A typical pool boiling curve, from the HTRI Design Manual, is
shown in Figure 200-5. It applies to boiling water at atmospheric pressure. This type
of boiling is obtained when the heated surface is surrounded by a fluid that is not
flowing. Agitation is produced by natural convection currents and bubble motion.
The physical condition associated with various parts of the curve is illustrated in
Figure 200-5 and briefly described below.
A-B Natural convection (no boiling)
B-C Incipient boiling (surface boiling with subcooled bulk fluid)
C-D Nucleate boiling (bulk fluid at saturation temperature)
D-E Transition to film boiling (unstable)
E-F-G Stable film boiling (heated surface is not wetted)

Commercial boiling equipment is intended to operate in the incipient boiling or

nucleate boiling regions. Operation in the nonboiling A-B region or the D-E-F-G
film boiling regions may result in severe fouling and/or mechanical failure of the
equipment.

December 1989  1989 Chevron USA Inc. All rights reserved. 200-9
200 Design Background Heat Exchanger and Cooling Tower Manual

The pool boiling curves for single component fluids have been correlated in terms
of critical pressure (Pc) and reduced pressure (Pr) and are given in Appendix D.
Critical pressure is the pressure above which the distinction between liquid and
vapor vanishes. Reduced pressure is the ratio of operating pressure to critical pres-
sure. Boiling is not possible at or above the critical pressure. The critical pressure of
most hydrocarbons is between 400 and 600 psia.

Fig. 200-5 Pool Boiling Curve for Water at Atmospheric Pressure

The nucleate boiling heat flux for multicomponent mixtures is less than that of pure
components at the same surface-to-bulk-temperature difference. More volatile
components boil first, concentrating less volatile components near the heat transfer
surface. This reduction in heat flux correlates with boiling range (dewpoint—bubble
point) and critical temperature. The maximum nucleate boiling heat flux for
mixtures (point D in Figure 200-5) satisfactorily correlates to mixture critical pres-
sure.
Boiling curves for horizontal tube bundles are markedly different than for single
tubes. Figure 200-6 shows an example from the HTRI Design Manual.
The bundle curve was obtained by boiling normal pentane in the shell at saturation
temperature with a large excess of saturated steam in the tubes to minimize tube
side resistance. The Overall Temperature Difference in Figure 200-6 is the steam
saturation temperature minus the n-pentane saturation temperature.

200-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

Fig. 200-6 Example: Comparison of Boiling Curves for a Single Tube and a Bundle, N-Pentane at 115 psia

Nucleate boiling heat flux for bundles is much higher than for single tubes, but
maximum bundle heat flux (dryout) is much lower. Vapor generated by each tube
enhances circulation and heat transfer for neighboring tubes at low heat flux. Exces-
sive vapor generation, however, prevents wetting some interior tubes above the
incipient dryout heat flux. Incipient dryout of central tubes usually occurs at 50% to
70% of the maximum bundle heat flux. Extensive dryout in the interior of the
bundle exists at the maximum bundle heat flux.
Enhanced nucleate boiling heat flux, incipient dryout heat flux, and maximum
bundle heat flux correlate with the ratio of heat transfer area to peripheral
inflow/outflow area. These correlations are given in Appendix D.
At or above incipient dryout heat flux, a fraction of the liquid entering the lower
part of the bundle becomes completely vaporized in the interior. Any solids in the
liquid will deposit at this point. Solids deposition in the bundle restricts flow,
reduces dryout heat flux, and extends the deposition region. Because most commer-
cial streams contain a few parts per million solids, the usual result of operation
above the incipient dryout heat flux is plugging of the bundle.
Operation of natural circulation boilers in a nonboiling region (A-B in Figure 200-5)
results in a similar plugging problem. Fluid shear is usually not adequate to keep
trace amounts of solids in the liquid from accumulating and adhering to the heated

December 1989  1989 Chevron USA Inc. All rights reserved. 200-11
200 Design Background Heat Exchanger and Cooling Tower Manual

surface and eventually stopping circulation. Vigorous agitation associated with

nucleate boiling is more than adequate to keep solids suspended. The minimum heat
flux to ensure nucleation is about 2000 Btu/hr ft2. This may impose a turn-down
limit, particularly for grossly oversized boilers.
Reboilers should be designed reasonably close to but not exceeding incipient dryout
heat flux. This minimizes the size and cost of the reboiler, maximizes allowable
turndown, and minimizes the required temperature of the heat medium. Design
considerations for reboilers are given in Section 360.

216 Condensing
This section discusses the principles of process condensation and gives rules-of-
thumb for steam condensation.
Condensing surfaces are below the dewpoint temperature of the condensing fluid
and are covered by a film of condensate. The resistance to heat transfer between the
vapor and the condensate is negligible for single component condensation. Essen-
tially all of the resistance to heat transfer is convection and conduction across the
condensate film. Major factors affecting condensate film thickness and heat transfer
are whether the condensing fluid flows vertically downward or horizontally,
whether the film is laminar or turbulent, whether the film is vapor shear controlled
(high vapor velocity) or gravity controlled (low vapor velocity), and whether the
condensing fluid is flowing in tubes or outside tube bundles.
The resistance to heat transfer between the vapor and the condensate film is signifi-
cant for multicomponent condensation. Least volatile components condense first,
concentrating more volatile components near the vapor-liquid interface. This tends
to inhibit condensation. This vapor phase resistance is governed by vapor shear and
turbulent mixing in the vapor.
Total condensers usually operate in the vapor shear controlled turbulent film regime
near the inlet, in gravity controlled regimes in the middle, and may have a liquid
flooded zone near the outlet. The heat transfer coefficients may vary from about a
thousand at the inlet to about one Btu/hr °F ft2 at the outlet. This type of
condenser is usually analyzed by solving for the limits of the various flow regimes
and applying the correlations appropriate to each regime. This is an iterative process
and involves a very large number of correlations. Computer simulation with the
HTRI CST program is recommended for these cases.
Some partial condensers operate with vapor shear controlled turbulent film conden-
sation throughout, which is very similar to two-phase heat transfer discussed earlier.
The two-phase methods discussed earlier may be applied to condensers in this case.
Condensing steam is a common and efficient heat medium. Steam condensing coef-
ficients are usually between 2000 and 3000 Btu/hr°F ft2 in the vapor shear
controlled regime and a few hundred in the gravity controlled regimes. Most steam
heated exchangers involve total steam condensation and have an outlet condensate
pot to avoid a condensate flooded zone in the exchanger. For these cases, a design
steam side heat transfer coefficient of 1000 Btu/hr°F  ft2 is recommended.

200-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

220 Economic Pressure Drop and Velocity

Pressure drop is required to change fluid momentum (increase velocity and/or
change direction), and to overcome friction. Friction pressure loss is the dominant
cause of pressure drop in most single- and two-phase heat exchangers. Acceleration
(momentum) loss is most important in boilers and is usually provided by differen-
tial static head. Momentum and friction losses dominate in the inlet regions of
condensers (and overall), with some vapor deceleration pressure recovery occurring
in the bundle. Both heat transfer and pressure drop increase with increasing fluid
velocity.
Optimum pressure drop for single-phase heat exchangers and some two-phase heat
exchangers (excluding boilers and condensers) is determined by balancing the cost
of pressure drop (e.g., pumping power) against the cost of heat transfer surface area.
Inevitable pressure losses associated with nozzles, including entrance and exit zones
of exchangers, do not substantially affect heat transfer. Exchanger nozzles usually
match the piping size and pipe size is usually based on the same principle of
balancing pipe costs against pumping costs. These pressure losses are therefore a
more-or-less constant fraction of the exchanger friction losses.
For pumped liquids, compressed vapors, or both (two-phase) in carbon steel
exchangers, the economic expenditure of pumping power on each side of the
exchanger (including nozzle, entrance, and exit losses) is about 5 hp per 1000 ft2 of
heat exchanger surface area. The optimum is relatively flat. Power expenditures
between 3 and 7 hp/1000 ft2 have a significant effect on exchanger and pump or
compressor costs, but result in approximately the same overall cost (5%). Designs
anywhere in this range are reasonable. Optimum power expenditure per unit area is
approximately the same for any fluid density and tube size.
The above economic range of power expenditure can be expressed in terms of
velocity and pressure gradient.

Economic velocity, ft/sec (independent of tube size):

(28 to 36)/(Density, lb/ft3)1/3 Tube side
(7 to 10)/(Density, lb/ft3)1/3 Shell side
The shell side velocity is for the “B” stream (crossflow) used in the HTRI programs
and is shown in Figure 200-3. “B” stream flow is usually about 70% of the total
shell side flow. Typical values for low viscosity hydrocarbon liquids and water are:
7 to 9 ft/sec Tube side
2 to 3 ft/sec Shell side

Economic pressure gradient, psi/ft (for 3/4-inch, 13 BWG tubes):

(0.05 to 0.08) (Density, lb/ft3)1/3 Tube side
(0.1 to 0.16) (Density, lb/ft3)1/3 Shell side

December 1989  1989 Chevron USA Inc. All rights reserved. 200-13
200 Design Background Heat Exchanger and Cooling Tower Manual

The tube side flow path is the straight tube length times the number of tube passes.
The shell side flow path is the axial shell length times the number of shell passes.
Typical values for low viscosity hydrocarbon liquids and water are:
0.2 to 0.3 psi/ft Tube side
0.4 to 0.6 psi/axial ft Shell side
The factor of two difference between shell and tube side pressure gradients simply
reflects the fact that actual shell side flow path is about twice the axial length.
Economic pressure gradient for 1-inch tubes is about 70% of that for 3/4-inch tubes.
These simple rules-of-thumb are insensitive to wide variations in energy costs,
because exchanger costs are energy intensive and track well with energy costs.
Exchangers with similar tube side and shell side heat transfer coefficients (the
norm) should be designed to the economic parameters particular to each side. If one
side of the exchanger limits (much lower coefficient than the other side), the above
economic parameters should be used for the limiting side, and less power expended
on the side that does not limit. Power expenditures on the high side, or more, are
justified for expensive alloy exchangers.
In some cases (e.g., desalter effluent water coolers), available pressure drop is
dictated by other process requirements (e.g., suppressing vaporization in the
desalter) and should be used to the maximum practical extent to reduce exchanger
size.
These guidelines do not apply to reboilers and condensers. Reboilers are usually
driven by natural circulation rather than pumps. Pressure drop in condensers
impacts column and reboiler design as well pumping/compression costs. At atmo-
spheric pressure or above, condenser pressure drop is typically the smaller of about
5 psi or 10% of the absolute pressure. At very low pressures, the type and perfor-
mance of vacuum equipment governs condenser pressure drop.
The economics of pumping/compression costs versus exchanger cost is one aspect
of optimum exchanger utilization. Equally important is balancing the value of heat
exchange against the cost of achieving it (i.e., determining the appropriate duty of
the exchanger). Determination of exchanger duty is beyond the scope of this
manual.

230 Flow Splitting

Flow splitting is sometimes a problem in networks and in multiple bundle services
with parallel branches. Flow splitting can always be controlled with valves, but that
wastes pumping power and is undesirable. Flow splitting in treated cooling tower
water systems has been a problem due to oversizing some condensers and throttling
water flow for control, or designing different services with different numbers of
bundles in series and with large differences in design pressure drop.
Throttling treated cooling tower water is always undesirable, as discussed in
Section 310. All water cooler services in a particular circulating system should be
designed with the same overall pressure drop and unrestricted water flow whenever
possible.

200-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

Splitting two-phase streams has been a problem when the piping is not symmet-
rical, even though care was taken to balance pressure drops in parallel paths. Reli-
able two-phase flow splitting is accomplished by splitting both branches along the
same axis and at right angles to the original combined stream in a standard pipe tee,
and designing for equal pressure drop in each branch.

240 Fouling
This section discusses principal fouling mechanisms, and services and conditions
where they may occur. In most cases, fouling conditions can be avoided by appro-
priate process and exchanger design. The principal fouling mechanisms are:
• Particulate fouling
• Salt precipitation fouling
• Chemical reaction fouling
• Filming amine fouling
• Biological fouling
• Corrosion fouling
Particulate fouling is possible for streams that contain a few parts per billion of
solids. Particle sizes between 0.001 and 1 micron contribute most to the deposit.
Particles over about 100 microns usually erode smaller particles and inhibit fouling.
Simple particulate deposits are weak and yield readily to fluid shear. Exchangers
that operate with fluid shear stresses greater than about 0.001 psi are usually not
subject to simple particulate fouling.
Economic liquid velocities discussed in Section 220 result in nearly four times the
shear stress (or twice the velocity) needed to prevent simple particulate fouling.
Economic velocities for most two-phase exchangers are also above the threshold for
particulate fouling, except at high vapor fraction. Economic gas velocities are not
adequate to keep small particles moving. For example, in FCC flue gas coolers the
process and/or exchanger design must be adjusted to control gas side fouling. This
example is discussed in Section 390.
Salt precipitation fouling usually involves liquid-to-solid phase transition at the
heat transfer surface. It occurs where an aqueous phase contacts the heat transfer
surface and the aqueous phase is supersaturated with respect to one or more of the
dissolved salts. Most salt deposits can not be eroded at economic velocities. A few
salt deposits are erodible at economic liquid velocities, and these are candidates for
aqueous phase corrosion inhibitors (See Section 310). Salt precipitation fouling may
occur alone or in combination with particulate fouling.
Gas-to-solid phase transition (sublimation) is a less common type of salt precipita-
tion that can occur in overhead condensers and effluent streams in hydroprocessing
plants. NH4Cl is usually the salt involved.
Addition of an aqueous phase and/or control of its composition is the normal
method of eliminating salt precipitation conditions in heat exchangers. This method
is practiced in cooling tower water exchangers, sea water coolers, crude oil preheat
exchangers, crude unit atmospheric overhead condensers and some other services.

December 1989  1989 Chevron USA Inc. All rights reserved. 200-15
200 Design Background Heat Exchanger and Cooling Tower Manual

Chemical reaction fouling involves chemical bonding between thermally unstable

organic compounds to the extent that liquid-to-solid phase transition occurs at the
heat transfer surface. This is a slow process in liquids and usually does not cause
fouling at economic velocities in the absence of particulate matter. However, chem-
ical reaction fouling can occur in combination with particulate fouling and affect
fouling by increasing the strength of predominantly particulate deposits. Chemical
reaction fouling increases exponentially with temperature above a certain threshold
temperature. It is controlled by maintaining heat transfer surface temperatures below
the threshold. The threshold fouling temperature for naturally occurring hydrocar-
bons is about 600°F for the heaviest components and higher for lighter components.
Residuum fouling is the most common type of chemical reaction fouling in refin-
eries.
Filming amine fouling usually occurs with liquids in combination with particulate
fouling and affects fouling by increasing the strength of predominantly particulate
deposits. Economic liquid velocities will not stop filming amine fouling. Filming
amines can cause fouling in two other ways. Filming amines dissolved in light
hydrocarbon will “deposit” as a liquid (like tar) where the light hydrocarbon is
evaporated to extinction. Most filming amines also decompose to solids above a
certain temperature, usually between 300°F and 500°F.
Filming amines are marketed as corrosion inhibitors, dispersants, antioxidants,
metal deactivators and antifoulants. They are commonly injected into crude unit
atmospheric overhead systems and dispersed throughout the refinery.
Biological fouling occurs spontaneously in oxygenated waters between 32°F and
120°F, unless significant toxic material is present. Cooling tower water systems and
sea water cooling systems are subject to biological fouling, usually in combination
with particulate fouling. Chlorination is the most common method of biological
fouling control. Redundant equipment and/or frequent cleaning is an alternative.
Corrosion fouling involves irregular loss of metal and accumulation of corrosion
products. Increased surface roughness improves convective heat transfer and
compensates for added thermal resistance of corrosion products that remain on the
surface. Pressure drop increases initially due to the roughness but may decrease if
metal loss is significant. These effects are usually small in heat exchangers and are
usually ignored in design. Quantitative fouling information is given in Section 300.

250 Overdesign
Typical Company practice is to rate plants and most component equipment at 80%
of their calculated capacity. This rating allows for normal uncertainty in design,
repairable deterioration during one turn-around cycle, and a margin for control. That
is, most plants could ideally run at 125% of rated capacity initially and should be
able to run at 100% of rated capacity at planned end of run.
Heat exchanger thermal and hydraulic overcapacity should be comparable to that of
other equipment in the plant. More overcapacity may be justified for unconven-
tional processes with little commercial experience.

200-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

Additional heat exchanger surface area is needed in services where fouling is

intended (corrosion inhibitor films) or is allowed to occur. Corrosion inhibitor
fouling is asymptotic (self-limiting) at temperatures below the thermal stability
limits of the protective films. Affected exchangers should be designed for 80% of
the asymptotically fouled performance (design U = 0.8  fouled U). Asymptotic
fouling resistance and film stability limits are given in Section 310 for treated
cooling tower water and in Section 380 for filming amine corrosion inhibitors.
Most fouling situations are not asymptotic (fouling resistance increases with time
indefinitely). Redundant (spare) heat exchangers with isolation block valves and
frequent cleaning are practical methods of accommodating nonasymptotic fouling.
Although 50% to 100% excess area for both the operating exchanger and its spare
with cleaning every month or two is usually effective, it may not be economical.
This approach has been used for some crude preheat systems, and to heat large
quantities of recovered oil with thermally unstable components.
Fouling prevention is usually more economical than fouling accommodation.
Sections 330, 350, 360 and 380 discuss fouling prevention for closed loop water
cooling, high temperature steam generators, reboilers, and crude unit heat
exchangers, respectively.
Biological fouling from sea water is nonasymptotic but is an order of magnitude
slower than other types of fouling. Both fouling prevention and fouling accommo-
dation are common in sea water exchangers. Section 320 discusses biological
fouling rates of untreated sea water, the conditions when it occurs, and other limita-
tions.
Although most commercial streams will foul under some conditions, most
exchangers are sized for clean operation only when the expected range of operation
conditions is outside the range where fouling occurs. Section 250 discusses prin-
cipal fouling mechanisms, and the services and conditions where they may occur.
The exchanger designer normally only specifies the calculated clean pressure drop.
Hydraulic overdesign is usually specified by the pump or compressor designer and
is usually 125% of design flow (150% of clean pressure drop), plus a 30% control
allowance. This factor is usually more than adequate to accommodate pressure drop
increase due to fouling.
The exchanger designer should check all shell and tube exchangers for tube vibra-
tion at 125% of design flow, as discussed in Section 260 below.

260 Tube Vibration

This section discusses tube vibration mechanisms, tube failure locations, and design
criteria to prevent failures. Damaging heat exchanger tube vibration mechanisms are
vortex shedding and fluid elastic whirling. Acoustic resonance can occur in
exchangers but does not damage tubes and is not covered here.
Vortex shedding downstream of a single tube in cross flow is illustrated in
Figure 200-7, and is a “snap shot” of a vortex forming near the tube and three other
vortices that were shed earlier in sequence.

December 1989  1989 Chevron USA Inc. All rights reserved. 200-17
200 Design Background Heat Exchanger and Cooling Tower Manual

Fig. 200-7 Illustration of Vortex Shedding from a Single Tube

Vortices are alternately shed on each side of the tube and exert an alternating pres-
sure (force) in the direction perpendicular to the crossflow direction. This alter-
nating pressure makes the upstream flow first favor one side of the tube and then the
other.
The vortex shedding phenomenon for a single tube occurs in heat exchanger
bundles at peripheral tubes and along unblocked pass partition lanes in the vicinity
of inlet and outlet nozzles, but does not penetrate far into the bundle.
The location of the next row of tubes affects the frequency of formation of vortices
on the first row. Interaction between vortices from adjacent tubes usually degener-
ates the vortices to harmless random noise after about the third row into the bundle.
That is why only peripheral tubes are of concern.
Vortex shedding frequency depends on tube layout angle. The distinction between
30 degrees and 60 degrees, and between 45 degrees and 90 degrees, however, is
academic because of the extreme divergence or convergence of nozzle flows. The
60-degree layout angle governs for all triangular layouts; the 45-degree layout angle
governs for all square layouts.
Damaging tube vibration occurs when the vortex shedding frequency matches one
of the tube natural frequencies. The maximum allowable unsupported tube span is
set so that the highest anticipated crossflow velocity will not excite the first mode
natural frequency of the tube. Higher mode resonant vibration occurs at higher
velocities.
Figure 1 in Standard Drawing GC-E1048 defines maximum unsupported tube spans
for inlet and outlet regions of shell and tube exchangers. Vibration control involves
adding partial support near nozzles as needed. These support plates do not affect
thermal or hydraulic performance.
Fluid elastic whirling may occur in the first two rows beyond the baffle cut in the
interior of the bundle only. These are the spans labeled “L4” in Figure 3 in Standard
Drawing GC-E1048. Clusters of at least three tubes in at least two different rows
vibrate in harmony, as illustrated in Figure 200-8.

200-18  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 200 Design Background

Fig. 200-8 Fluid Elastic Whirling Tube Motion

Figure 2 in Standard Drawing GC-E1048 defines maximum unsupported tube spans

for interior tubes. The velocities that initiate fluid elastic whirling are greater than
economic velocities and therefore rarely affect exchanger design. Problems occur
when abnormally high velocities and/or abnormally long unsupported spans are
used.

270 Enhanced Surfaces

Enhanced surfaces are justified when one side of a heat exchanger has an inherently
lower heat transfer coefficient than the other side (e.g., low gas side coefficient and
high liquid coefficient). Air cooled exchangers are the most common example. See
Section 600.
Water cooled compressor inter- and after-coolers are also appropriately designed
with enhanced surfaces. Low fin tubes with area ratios between three and five to one
are commonly used. Optimum fin area depends on the gas pressure. The HTRI ST-5
program can be used to select the most appropriate type of tube.
There are few other appropriate exchanger applications of enhanced surfaces in the
petroleum and petrochemical business.

280 Computer Program Abstracts

The following HTRI heat exchanger simulation computer programs are available on
the Company’s mainframe computer system.
ST-5—Single-phase shell and tube heat exchangers
This program is primarily used to model single-phase vapor and liquid. It is also
used for two-phase mixtures modeled as pseudo single-phase fluids, described in
Section 214. This program should also be used when the limiting side is single-
phase even though boiling or condensing occurs on the other side. Steam condensers
and steam generators are usually modeled on ST-5.

December 1989  1989 Chevron USA Inc. All rights reserved. 200-19
200 Design Background Heat Exchanger and Cooling Tower Manual

CST-2—Condensing shell and tube heat exchangers

This program is used to model multicomponent condensation in horizontal and
vertical down-flow shell and tube exchangers. The coolant may be single-phase or
boiling.
RTF-2—In-tube boiling shell and tube exchangers
This program is used to model vertical thermosiphon reboilers, and horizontal or
vertical forced flow reboilers.
RKH-2—Boiling outside horizontal bundles
This program is used to model boiling in horizontal bundles in enlarged shells
(kettle reboilers) or without shells (internal column reboilers). The program does a
simplified analysis of flow in inlet and outlet piping. It uses average fluid properties
and does not calculate circulation rigorously. The program should be used with
caution.
ACE-2—Air cooler heat exchangers
This program is used to model horizontal air coolers with single-phase or
condensing fluids in the tubes.
Instructions for access and use of these computer programs are given in the Heat
Exchanger Design Program Users Guide maintained separately.

200-20  1989 Chevron USA Inc. All rights reserved. December 1989
300 Service Considerations

Abstract
This section provides design and operating guidelines that apply to specific service
classes, including treated cooling water, untreated sea water, closed loop water
cooling, low temperature process steam generators, high temperature waste heat
steam generators, reboilers, crude unit heat exchangers, and FCC flue gas coolers.

Contents Page
310 Treated Cooling Water 300-2
320 Untreated Sea Water 300-4
330 Closed Loop Water Cooling 300-5
340 Moderate Temperature Process Steam Generators 300-6
350 Very High Temperature Waste Heat Steam Generators 300-8
351 Components
352 Field Modifications
360 Reboilers 300-11
370 Condensers 300-20
380 Crude Unit Heat Exchangers 300-21
390 FCC Flue Gas Coolers 300-24

December 1989  1989 Chevron USA Inc. All rights reserved. 300-1
300 Service Considerations Heat Exchanger and Cooling Tower Manual

310 Treated Cooling Water

Heat exchangers in treated cooling water service should be designed and operated
consistent with the water treatment. The temperature limits and thermal resistance
of corrosion inhibitor films affect the design and operating limits of cooling water
exchangers.
Cooling tower water is normally acidified to suppress alkaline scaling, and concen-
trated (“cycled”) close to but not exceeding the solubility limit of the least soluble
salt at the maximum design tube wall temperature. Chlorination is normally used to
suppress biological fouling. These procedures eliminate natural causes of water
fouling.
Corrosion inhibitors are added to cooling tower water to provide a protective barrier
between the oxygen saturated water and carbon steel components of the system.
Corrosion inhibitor films are effective in piping and on the tube side of heat
exchangers. Movement of tubes in baffle holes prevents effective film formation on
the outside (shell side) of baffled tubes. Where shell side water is used, tube mate-
rial should be selected to resist oxygenated water.
Thermal resistance of corrosion inhibitor films is significant and should be included
in heat exchanger design. Recommended design resistance for corrosion inhibitor
films is given in Figure 300-1. Figure 300-1 is based on HTRI data obtained in
member plants, and includes the effects of imperfect control in the commercial envi-
ronment.

Fig. 300-1 Asymptotic Fouling Resistance for Cooling Water Corrosion Inhibitor Films

300-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

The temperature limits of corrosion inhibitor films are more important than their
thermal resistance. The protective films break down between 160F and 220F and
result in rapid fouling and loss of corrosion protection. The interface temperature
between the metal and the inhibitor film should be kept below 160F throughout the
exchanger. Figure 300-2 expresses this criterion in a plot of maximum process
temperature versus the process-side heat transfer coefficient for various water-side
velocities. The curves are based on 120F bulk water.

Fig. 300-2 Corrosion Inhibitor Stability Limits for Treated Cooling Tower Water

The shell side heat transfer coefficients for liquid hydrocarbons from Figure 200-4
are superimposed on Figure 300-2 for convenience. Cooling liquid hydrocarbons at
economic velocities is not a problem in the usual temperature range.
Condenser design within the inhibitor stability limits is difficult when condensing
temperatures are over 200F, since the condensing heat transfer coefficients in inlet
regions are usually between 500 and 1000 Btu/hrFft2. Air cooled condensers are
preferred in those cases.
Figure 300-2 can also be used to evaluate the effect of throttling water to control
condenser duty. Throttling reduces water velocity, increases water temperature, and
is likely to exceed inhibitor stability limits if the condensing-side temperature is
over 160F. Bypassing part of the process fluid around the condenser is a better way
to control a water cooled condenser.
Corrosion inhibitor film resistance at economic liquid velocity is not large enough
to justify the use of all alloy cooling water systems without corrosion inhibitors.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-3
300 Service Considerations Heat Exchanger and Cooling Tower Manual

320 Untreated Sea Water

Heat exchangers using untreated sea water should be designed and operated to avoid
inorganic fouling, and to avoid or accommodate biological fouling. By operating
exchangers near economic velocities and suppressing water vaporization, inorganic
fouling can usually be avoided. Periodic cleaning is usually required to accommo-
date biological fouling.
Solubility of calcium sulfate usually limits the applications of sea water cooling.
Calcium sulfate hemi-hydrate determines the limits in nonboiling applications; it
precipitates at surfaces over 310F. Calcium sulfate anhydrite determines the limits
where subcooled boiling occurs; it precipitates on boiling surfaces over 230F.
Alkaline scaling (e.g., CaCO3 or Mg(OH)2 precipitation) may occur in some coastal
waters where the pH exceeds about 7.6. Acidification of such water is necessary to
make it suitable for cooling purposes. Particulates do not cause fouling unless veloc-
ities are less than about half the economic velocity.
Biological fouling can be prevented by chlorination or by maintaining surfaces over
120F. Chlorination is not always environmentally acceptable or practical. Main-
taining heat transfer surfaces over 120F has been used in subsea natural convec-
tion crude oil coolers needed to protect nonmetallic subsea production hoses.
Keeping tube surface temperatures over 120F is not practical in most cases.
Biological fouling occurs spontaneously in natural sea water between 32F and
120F. Particulates make up a major portion of the deposit and are bonded to each
other and to the tube by biological material. Observed average fouling rates for tita-
nium and 90-10 Cu-Ni tubes in sea water are shown in Figure 300-3.

Fig. 300-3 Average Observed Biological Fouling Rates for Titanium and Cu-Ni Tubes in Sea Water

300-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

The lower average fouling rate for 90-10 Cu-Ni tubes reflects the tubes periodic
sloughing of deposit from itself. Copper slowly dissolves in sea water (corrodes)
through the deposit and is toxic to lower forms of aquatic life. Copper corrosion and
toxicity, however, are not sufficient to break the bond at the deposit-metal interface.
Titanium is immune to corrosion, is not toxic, and therefore, is generally recom-
mended for sea water service.
Larvae and small marine creatures pass through line filters (screens) and can grow
large enough to block heat exchanger tubes. Back flushing lines once a shift elimi-
nates this possibility.

330 Closed Loop Water Cooling

Closed loop water cooling is used where the coolant temperatures must be main-
tained over a certain temperature to avoid solidifying the process stream (“tempered
water cooling”), or for high process temperature applications that would exceed the
temperature limits of treated cooling tower water or sea water discussed above.
Closed loop water is usually boiler quality and either inhibited with chromate or
nitrogen blanketed. Chromate is effective if tube temperatures are below the water
boiling point. If tube temperature could exceed the water boiling point, nitrogen
blanketing without inhibitor is recommended.
Heavy oils may freeze or become immobile at temperatures above normal coolant
temperature. Cooling such fluids is usually accomplished with “tempered water
cooling,” as illustrated in Figure 300-4. The coolant temperature should be
controlled about 10F above the freezing/pour point temperature of the stock being
cooled.

Fig. 300-4 Typical Tempered Water Cooling System

December 1989  1989 Chevron USA Inc. All rights reserved. 300-5
300 Service Considerations Heat Exchanger and Cooling Tower Manual

340 Moderate Temperature Process Steam Generators

This section applies to steam generators heated by process fluids with temperatures
less than 1000F.
Many process steam generators operate by natural circulation from an elevated
steam drum. Circulation is driven by the density difference between the liquid in the
downcomer and the two-phase mixture in the riser (external to the exchanger). A
typical steam generator is shown in Figure 300-5. Design guidelines for this type of
steam generator are listed below.
The steam/water riser to the steam drum should operate in annular or churn flow.
• Side-to-side flow with 90 degree tube layout is used for good circulation at low
pressure drop.
• Target vaporization is usually 5% of total circulation and is obtained by appro-
priate pipe sizes and drum elevation.
• Blowdown is from the steam drum near the vapor-liquid interface where water
is most concentrated.
• Heat flux should be 2000 Btu/hrft2 or higher to assure nucleate boiling
throughout the bundle.
• A support plate at the steam/water outlet is usually required to avoid tube vibra-
tion.

Fig. 300-5 Typical Conventional E-Shell Steam Generator

300-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

A single steam generator and steam drum are sometimes combined in a kettle-type
steam generator as shown in Figure 300-6.

Fig. 300-6 Typical Kettle Steam Generator

The bundle is elevated relative to the lower part of the shell to provide a minimum
bundle-to-shell gap equal to about 10% of the bundle diameter. Cutouts in the lower
part of the full tube support plates are required to facilitate free axial flow of water
along the bottom of the bundle. Circulation is driven by the density difference
between the liquid at the sides of the bundle and the two-phase mixture in the
central part of the bundle (internal circulation). Blowdown is from the bottom of the
shell near the tubesheet. Liquid level is maintained a minimum of 3 inches above
the top tube, plus a control allowance. The kettle should be sized to provide
adequate vaporliquid disengaging space between the normal operating liquid level
and the top of the kettle. Horizontal mass velocity of steam approaching the steam
outlet nozzle(s) in the disengaging space (Figure 300-7) should not exceed the
following criteria:
Operating Pressure, psig Mass Velocity, lb/hrft2
15 1900
40 2900
60 3400
150 4800
300 6400
600 8500

December 1989  1989 Chevron USA Inc. All rights reserved. 300-7
300 Service Considerations Heat Exchanger and Cooling Tower Manual

Fig. 300-7 Steam Disengaging Velocity

The above criteria provide adequate steam quality for most process uses (e.g., strip-
ping steam), but are not adequate for steam turbines. Knockout vessels are occasion-
ally provided immediately upstream of turbines. This arrangement removes
condensate formed in the piping as well as that entrained from the steam generator.
U-tube construction described above is acceptable if the difference between process
inlet and outlet temperature is less than 200F. Otherwise excessive thermal stresses
develop in the tubesheet. Thin, stayed fixed tubesheets are used when thermal
stresses would otherwise be too high. Stayed tubesheets depend on the tubes to
contain the shell side pressure. Stayed tubesheets are governed by the rules in
Section I of the ASME Code.

350 Very High Temperature Waste Heat Steam Generators

High temperature waste heat steam generators are services where the heating gas is
over 1000F and can cause rapid failure of pressure containing parts that are not
continuously water wetted. High temperature in-shell steam generation is discussed
below for horizontal units, and in Section 390 for vertical units. This manual does
not address tube side steam generation. Figure 300-8 is a schematic of a high
temperature horizontal shell side steam generator.
This type of unit is used in hydrogen, ammonia and sulfur plants. Inlet gas tempera-
tures range from about 1500F to 3000F. The bypass pipe shown in Figure 300-8 is
required in hydrogen and ammonia plants to control downstream temperature.
Bypass is not used in sulfur plants.
Shell side circulation is driven by the density difference between the fluid in the
annular gap (mostly liquid) and the lighter two-phase mixture in the bundle (internal
circulation). For best circulation, 90 degree square tube layout is used. The annular
gap should be about 10% of the bundle diameter.

300-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Fig. 300-8 Typical High Temperature Steam Generator

Operation is limited by “dryout” just downstream of inlet tube ferrules near the
center of the bundle. When dryout occurs, the tube temperature jumps to within a
few hundred degrees of process inlet temperature, and the tube may fail within
minutes in very high temperature units. In lower temperature units, local dryout
causes local deposition of solids, restricted water circulation, more extensive dryout,
and eventual tube failure a few months after initial dryout.
Dryout heat flux in submerged bundles increases with process gas temperature and
pressure, and with the ratio of heat transfer surface area to peripheral inflow/outflow
area. Appendix E provides methods to evaluate thermally induced dryout, and
calculate maximum recommended heat flux.
Water concentration is maximum at the same location where dryout is most likely to
occur (central tubes near ends of ferrules). Blowdown should be located close to this
region.
Higher quality water is required in horizontal shell side steam generators with
internal circulation than with external circulation because the maximum percentage
of vapor is higher. Within the recommended operating range (without dryout), water
is about 25% more concentrated than in systems designed for a maximum of 5%
vaporization.

351 Components
Tubesheets
Thin stayed tubesheets are required to avoid excessive thermal stresses. They should
be designed in accordance with ASME Boiler and Pressure Vessel Code, Section I.
Tubesheet thickness is governed by the largest unstayed area, which is usually the
annular space between the bundle and the shell.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-9
300 Service Considerations Heat Exchanger and Cooling Tower Manual

Tube-to-Tubesheet Joints
Tube ends should be rolled and strength-welded to the tubesheet. Roll 95% of the
tubesheet thickness to minimize the crevice between the tubes and the tubesheet.
Rolling at or beyond the tubesheet thickness may damage the tube.

Tube Supports
Tube support spacing and thickness should conform to TEMA rules. Bolt the tube
supports to clips on the shell to avoid cracking due to pressure dilation of the shell.
Provide peripheral cut-outs on the tube supports so axial flow in the annulus is unre-
stricted.

Tubesheet Refractory
The purpose of tubesheet refractory is to protect the tubesheet from high tube side
temperatures. Only about 1 inch of refractory is useful for insulation. More thick-
ness is needed to keep the refractory in place when it cracks. See Insulation and
Refractory Manual for recommendations.

Ferrules
The purpose of ferrules is to protect the tubesheet and seal welds from high temper-
atures, and to provide a hole through the hot tubesheet refractory. Ferrules are
usually high alumina ceramic or an alloy, such as Inconel. Metal ferrules are
preferred, if temperatures allow, because they can be made thinner than ceramic,
and thus reduce the geometry discontinuity between the downstream end of the
ferrule and the tube I.D. Wall thickness of ceramic ferrules should be at least
1/8 inch to facilitate their manufacture.
Both ceramic and metal ferrules should be wrapped with 1/32 inch thick ceramic
paper insulation, glued in place. About 1 inch of the outlet end of metal ferrules
should be expanded to the tube I.D. dimension before insertion into the tube. This
snug fit keeps the ferrule in place during installation of the refractory and keeps the
paper insulation in place during operation.

Downcomers and Risers

A single downcomer is usually sufficient. Multiple risers are usually required, and
should be sized for annular or churn flow. Consider the axial steam production
profile along the unit in selecting riser locations for horizontal units; i.e., put one
riser near the hot tubesheet and shift the others toward the hot end as appropriate.
Equations for calculating the axial steam production profile are given in
Appendix F.

Internal vs. External Bypass

An internal bypass is preferred to an external bypass because internal bypasses are
cheaper and more reliable. If the internal bypass is uninsulated, the surface area
should be included in the dryout heat flux calculations. If dryout is a problem,
provide an insulated internal sleeve, welded on one end only, for the first 2 feet.
Check the thermal stress between the bypass pipe and nearby tubes. If thermal stress
is a problem, the insulated sleeve can be extended for the full length.

300-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Dryout Sensors
If it is necessary to operate a steam generating bundle close to the dryout limit, it
may be advantageous to install dryout sensors. Local dryout usually results in
mechanical failure before reduction in thermal performance is noticed. When dryout
occurs, the tube temperature jumps from very near the water/steam temperature to
near the tube side fluid temperature. This large temperature change can be easily
detected by a thermocouple.
A dryout sensor consists of a thermocouple tack welded to the tube O.D. in the high
heat flux region downstream of the ferrule, with the lead being taken out through a
high pressure fitting in the shell. This can only be done during manufacture or
retubing. Several dryout sensors are usually installed in areas most prone to dryout.

Piping and Instrumentation

Section I of the ASME Code defines instrumentation and piping requirements.

352 Field Modifications

Where thermally induced dryout is a problem, the dryout heat flux can be increased
by judicious removal or plugging of tubes as indicated in Figure 300-9.
Either tube removal or tube plugging provides a low resistance path for vapor free
water to flood the central region of the bundle. Both have been done successfully.
Tube removal requires new tubesheets and added stay rods. Consult the Engi-
neering Analysis Division of ETD for the design of bundle modifications to avoid
dryout.

Fig. 300-9 Modifications to Increase Dryout Heat Flux

360 Reboilers
Reboilers are designed in many configurations, including vertical tube side boiling
thermosiphons, horizontal shell side boiling thermosiphons, submerged horizontal
shell side boiling with internal natural circulation, and horizontal tube side boiling

December 1989  1989 Chevron USA Inc. All rights reserved. 300-11
300 Service Considerations Heat Exchanger and Cooling Tower Manual

with pumped circulation. Vertical tube side boiling thermosiphons and submerged
horizontal shell side boiling with internal natural circulation are the two most
common configurations.
Vertical tube side boiling thermosiphons are normally designed using the HTRI RTF
computer program. This is a mature program that rigorously simulates thermal and
hydraulic performance and flags inappropriate operating conditions, including
choked flow, mist flow, hydraulic resonance, and excessive subcooled boiling
zones.
Horizontal shell side boiling is more efficient than vertical tube side thermosiphon
boiling, but rigorous integrated thermal and hydraulic simulation methods are not
currently available. Horizontal reboilers may be sized and thermally evaluated using
the simple methods given later in this section, or with the HTRI RKH computer
program. In either case, independent design/evaluation of the external circulation is
required.

General Guidelines
The following guidelines apply to all reboiler configurations.
1. The boiling regime should always be nucleate. This is assured if the heat flux is
maintained at 2000 Btu/hrft2 or higher everywhere in the exchanger.
2. Sensible heating medium should be routed cocurrent to the boiling fluid to
facilitate control without fouling. Control by bypassing heat medium results in
a temperature pinch and nil heat transfer near the heat medium outlet, which
should correspond to the high velocity two-phase boiling outlet. Heat medium
inlet near the boiling fluid inlet promotes nucleate boiling in this region and
avoids a stagnant zone on the boiling side where particulate fouling may other-
wise occur.
3. Steam heated reboilers should be controlled by throttling steam pressure and
should have a downstream condensate receiving vessel to keep the reboiler
drained of condensate, and a vent for noncondensables. Otherwise, steam side
corrosion at the condensate interface and boiling side fouling in the nil heat
transfer condensate flooded region are likely.
Specific configuration guidelines are given below.

Vertical Tube Side Boiling

Vertical thermosiphons have single shell and tube passes with expansion provisions
(packed joint or bellows) between the floating head and shell cover. Tube diameter
is usually 1 inch or larger, and tube lengths are usually limited to 8 to 10 feet for
good circulation. Shell side geometry is typical of TEMA “E” shells with 45 degree
tube layout and 20% cut segmental baffles. Baffle spacing is 20 to 50% for sensible
heat mediums, and as needed for vibration control in steam heated units. A vent
near the top tubesheet is needed for steam heated units to prevent accumulation of
CO2 and carbonic acid corrosion of the tubes. Tubes are flush with the top tubesheet
(instead of the usual 1/8 inch protrusion) to facilitate drainage. Target vaporization
is 30 weight percent maximum.

300-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Figure 300-10 shows the typical configuration for steam heated and oil heated
vertical thermosiphon reboilers. Four alignment bars around the bellows permit
axial movement only and protect the bellows from non-axial loads during assembly,
disassembly and handling.

Fig. 300-10 Typical Vertical Thermosiphon Reboilers

December 1989  1989 Chevron USA Inc. All rights reserved. 300-13
300 Service Considerations Heat Exchanger and Cooling Tower Manual

The flow area of the process outlet line should be 100 to 150% of the total tube side
flow area. The size of the process inlet line may be estimated by: (Dpi)2/Q = 9 to 11.
Dpi is the inlet pipe I.D. in inches and Q is the heat duty in millions of Btu/hr.
A packed tail pipe, instead of a bellows, may be used for noncorrosive, nontoxic
heat media below 300F. A packed tail pipe is illustrated for a split ring floating
head (Type “S”) in Figure 400-1 of Section 410.

Horizontal Shell Side Boiling

Most horizontal reboilers have internal circulation with up-flow through a 90 degree
square layout bundle and down-flow in an annular space around the bundle. Internal
circulation is independent of external circulation provided that external circulation
is adequate to keep the bundle flooded. The annular gap around the bundle should
be at least 10% of the bundle diameter. Full tube support plates are used with
cutouts top and bottom to facilitate axial flow within the exchanger.
When vapor-liquid separation in the reboiler is desired, a kettle-type shell is used.
The bundle is located near the bottom of the shell with a gap at the bottom equal to
about 10% of the bundle diameter. There is a much larger space above the bundle to
effect vapor-liquid separation. Full tube support plates are used with cutouts at the
bottom for unrestricted axial flow of liquid. A kettle reboiler provides one theoret-
ical distillation stage.
Horizontal reboilers are sometimes inserted directly into the bottoms compartment
of a column without a shell. This is practical for very small reboilers.
Horizontal shell side reboilers with circulation driven by the difference in fluid
density between the inlet and outlet piping have no advantage over horizontal
reboilers with internal circulation. The boiling-side pressure drop is much higher
than for internal circulation. Horizontal shell side reboilers have been used when
surplus exchangers designed for other services were available that were large
enough for adequate circulation and small enough to assure nucleate boiling
throughout the bundle.

Horizontal Forced Flow Tube Side Boiling

Forced flow boiling is rare. It is used for boiling light components from viscous
mixtures and for polymerizing fluids where higher fluid shear is needed than can be
provided by natural circulation.

Sizing Shell Side Boilers

This section presents a simple graphical method to size or evaluate shell side
reboilers. The graphs are based on the equations in Appendix D. They apply to hori-
zontal shell side hydrocarbon reboilers with internal circulation (i.e., with annular
gap between bundle and shell, in kettles, and for stab-in reboilers without shells).
Figure 300-11 is a plot the of maximum recommended heat flux (qref), boiling heat
transfer coefficient (href), and tube wall-to-boiling fluid temperature difference
([Twall - Tsat]ref) versus bundle diameter, with hydrocarbon critical pressure as a
parameter. This graph applies directly to pure component boiling at a reduced pres-
sure of 0.029 (near atmospheric pressure for hydrocarbons) with isothermal heat

300-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

medium. The heat flux curves include a factor of 0.7 to account for correlation inac-
curacies and an additional 0.8 factor for operating flexibility; 3/4 inch tubes on
1 inch 90 degree square layout is assumed. Figure 300-12 is a plot of correction
factors for heat flux, boiling coefficient, and temperature difference to extend
Figure 300-11 to all pressures.

Fig. 300-11 Reference Boiling Parameters

Figures 300-11 and 300-12 define the minimum required heat transfer area for a
given heat duty, or the maximum recommended heat duty for a given bundle, for
isothermal heat media (e.g., steam heated). To determine the minimum heat transfer
area for a particular heat duty, simply divide the duty by the heat flux from
Figure 300-11 and by the pressure correction factor (Fqp) from Figure 300-12. To
determine the maximum recommended heat duty for a particular bundle, multiply
the heat transfer area by the heat flux from Figure 300-11 and by the pressure
correction factor (Fqp) from Figure 300-12. These procedures are for operation at
56% (0.70.8100) of incipient dryout heat flux.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-15
300 Service Considerations Heat Exchanger and Cooling Tower Manual

Fig. 300-12 Pressure Corrections

Heat transfer coefficients and temperature differences have to be considered to

select or evaluate appropriate heat medium temperatures. Fouling can be compen-
sated for by raising the heat medium temperature. However, the capacity of the
reboiler is still limited by the surface area and the maximum recommended heat
flux.
Figure 300-13 presents correction factors for heat flux and the boiling heat transfer
coefficient when sensible (nonisothermal) heat media are used. Fqsh is the ratio of
maximum to average heat flux along the sensible heat medium flow path. Fhsh
corrects the boiling coefficient for the lower average heat flux. The equations for
Fqsh and Fhsh are given below.

Fqsh = (1-e-N)/N
where:
N = U A/(M Cp), dimensionless
U = Overall heat transfer coefficient
A = Heat transfer area
M = Mass flow rate of heat medium
Cp = Specific heat of heat medium
Fhsh = Fqsh2/3
This correction results in a maximum heat flux (at the heat medium inlet) equal to
56% of incipient dryout heat flux.
Figure 300-14 is a boiling range (dew point to—bubble point) correction factor, Fhc,
for multicomponent boiling mixtures that applies to the boiling heat transfer coeffi-
cient.

300-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Fig. 300-13 Corrections for Sensible Heat Medium

Fig. 300-14 Correction for Multi-Component Mixtures

Correction for Multi-Component Mixtures

If the feed to the reboiler is subcooled, the following additional correction to the
boiling coefficient is needed.
Fhsc = (1 - Qsc/Qt)2/3
where Qsc and Qt are subcooled duty and total duty, respectively.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-17
300 Service Considerations Heat Exchanger and Cooling Tower Manual

The maximum recommended heat flux is:

qdmax = qref  Fqp  Fqsh
The corresponding boiling heat transfer coefficient is:
hb = href  Fhp  Fhsc  Fhsh  Fhc
The tube wall to boiling fluid temperature difference is:
(Twall - Tsat) = (Twall - Tsat)ref  Ftp  qdmax/hb
The required heat medium temperature depends on the heat medium heat transfer
coefficient, hi, inside the tubes and on the tube wall resistance, rw. The overall clean
heat transfer coefficient is
Uc = 1/[1/hb + rw + (1/hi)(0.75/0.56)]
where (0.75/0.56) is the ratio of tube O.D. to I.D.
The required mean temperature difference (MTD) between the heat medium and the
boiling fluid is:
MTD = qdmax/Uc
If steam is the heat medium, MTD is the temperature difference between the steam
and the boiling fluid, Tstm - Tsat. The heat medium is selected to provide a tempera-
ture driving force greater than or equal to that indicated above. However, the
reboiler should not be operated above the recommended maximum heat flux.
For 3/4 inch tube O.D. (dto) on 1 inch square pitch, there is one tube per square inch
of bundle cross-sectional area. Ignoring tie rods and pull holes, the cross-sectional
area of the bundle in square inches is numerically equal to the number of tube holes
in the tubesheet (nt). The bundle diameter can be approximated as
Db = (4nt/)0.5 inches
and the heat transfer surface per unit length as
A/L = nt dto/12 ft2/ft

Example
Consider boiling normal pentane using steam heat.
Duty Q = 14.8  106 Btu/hr
Pentane properties:
Operating pressure P = 100 psia
Critical pressure Pc = 489 psia
Reduced pressure Pr = P/Pc = 0.20
Saturation temperature Tsat = 220F

300-18  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Assume bundle diameter, Db = 20 inches

Number of tubes nt = (20)2/4 = 314
Area/length A/L = 314 (0.75/12) = 61.7 ft2/ft
From Figure 300-11, qref = 16,000 Btu/hrft2
From Figure 300-12, Fqp = 1.5
Maximum design heat flux qdmax = 1.5(16,000) = 24,000 Btu/hrft2
Required area A = Q/qdmax = 14.8  106/24000 = 617 ft2
Required length L = A/(A/L) = 617/61.7 = 10 ft
Many combinations of bundle diameter and length would work. Even integer
lengths from 10 to 20 feet are standard. Since the above calculated length is reason-
able, the assumed bundle diameter is OK.
From Figure 300-11, href = 1000 Btu/hrFft2
From Figure 300-12, Fhp = 2.5
The boiling coefficient hb = 2.5(1000) = 2500 Btu/hrFft2
The steam side coefficient (Section 216) hi = 1000 Btu/hrFft2
Tube wall resistance (Section 212):
rw = 0.75  ln(0.75/0.56)/(30  2  12) = 0.0003 hrFft2/Btu
The overall clean coefficient (Section 212):
Uc = 1/[1/2500+.0003+(1/1000)(0.75/0.56)] = 491 Btu/hrFft2
The required steam temperature is 220 + 24000/491 = 268.9F
corresponding to a steam pressure of 41.1 psia.
The steam temperature required to achieve 125% of design duty is
220 + 1.25(24000/491) = 281F, corresponding to a steam pressure of 50.0 psia.
Operating at higher duties would risk partial dryout and plugging of the bundle.
The nearest available steam pressure that exceeds the required steam pressure would
normally be selected, and throttled as needed to control duty.
The minimum heat flux to assure nucleate boiling is about 2000 Btu/hrft2. 
Operation below this heat flux may result in fouling and plugging the bundle. The
allowable turndown for the above example is 12:1.

External Circulation
Internal circulation through and around the bundle is independent of external circu-
lation between the reboiler and the column, provided that sufficient liquid flows
from the column to keep the bundle flooded. The piping is usually sized to limit
vaporization to about 30% of the total external natural circulation.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-19
300 Service Considerations Heat Exchanger and Cooling Tower Manual

The piping should be designed to minimize elbows and horizontal runs. The liquid
feed line is usually sized for a liquid velocity about one third the economic velocity
for pumped liquid. The exit pipe is usually sized so the flow regime is annular but
near the transition to slug flow, at maximum duty operation. This maximizes
external circulation. Larger exit piping can cause slug flow with reduced static
differential head and reduced circulation. Smaller exit pipes can cause excessive
friction losses and reduced circulation.
Detailed analysis is required to assure satisfactory circulation. The following rules
of thumb may be used for initial estimates.
Inlet line: Dpi2 = 9 to 11
Outlet Line: Dpo5/Q2=1800 to 2500
Where Dpi and Dpo are in inches and Q is in millions of Btu/hr.
The HTRI RKH computer program may also be used to rate or design horizontal
reboilers. The current version is not rigorous and should be used with caution.

370 Condensers
Condensers are normally designed using the HTRI CST computer program for shell
and tube exchangers and the HTRI ACE computer program for air coolers. At very
low pressures, vacuum equipment performance significantly affects condenser
design. Vacuum systems are usually designed as a package by vendors that
specialize in this type of equipment.
Condenser configuration is usually dictated by the coolant. Air-cooled condensers
are horizontal with in-tube condensation. Water-cooled condensers normally have
the water tube side to be consistent with water treating practice. Viscous coolants
(e.g., crude oil) should be in the shell for best heat transfer.
Condensing side pressure drop may govern the condenser configuration at low pres-
sures. Pressure drop in condensers is typically 10% of the absolute pressure or 5 psi,
whichever is less. A single down-flow shell side pass over a horizontal bundle
(e.g., TEMA X shell) provides the lowest pressure drop. The effect of pressure drop
on the condensing temperature profile should be considered.
Wide condensing range mixtures should be kept well mixed along the condensing
path. Separation of liquid and vapor reduces heat transfer coefficients and tempera-
ture driving force. Considerable pressure drop is required to keep phases mixed in
horizontal shell side condensers with side-to-side flow. Single pass down-flow
through either the shell side or the tube side requires the least pressure drop to keep
phases mixed. Divided horizontal flow in the shell (e.g., TEMA J shell) or hori-
zontal in-tube condensation are next best. Pure component condensation coefficients
(e.g., steam) are usually very high and insensitive to configuration and flow regime.
Condensate receiver vessels below the condenser are usually provided to keep the
condenser well drained of condensate and maintain good performance. Condensate
receiver vessels and most condensers should have vents at high points to remove
noncondensables. Noncondensable accumulation in the condenser reduces heat

300-20  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

transfer and may cause corrosion in some services (e.g., steam). Vertical down-flow
in-tube partial condensers do not require periodic venting.
Subcooling the condensate, when required, is usually best accomplished in a sepa-
rate bundle where reasonable flow at the heat transfer surface can be obtained.
Subcooling is also practical in single pass down-flow in-tube partial condensers
where reasonable condensate flow in contact with the heat transfer surface occurs.

380 Crude Unit Heat Exchangers

Several interrelated crude unit heat exchanger problems are addressed in this
section. Crude-preheat-exchanger fouling is a problem in some crude units. Crude
side fouling control in the crude-residuum exchanger affects residuum fouling
control. Corrosion control in the atmospheric overhead condenser may affect the
performance of the condenser as well as the desalter water exchanger and the crude
preheat exchangers.

Crude Preheat Exchangers

Crude preheat exchangers are subject to salt-precipitation fouling, filming amine
fouling, and particulate fouling. Chemical reaction fouling of natural crude oil
components occurs at temperatures (600+ F) above those encountered in crude
preheat exchangers. Crude preheat exchanger fouling can be prevented by main-
taining adequate velocities to keep particles moving, preventing salt-precipitation
conditions in the heat exchangers, and avoiding film forming and water-oil stabi-
lizing additives in the crude oil. Figure 300-15 and the associated notes give the
criteria to prevent crude oil fouling.
Crude-Residuum Exchangers
Residuum exchangers are subject to chemical reaction fouling (“coking”) and
particulate fouling. Residuum fouling can be prevented by operating exchangers
near the economic velocity and controlling tube surface temperatures below the
threshold of chemical reaction fouling. Figure 300-16 illustrates the normal method
of controlling residuum fouling.
The heat transfer coefficient on the crude side is usually higher than on the
residuum side. The tube wall temperature is therefore closer to the crude tempera-
ture than to the residuum temperature. Quenching the residuum to 650F results in a
maximum tube wall temperature less than 600F, if there is no crude fouling.
The threshold fouling temperature of residuum varies with crude source and flash
zone temperature. Normal pump overdesign (125% of design flow) is usually
adequate to compensate for crude variations. The optimum operating temperature is
just below the threshold temperature for fouling. If crude-side fouling is allowed to
occur, the residuum would have to be reduced another 50F or so to prevent
residuum-side fouling.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-21
300 Service Considerations Heat Exchanger and Cooling Tower Manual

Fig. 300-15 Crude Oil Fouling Control

(1) Maintain exchanger velocities at half the economic velocity or higher. This is equivalent to maintaining shell and tube
side friction pressure gradients over 0.2 psi/ft (axial) and 0.1 psi/ft, respectively.
(2) Design heat exchangers upstream of the desalter with tube wall temperatures less than or equal to 350F.
(3) Maintain back pressure on the exchangers upstream of the desalter over the vapor pressure of water-saturated
crude oil at the maximum tube wall temperature (350F).
(4) Inject desalter water into cold crude oil at a minimum rate of 2 liquid volume percent of crude oil flow rate. More is
preferred.
(5) Maintain aqueous phase pH in the exchangers and desalter between 5.5 and 6.5. This may require acid injection
into cold crude for some very alkaline crudes. Careful design of injection and mixing equipment is required.
(6) Maintain flash drum bottoms temperature at least 10F above the water boiling point at flash drum pressure.
(7) Avoid film forming and water oil stabilizing additives in crude oil feed and desalting water.
(8) Caustic injection, if any, should be downstream of the last preheat exchanger.

Fig. 300-16 Residuum Fouling Control

Atmospheric Overhead Condenser

Crude unit overhead condensers usually exchange heat with viscous crude oil. The
crude oil should be on the shell side with a 45 degree tube layout. This arrangement
results a seven-fold improvement in thermal performance relative to viscous oil in
the tubes. (See Figure 200-4 in Section 213.)

300-22  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Atmospheric overhead exchange is subject to fouling on the condensing side by

salt-precipitation fouling (NH4Cl) and particulate fouling (from condenser corro-
sion products). Condenser-side fouling can be eliminated with saturation water
recycle shown in Figure 300-17. Saturation water recycle is consistent with good
corrosion control as discussed in the Corrosion Prevention and Metallurgy Manual.
Fig. 300-17 Fouling (and Corrosion) Control in Crude Unit Atmospheric Overhead Condensers

Vertical single pass down-flow condensers are preferred to obtain high condenser-
side heat transfer with low pressure drop. An expansion bellows is provided
between the floating head and the shell cover.
Filming amines are commonly injected into crude unit overhead systems. They are
present in the naphtha and are dispersed to other distillate streams via the top reflux,
and to the desalter water exchanger and to crude feed via the net overhead water.
The presence of filming amine can significantly affect thermal performance of
downstream exchangers.
Filming amine fouling occurs rapidly and then levels out (i.e., is asymptotic) at
moderate temperatures. The observed amine films are mostly particulates bound
together by the amine. Above the amine decomposition temperature (usually 300F
to 400F), nonasymptotic filming amine fouling has been observed.
Moderate temperature asymptotic filming amine fouling is shown in Figure 300-18
for amine contaminated desalter water and crude oil.
Where it occurs, filming amine fouling is usually the dominant thermal resistance in
heat exchangers.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-23
300 Service Considerations Heat Exchanger and Cooling Tower Manual

Fig. 300-18 Filming Amine Fouling Resistance for Hydrocarbon and Water Streams below
350F

390 FCC Flue Gas Coolers

Flue gas coolers in Fluid Catalytic Crackers (FCCs) cool particle-laden flue gas
from the regenerator. The flue gas flows downward in a single tube pass. Natural
circulation steam generation occurs in the shell. Rapid gas-side fouling has been a
problem in some units. Dryout on the steam-side near the top of the exchanger has
also been a problem in a few units.
Flue gas from single stage cyclones contain enough large particles to keep the small
particles scrubbed off the tubes at gas velocities of about 100 ft/s. With two-stage
cyclones, about 200 ft/s is required too keep tubes clean.
Several FCC units have two-stage cyclones and flue gas coolers that operate at
about 100 ft/s. Flue gas cooler duty falls at a rate of about 30% per shift. This
fouling can be controlled by injecting walnut shells into the cooler twice a shift. The
shells scrub the fine solids off the tubes before burning to extinction. Without
fouling control, these coolers quickly become about half plugged and remain in that
condition until mechanically cleaned.
Steam side dryout problems in a few units have been associated with poor hydraulic
design. Figure 300-19 illustrates the configuration designed to keep all steam-side
components water wetted.

300-24  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 300 Service Considerations

Fig. 300-19 FCC Flue Gas Cooler

The donut baffle near the top of the cooler is located and sized so that inertia of the
liquid flowing up through the donut hole is sufficient to impact on the top tubesheet.
The peripheral weir encourages liquid hold-up above the donut baffle. The high
velocity jet over the weir keeps the outer annular region well mixed (a vent is not
required).
Tube ferrule ends are located so that the high heat flux zone just downstream of the
ferrules is in a well-wetted region. These units have been designed to operate with
flue gas temperature up to 1400F. Engineering Analysis Division of ETD can assist
in developing specific designs.

December 1989  1989 Chevron USA Inc. All rights reserved. 300-25
300 Service Considerations Heat Exchanger and Cooling Tower Manual

300-26  1989 Chevron USA Inc. All rights reserved. December 1989
400 Shell and Tube Exchanger Design
and Selection

Abstract
This section contains information on TEMA nomenclature, selecting the most
economic exchanger configuration for a defined service, allocating the streams to
shell or tube side, specifying appropriate mechanical components, defining baffle
layout, deciding if a small predesigned exchanger is appropriate, and estimating the
size and cost of shell and tube exchangers.

Contents Page
410 TEMA (Tubular Exchanger Manufacturers Assoc.) 
Nomenclature 400-2
420 General Design Considerations 400-7
430 Stream Placement 400-11
440 Pass Arrangements and Multiple Shells 400-12
450 Bundle and Tubesheet Arrangements 400-13
451 Front Head Design
452 Fixed Tubesheets
453 U-tubes Versus Floating Rear Heads
454 TEMA F Shell
460 Shell Side Baffle and End Spaces 400-15
470 Small Exchangers 400-16
480 Estimating Methods 400-17
481 Step by Step Procedure
482 Surface Area Calculations
483 Tube Count and Number of Tube Passes
484 Shell Diameter
485 Exchanger Investment Cost

December 1989  1989 Chevron USA Inc. All rights reserved. 400-1
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

410 TEMA (Tubular Exchanger Manufacturers Assoc.) Nomenclature

The Tubular Exchanger Manufacturers Association (TEMA) has developed nomen-
clature for describing shell and tube heat exchangers. It includes a simple code for
designating the size and type of the exchanger. In addition, standard terminology
has been set up to specify typical parts and connections.
TEMA size is the shell inside diameter in inches rounded to the nearest integer,
followed by the straight length of the tubes in inches rounded to the nearest integer.
The two dimensions are separated by a hyphen (-).
For kettle reboilers, the port diameter in inches precedes the shell inside diameter.
The two dimensions are separated by a slash (/). Port diameter is the size of the
opening the bundle slides through.
TEMA type consists of three letters describing the stationary or front end head,
shell, and rear head, in that order. The letter designations are shown on
Figure 400-1.
For example, a 20-foot straight length U-tube bundle, 3-foot shell diameter, with a
single shell pass and removable shell cover would be a TEMA SIZE 36-240 TYPE
AEU. The same bundle installed in a 5-foot diameter kettle reboiler would be a
TEMA SIZE 36/60-240 TYPE AKU.
Standard terminology to describe components and connections of shell and tube
exchangers is provided in Figure 400-2.
TEMA sets mechanical standards for three classes of exchangers reflecting the
severity of the service. For most refinery services, the most restrictive class is
used—TEMA Class R. For other services (chemical plants for example), TEMA
Class C or B exchangers are used. In general, Class R exchangers have thicker
shells, larger and thicker heads, thicker tubes, and larger miscellaneous parts.
TEMA requirements are noted where appropriate throughout this manual.

400-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

Fig. 400-1 Heat Exchanger Nomenclature (TEMA, Figure N-1.2) (Courtesy of TEMA)

December 1989  1989 Chevron USA Inc. All rights reserved. 400-3
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Fig. 400-2 Heat Exchanger Components (1 of 3) (TEMA, Table N-2 and Figure N-2) (Courtesy of TEMA)
1. Stationary Head—Channel 21. Floating Head Cover—External
2. Stationary Head—Bonnet 22. Floating Tubesheet Skirt
3. Stationary Head Flange—Channel or Bonnet 23. Packing Box
4. Channel Cover 24. Packing
5. Stationary Head Nozzle 25. Packing Gland
6. Stationary Tubesheet 26. Lantern Ring
7. Tubes 27. Tierods and Spacers
8. Shell 28. Transverse Baffles or Support Plates
9. Shell Cover 29. Impingement Plate
10. Shell Flange—Stationary Head End 30. Longitudinal Baffle
11. Shell Flange—Read Head End 31. Pass Partition
12. Shell Nozzle 32. Vent Connection
13. Shell Cover Flange 33. Drain Connection
14. Expansion Joint 34. Instrument Connection
15. Floating Tubesheet 35. Support Saddle
16. Floating Head Cover 36. Lifting Lug
17. Floating Head Flange 37. Support Bracket
18. Floating Head Backing Device 38. Weir
19. Split Shear Ring 39. Liquid Level Connection
20. Slip-on Backing Flange

400-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

Fig. 400-2 Heat Exchanger Components (2 of 3) (TEMA, Table N-2 and Figure N-2) (Courtesy of TEMA)

December 1989  1989 Chevron USA Inc. All rights reserved. 400-5
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Fig. 400-2 Heat Exchanger Components (3 of 3) (TEMA, Table N-2 and Figure N-2) (Courtesy of TEMA)

400-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

420 General Design Considerations

Single- and two-phase exchangers and most condensers have very similar configu-
rations. The typical layout is summarized in the following list and shown in Figures
400-3 and 400-4. (Steam generators (2 types), reboilers, and condensers are
described in Sections 340, 350, 360 and 370.)
The typical shell and tube exchanger geometry includes the following items:
• TEMA E shell style
• U-tubes for rear head type with full support plate at tangent
• TEMA A-type front head
• Single segmental baffles with cut of 18 to 25% of shell I.D. and with cut
oriented vertically
• Baffle spacing of 20 to 100% of shell I.D
• 3/4-inch O.D., 14 BWG (average) thickness (0.584 inch I.D.) carbon steel tubes
• Tube length variable with one or two tube passes depending on service
• 45 degree rotated square layout with tube pitch = 1.25  tube O.D. for liquid
and two-phase hydroprocessing shell side service
• 90 degree square layout with tube pitch = 1.25  tube O.D. for boiling,
condensing, and single-phase gas shell side service
• Two or more pairs of sealing strips (bars)
• Dummy tubes in pass partition lane when two tube passes
• Two rows of impingement rods at inlet nozzle when warranted

December 1989  1989 Chevron USA Inc. All rights reserved. 400-7
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual
Typical Longitudinal Section Shell and Tube Exchanger
Fig. 400-3
400-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

Fig. 400-4 Typical Cross Section, Shell and Tube Exchanger

Overall Exchanger Configuration

The Company preference is a TEMA AEU exchanger for most services. U-tubes are
the cheapest rear head type that allows for thermal expansion of the tubes. The
TEMA A type front head has a removable channel cover. This allows for inspection
and cleaning of the tube side without pulling spool pieces in the piping.

Shell Side Nozzle Placement

Single inlet and outlet shell side nozzles are normally located at opposite ends of the
exchanger with one on the top and one on the bottom of the shell. This arrangement
allows vents and drains to be located in piping.
Route two-phase flow based on the following rule: “Heat up and cool down.” This
means hot fluid being condensed should enter on the top and exit on the bottom of
the exchanger. Likewise, cold fluid being boiled should enter on the bottom and exit
on the top. The “Heat up and cool down” rule does not apply to single-phase flow.

December 1989  1989 Chevron USA Inc. All rights reserved. 400-9
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Transverse and Support Baffles

The normal configuration for the tube side consists of U-tubes with a full support
plate at the tangent. This is shown in Figure 400-3. The plate blocks flow over the
U-bends. Otherwise, the bends must be supported to protect against vibration.
For baffles, use single segmental baffles with a cut of around 18 to 25% of the shell
I.D. for most efficient conversion of pressure drop to heat transfer. The baffle cut
should be vertical for best drainage of the shell side at shutdown. Baffle thickness is
set by TEMA.
Baffle spacing should be 20 to 50% of the shell I.D. It is usually set to maintain
good heat transfer (economic pressure gradient or shear controlled flow regime).
Guidelines for economic exchanger velocity and pressure drop are provided in
Section 220 of this manual. In some cases (particularly for gas and two-phase flow
shell side), additional supports may be required to prevent vibration. See
Section 260 of this manual for more information.

Tube Selection
Tubes are normally 3/4-inch outside diameter, 14 BWG (minimum) thickness
(0.56-inch inside diameter), and made of carbon steel. Length is limited by the plot
space for pulling the bundle and standard bundle pulling equipment. TEMA has
named 8, 10, 12, 16 and 20 feet as standard tube lengths. Other lengths are possible.
Alloy tubes are appropriate for some services. The cost of upgrading to alloy tubes
should always be weighed against possible process adjustments to permit carbon
steel construction. Section 800 of this manual discusses materials selection for
different services.

Tubepass Layout
Most exchangers should be limited to one or two tube passes. Using U-tubes with
two passes is best and cheapest, however some services dictate 1 pass with a more
expensive rear head (vertical thermosiphon reboilers or crude/overhead condensers,
for example).

Tube Pitch
For liquid and two-phase services, use 1-inch, 45 degree rotated square pitch. This
promotes mixing. Use 1-inch, 90 degree square pitch for boiling, condensing, and
single-phase gas on the shell side. For boiling, the vertically oriented lanes promote
circulation. For condensing and single-phase gas, in-line tubes minimize pressure
drop without sacrificing heat transfer. Both 45 and 90 degree pitch provide
0.25-inch inspection and cleaning lanes through the bundle.

Preventing Shell Side Flow Bypassing

Single- and two-phase exchangers with impingement protection typically include
two pairs of sealing strips (bars). The bars block the leakage stream flowing around
the baffles between the bundle and shell (“C” stream shown in Figure 200-3 in
Section 213). For vertical cut baffles, the bars straddle the nozzles (located at the

400-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

top and bottom of the bundle). Note that the bars on the bottom act as skid bars for
bundle removal.
For an exchanger with two tube passes, the single pass partition lane runs perpen-
dicular to the baffle cuts. Dummy tubes are positioned in the pass partition lane to
block flow bypassing (“F” stream shown in Figure 200-3 in Section 213). Dummy
tubes are spaced four to six tube rows apart between baffle cuts and are the same
diameter as the tubes.

Impingement Protection
When impingement protection is warranted, the preferred method is to install two
rows of rods (typically tubes over solid rods) adjacent to the inlet nozzle.
Section 525 contains design details and applications of impingement rods along
with descriptions of other types of impingement protection.

Tolerances and Clearances

All tolerances and clearances are TEMA.

430 Stream Placement

Allocating the streams to the shell or tube side is determined by weighing factors
which sometimes conflict. These factors include stream temperature, pressure, rela-
tive flowrate, viscosity, corrosiveness, relative heat transfer film coefficient, and
pressure drop limitations. Guidelines for allocating the streams to the shell or tube
side are given in Figure 400-5.

Fig. 400-5 Allocating the Streams for Shell and Tube Heat Exchangers (1 of 2)
In Order of Decreasing Priority:
Preferred Side
Stream Property 
Compared to Other Stream Shell Tube Reasons for This Choice
Match Coefficients and Pumping — — Minimize cost
Power
Lower Film Coefficient Expected X Enhance outside surface to raise
(hshell / htube <0.3) limiting side coefficient (single-phase
gas only)
Condensing — — Determined by coolant
Treated Cooling Tower Water X Corrosion inhibitors effective tube-
side; otherwise use alloy tubes
Viscosity above 2 cP X Staggered tube layout induces good
heat transfer at low Reynold’s number
Alloy Required for Corrosion X Allows cheaper shellside components
Very Low System Pressure or P X Can use J or X shell style to shorten
Available flow path and reduce pressure drop

December 1989  1989 Chevron USA Inc. All rights reserved. 400-11
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Fig. 400-5 Allocating the Streams for Shell and Tube Heat Exchangers (2 of 2)
In Order of Decreasing Priority:
Preferred Side
Stream Property 
Compared to Other Stream Shell Tube Reasons for This Choice
High System Pressure X Reduces shell thickness; however,
tube rupture design sometimes
controls
High T across one Bundle (Over X Excessive T in stationary tubesheet
200°F) if placed on tubeside
Normal Fouling — — Does not matter
Deposits Too Hard to  X Use floating rear head for straight
Hydroblast (Rare) tubes
Complete Tube Plugging (Rare) X Use floating rear head for straight
tubes

440 Pass Arrangements and Multiple Shells

The appropriate stream pass arrangements for a particular service are based on:
• Providing economic pressure gradient on both sides of exchanger
• Operating in shear controlled flow regime for two-phase flow
• Limiting pressure drop
• Controlling temperature efficiency
On the tube side, the pressure gradient is adjusted by changing the number of tubes
per pass. To get more area, increase the flow path length either by using longer
tubes, by adding more shells in series, or by increasing the number of tube passes.
Note that two tube passes are typical because more passes dramatically increase
pressure drop. Not only does the pressure drop increase proportionally to the
increase in flow path length, but to the square of velocity. For example, going from
two to four passes increases the pressure drop by a factor of eight with the tube
count held constant.
On the shell side, the pressure gradient is adjusted by changing the baffle spacing.
To get more area, the exchanger (tube length) is made longer. When more area is
needed and the tube length is maximum, add another shell with the shell side flow
in series.
The shell style is changed from a TEMA E-type to a TEMA J- or X- type when the
resulting pressure drop is too large at the target pressure gradient. This shortens the
flow path allowing the pressure gradient to be maintained.
Use parallel exchangers only when a single exchanger is too large, and the pressure
drops can not be increased at the target pressure gradients. Exchanger size is limited
by the manufacturer’s fabricating equipment and the user’s maintenance equip-
ment. Space availability may also limit size, especially when modifying an existing
unit.

400-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

Parallel units with isolation valves have been used to provide an installed spare or
when flow rates will vary more than 50% from normal. When the flow rate varies,
the number of units onstream is changed to maintain reasonable operating pressure
drop.
Consider using a mixed parallel/series arrangement of shell and tube passes in
multiple units only when required to meet pressure drop restrictions. The overall
temperature efficiency of the units is reduced. Note that the F-factor described in
Section 211 of this manual is the common measure of temperature efficiency.
Temperature efficiency will vary with service. Area is most effectively used when
shell and tube side stream routing approaches pure countercurrent flow (F-factor of
1.0). Going to multiple units in series increases the temperature efficiency. Keep the
F-factor above approximately 0.85.
When performance is limited by a temperature pinch between the streams (small
local temperature difference reflected as low F-factor), multiple shells become cost
effective by reducing the total area requirement. Countercurrent flow of both fluids
through the shells maximizes efficiency.
For condensing services, significant subcooling loads are usually processed in a
separate exchanger following the condenser. This allows the geometry to be
changed to accommodate the much lower volumetric rate of the liquid. As a result,
the area needed for subcooling is reduced.

450 Bundle and Tubesheet Arrangements

This section covers front head selection, fixed tubesheet applications, U-tubes
versus floating rear heads, and TEMA F shells (two shell pass exchangers).

451 Front Head Design

The TEMA Type A front stationary head is normally used. It has a removal channel
cover so the tube side can be inspected without disconnecting nozzles or removing
pipe spools. The bonnet channel (Type B) is cheaper and is appropriate for small
exchangers with small easily removed pipe spools. For operating pressures above
1000 psig, a special front head is required. Options are discussed in Section 532.

452 Fixed Tubesheets

Fixed tubesheets are the cheapest type of head. They are typically used when the
shell side service is nonfouling and noncorrosive, and the metal temperature of the
shell and tubes operate within 50°F (including startup, shutdown and steam out
conditions). The bundle is not removable.
The shell side is not accessible for inspection or mechanical cleaning since the
tubesheets are seal welded to the shell. If the temperature difference is larger than
50°F, an expansion joint may be required in the shell.

December 1989  1989 Chevron USA Inc. All rights reserved. 400-13
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Steam generators with very high (1000°F and above) process side temperatures and
water on the shell side must have fixed tubesheets. See Section 350 of this manual
for more information.

453 U-tubes Versus Floating Rear Heads

U-tube and floating head bundles are removable. Both permit thermal expansion of
the tubes. The various types of rear heads are shown on Figure 400-1.
U-tubes (TEMA Type U) are the cheapest of the two types and are preferred. The
bends can be mechanically cleaned by hydroblasting for typical fouling deposits—
as long as complete plugging does not occur.
One disadvantage of U-tube bundles is that corrosion is difficult to monitor. Spec-
imen tubes can only be taken from the outside perimeter of the bundle.
TEMA Type S and T floating rear heads cost more than U-tubes. Maintenance is
complicated by the added bundle flange. Floating heads can be taken apart and the
straight tubes drilled out. Floating heads are recommended for services leaving
deposits too hard to hydroblast.
The differences between the S- and T-type heads are minor. The split ring (S) type
allows for tight clearance between the shell and bundle. However, a shell body
flange and the split ring flange must be taken apart before the bundle can be pulled.
The pull through (T) type allows the bundle to be removed prior to taking apart the
floating head and does not require a shell body flange. However, the shell is over-
sized to allow the floating head to pass through.
Floating heads (versus fixed tubesheets) are usually necessary for single tube pass
exchangers to accommodate thermal expansion. Head design must account for
startup, shutdown, and steam out conditions. Single pass exchangers with a floating
head are commonly used for atmospheric column overhead condensers in crude
units and vertical thermosiphon reboilers.

454 TEMA F Shell

The TEMA F shell has a longitudinal baffle running through the middle of the
exchanger. This provides two shell passes within one shell. Both the inlet and outlet
shell side nozzles are located adjacent to the tubesheet (channel end).
When coupled with two tube passes, the F shell provides pure countercurrent flow.
F shells have been used instead of multiple shells in series to avoid temperature
pinches. F shells are cheaper than multiple shells in series. However, experience has
shown the seal between the two shell passes to be very difficult to maintain.
Increased maintenance time and performance loss due to leakage by the longitu-
dinal baffle is reported frequently.
As a result, TEMA F shells are currently recommended for noncorroding and
nonfouling services only—where the tube bundle is rarely if ever pulled for mainte-
nance.

400-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

If the bundle from a F shell is pulled, the seal (described in Section 524) is usually
replaced. The bundle must be handled carefully when reinstalled. The seal is easily
ruined if the slings twist the seal or the bundle goes in crooked.

460 Shell Side Baffle and End Spaces

The number of crosspasses, the baffle spacing (central, inlet and outlet spacing), and
the straight tube length are related mathematically. For a TEMA E shell with
U-tubes fully supported at the bend tangent, the following relationship applies.

F = C (cp - 2) + D + E + tbst
(Eq. 400-1)
where:
F = Straight (total) tube length in inches
C = Central baffle spacing in inches
cp = Number of crosspasses per shellpass
D = Inlet baffle spacing in inches
E = Outlet baffle spacing in inches
tbst = Tubesheet thickness in inches
End (inlet and outlet) spaces are set to keep the transverse baffles clear of the inlet
and outlet nozzles. The spacing accounts for mechanical constraints which force the
nozzle position. These include flange thickness, body and nozzle flange clearances,
nozzle reinforcement and access. For a TEMA E shell with U-tubes, end spaces can
be estimated using the following equations.
End space at channel or tubesheet in inches:
1.1 (nozzle I.D., inches) + 0.1 (shell I.D., inches) + 8.0
End space at rear end or free end of bundle in inches:
1.1 (nozzle I.D., inches) + 2.0
The actual spacing can be wider, but should not be excessive. Heat transfer in the
end spaces is not as good as between transverse baffles.

December 1989  1989 Chevron USA Inc. All rights reserved. 400-15
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

470 Small Exchangers

There are two types of small exchangers: the double pipe and the multitube hairpin.
Both are predesigned in set configurations, and provided by vendors off the shelf.
They are designed to be stacked nozzle to nozzle as shown in Figure 400-6.

Fig. 400-6 Typical Stack of Small Exchangers

Figure 400-7 is diagram of a double pipe exchanger. It is simply a single pipe within
a pipe. Fluid flow on the shell side simplifies to flow through an annulus.
Figure 400-8 is a diagram of a multitube hairpin exchanger. It is a shell and tube
exchanger with one U-shell and one U-tube pass. Figure 400-9 gives typical
exchanger geometries.

Fig. 400-7 Double Pipe Exchanger Fig. 400-8 Multitube Hairpin Exchanger

The same economic considerations for setting pressure gradient or velocity apply to
small exchangers as to conventional shell and tube exchangers. Small exchangers
are cost effective when the required surface area is less than about 250 ft2 for double
pipes and less than 1000 ft2 for multitube hairpins.

400-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

Because the configuration is already fixed, you should confirm exchanger geometry
with the vendor. The HTRI programs can be used to model double pipe and multi-
tube hairpin exchangers. See the Heat Exchanger Design Program User’s Guide for
details.

480 Estimating Methods

This section gives procedures for estimating the size and cost of a shell and tube
exchanger. The procedures are recommended for:
• Preliminary sizing and layout of a new exchanger prior to rigorous computer
modeling
• Developing project economics
• Comparing performance or configuration of an existing exchanger to a defined
standard or baseline exchanger

December 1989  1989 Chevron USA Inc. All rights reserved. 400-17
 400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual
Multitube Hairpin Exchanger Information (1 of 2)
Fig. 400-9
400-18  1989 Chevron USA Inc. All rights reserved. December 1989
 Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection
Multitube Hairpin Exchanger Information (2 of 2)
Fig. 400-9
December 1989  1989 Chevron USA Inc. All rights reserved. 400-19
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

481 Step by Step Procedure

Step 1. Estimate physical and thermal properties for streams. Calculate exchanger duty (MMBtu/hr). Allocate
streams to shell and tube sides using the guidelines in Section 430.
Step 2. Plot heat release curve for stream(s) undergoing phase change. Estimating techniques must be carefully
applied to streams with dramatic slope changes. More information and an example are provided in
Section 211.
Step 3. Estimate the actual mean temperature difference (MTD). This depends on the service of the exchanger. See
Section 211.
Step 4. Determine if multiple shells in series are required. See Section 440.
Step 5. Select appropriate film coefficients from Figure 400-10. Appropriate sections of this manual that contain
more accurate methods are referenced in Figure 400-10.
Step 6. Calculate a overall service heat transfer coefficient U. See Section 212.
Step 7. Calculate the required surface area for the service. See Section 482.
Step 8. Determine the number of tubes per shell and pass configuration. See Section 483.
Step 9. Estimate the shell diameter for given tube count. See Section 484.
Step 10. Estimate shell and tube side pressure drop, if needed. See Section 220.
Step 11. Cost the exchanger, if needed. See Section 485.

Fig. 400-10 Approximate Heat Transfer Film Coefficients for a Well Designed Heat Exchanger(1) (2)
Shell or Tube Side Coefficient,
Btu/hrFft2 [based on bare
Service or Fluid outside area] Reference
SENSIBLE
Pure Water 1400
C.T. Water(3) 450 Figure 200-4
HC, 0.5 cP 400 Section 213
HC, 2 cP 250
HC, 10 cP 150
GASES
Light HC, 150 psig 100
Air, 10 psig 15 Appendix B
Air, 300 psig 60
CONDENSING
Steam 1000
Light HC 200 Section 370
Heavy HC 100
Subcooling(4) 50
BOILING
Water 1000
Light HC 300 Section 360
Heavy HC 150
AIR COOLED (FIN FAN)
Air Side(5) 175 Section 600
(1) This table applies to well designed exchangers (fouling is controlled and flow regime is shear controlled or turbulent to
promote heat transfer).
(2) The film coefficients are on a clean basis. Allowance for extra area is applied separately.
(3) Cooling tower water film coefficient includes thermal resistance of corrosion inhibitor film.
(4) Subcooling coefficient applies for condensate cooling in the condenser. Typically subcooling is accomplished in a sepa-
rate condensate cooler.
(5) Tubes are finned.

400-20  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

482 Surface Area Calculations

Area (A) for heat transfer is calculated from the overall heat transfer expression for
the service.

Q
A = ----------------------
U  MTD
(Eq. 400-2)
where:
A = Surface area for heat transfer, ft2
Q = Heat duty for service, Btu/hr
MTD = Mean temperature difference for service, °F
U = Overall service heat transfer coefficient, Btu/hr°Fft2

483 Tube Count and Number of Tube Passes

The number and length of the tubes is determined through trial and error. The two
are related by the necessary mechanical configuration of the exchanger to provide
the surface area (A) calculated in Section 482.

A = (#tubes) (L) () (O.D.) / 12

(Eq. 400-3)
where:
#tubes = Number of tubes per pass, dimensionless
= Mt / [(t) (Vt) (3600) (At)]
At = Cross sectional area of single tube, ft2
= () (I.D./12)2/4
L = Flow path length, ft (in heat transfer)
 = 3.142
O.D. = Tube outside diameter, inches
I.D. = Tube inside diameter, inches
Mt = Mass flow rate of tube side fluid, lb/hr
t = Fluid density of tube side fluid, lb/ft3
Vt = Velocity of tube side fluid, ft/sec
For single-phase, two-phase, and some condensing services, use the economic
sizing guidelines (Section 220) to select velocity (Vt), or a range of reasonable
velocities. Use an initial flow path length of 40 feet. This assumes a 20-foot long
U-tube exchanger with two tube passes and a full support plate at the bend tangent.

December 1989  1989 Chevron USA Inc. All rights reserved. 400-21
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

Through trial and error calculations, determine a tube count that meets the area and
velocity requirements. The flow path length may change. Consider leaving the flow
path at 40 feet, and ending up with more excess area. Be careful when specifying
exchangers with other than two tube passes. Be careful of a long flow path. The
pressure drop can be excessive.
Note that multiple tube pass exchangers have an even number of tube passes to
accommodate thermal expansion.
If tube side fluid is pure component condensing or boiling, velocity can generally be
ignored. Set tube length and calculate tube count for area. For vertical thermosi-
phons (VTSR) with tube side boiling, 8- to 12-foot tubes are typical with only one
tube pass. The actual length depends on the service as well as velocity and exit pipe
flow regime. Further definition is beyond the scope of this section.

484 Shell Diameter

For a typical U-tube exchanger with two tube passes, 0.75-inch tubes on a 1-inch
rotated square (45 degree) pitch and impingement rods, the shell diameter in inches
is given by:

Shell I.D. = 1.95 [#tubes]0.433 (for shell I.D. between 15 and 51 inches)
(Eq. 400-4)
Shell diameter should be rounded to the nearest 1/16 inch. The correlation is based
on shell side nozzle diameters between 20 to 30% of the shell inside diameter.
Within the constraints, the correlation is good to plus or minus 2%. If nozzles are
relatively smaller, the tubes may fit into a smaller shell. And, if nozzles are larger, a
larger shell may be required to accommodate all the tubes.

485 Exchanger Investment Cost

Exchanger investment cost is calculated using techniques from the Company Cost
Estimating Books. For shell and tube heat exchangers with design pressure below
600 psi for both sides, the installed cost is:

HEX = (EDMI/655) (I) (T) [ (MTL) (A) + CMP (F + m A) ]

(Eq. 400-5)
where:
HEX = Installed cost of exchanger, $
EDMI = Chevron material index, dimensionless
I = Installation factor for heat exchanger, dimensionless
T = Multiplier for geographic location adjustment, sales and other
taxes, dimensionless
MTL = Tube material adjustment, $/ft2 (at 655 EDMI)

400-22  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 400 Shell and Tube Exchanger Design and Selection

A = Area for heat transfer, ft2 (Note that installed cost is directly
proportional to area—exponent of 1.0.)
CMP = Configuration and component adjustment including component
material multipliers, dimensionless (See Cost Estimating Book)
F = Fixed cost add on which is a function of exchanger class (small or
large) and design pressure, $ (at 655 EDMI)
m = Multiplier reflecting linear cost change with area, $/ft2; the multi-
plier is a function of exchanger class—small or large—and design
pressure
For a typical exchanger configuration with all carbon steel construction and 300 psi
design pressure (both sides), the expression simplifies to:

HEX = (EDMI/655) 5.5 [13,100 + (8.8 A)]

(Eq. 400-6)
This equation assumes that the installation factor (I) is 5.5, area and tax adjustment
(T) is 1.065, material add on (MTL) is 0, and configuration adjustment (CMP) is
0.935. CMP is for 20 feet (straight length) U-tubes. F and m are for exchangers with
1000 ft2 or more.
The cost expression is different for high pressure shell and tube heat exchangers
(design pressure well above 600 psi). Cost varies with the area to the 0.64 power.
See the Cost Estimating Books for details.

December 1989  1989 Chevron USA Inc. All rights reserved. 400-23
400 Shell and Tube Exchanger Design and Selection Heat Exchanger and Cooling Tower Manual

400-24  1989 Chevron USA Inc. All rights reserved. December 1989
500 Shell and Tube Exchanger
Component Design
Considerations

Abstract
This section discusses the mechanical design of shell and tube heat exchangers and
their components. Emphasis is placed on company practices which differ from
industry standards.

Contents Page
510 Design Pressure and Temperature 500-3
511 Design Pressure
512 Design Temperature
513 Relief Valves
514 Rupture Surge Pressure
520 Bundle Design 500-7
521 Tubesheet Design
522 Tube-to-Tubesheet Connection
523 Transverse Shell Baffles
524 Longitudinal Shell Baffles
525 Impingement Devices
526 Retrofitting Bundles for Ease of Maintenance or Enhanced Performance
530 Channel and Shell Design 500-24
531 General
532 Body Flanges
533 High Pressure Closures
534 Connections

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-1
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

540 Gaskets 500-44

550 Insulation 500-48
551 Reasons for Insulating
552 Types of Insulation
553 Weatherjacketing
554 Flange Insulation

500-2 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

510 Design Pressure and Temperature

511 Design Pressure

Internal Pressure
The maximum allowable working pressure (MAWP) should normally exceed the
maximum expected operating pressure as follows:
Minimum Amount by Which MAWP
Maximum Expected Operating Pres- Exceeds the Maximum Expected
sure, psig (barg) Operating Pressure
0–170 (0–11.7) 25 psi (1.7 bar)
170–300 (11.7– 0.7) 15% of max. op. press.
300–450 (20.7–31.0) 45 psi (3.1 bar)
450–1000 (31.0–68.9) 10% of max. op. press.
1000 + (68.9 +) Not less than 8% of max. op.
press.

The maximum expected operating pressure is the maximum expected pressure

inside the heat exchanger under any operating case; startup, or shutdown condition.
• All exchangers with liquid or vapor-liquid mixtures on the low pressure side
should be designed for tube rupture safety. See Section 514 for further
discussion.
• In the case of large vertical exchangers, the nameplate design pressure is the
maximum pressure permissible at the top of the exchanger. Therefore, design
pressure must be adjusted for any difference in static head that may exist
between the part considered and top of vessel.
• Exchangers should be designed to a minimum internal pressure of 50 psig at
300°F (3.4 barg at 149°C).
• For exchangers in hydrogen service, the design pressure should meet the
criteria specified in Section 400, Prevention of High-Temperature Hydrogen
Attack, of the CES Corrosion Prevention and Metallurgy Manual.
• See Section 400, Determining Design Conditions, of the CES Pressure Vessel
Manual for a detailed discussion regarding operating pressure, design pressure,
and MAWP.
The MAWP, maximum allowable pressure (MAP) new and cold, design
temperature, and the hydrostatic test pressure of the heat exchanger should not be
limited by the mechanical design of minor components such as nozzles, nozzle
flanges, manholes, or reinforcing pads. Such items are relatively inexpensive
compared to the total equipment and should be upgraded accordingly.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-3
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

External Pressure
Exchangers in steam service or operating at less than atmospheric pressure should
be designed for an external pressure (vacuum) of 15 psi (1.0 bar). All exchangers
designed for internal pressure should be designed for at least 7.5 psi (0.52 bar)
external pressure at 450°F (232°C), excluding corrosion allowance.

512 Design Temperature

The design temperature for any part of a heat exchanger is the maximum allowable
operating temperature of the fluid inside that part (or minimum for cold service
design) plus (or minus for cold service) a temperature margin as defined below. The
design temperature controls vessel design by establishing the maximum allowable
stresses for the selected materials of construction. The following temperature
margins are recommended:
1. The hot service design temperature (−20°F [−29°C] and above) for each side of
a unit should be at least 25°F (14°C) above the maximum operating
temperature for the fluid on that side. (Note that tubes and tubesheets are
exposed both to shell side and tube side fluids.)
2. The cold service design temperature (below −20°F [−29°C]) for each side of a
unit should be at least 5°F (3°C) below the minimum expected operating
temperature of the fluid on that side.
3. For exchangers in hydrogen service, the design temperature should meet the
criteria specified in Section 400, Prevention of High-Temperature Hydrogen
Attack, of the CES Corrosion Prevention and Metallurgy Manual.
The maximum design temperature that is on the name plate of the heat exchanger
is the temperature at which the ASME Code allowable stress in tension for the
component is determined, and must be above the maximum expected operating
temperature. Normal operating temperature is only occasionally related to design
temperature. For example, tubes exposed to treated cooling tower water, irrespective
of metallurgy, plug solid if tube surface temperatures exceed ~160°F (71°C).
The intent of the exchanger design and control system is to maintain temperatures
within the functional range. However, the name plate design temperature is usually
at least the highest temperature at which the specific material maintains its
maximum allowable stress in tension. For carbon steel, the maximum design
temperature should at least be:
• 400°F (200°C) for forged material.
• 500°F (260°C) for plate and tube material.
• 650°F (340°C) for pipe material.
If external pressure controls the design, the design temperature at which the
allowable compressive stress is maximum could be as low as 300°F (150°C),
provided this temperature exceeds the maximum operating temperature by a margin
as noted above.

500-4 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Minimum Design Metal Temperature (MDMT) (a.k.a.: minimum pressurizing

temperature or MPT), an important design parameter, is discussed in detail in
Section 400, ASME Code, Section VIII, Division 1, Design Temperatures, of the
CES Pressure Vessel Manual; MDMT is also discussed in detail in Sections 343
through 346 in the CES Corrosion Prevention and Metallurgy Manual. In brief, the
reason for establishing a minimum pressurizing temperature is to avoid a brittle
fracture. Ordinary carbon steels, for example, become brittle at low temperatures.
The ductile-to-brittle transition temperature may range from well above ambient to
well below ambient depending on grade of steel used, its thickness, and Charpy
impact testing requirements specified, if any. The aim is to choose a material which
will not suffer brittle fracture under the design operating conditions of the
exchanger. This includes hydrotest, which should be done at a temperature at least
30°F (17°C) above the minimum pressurizing temperature, but not greater than
120°F (49°C), as recommended by ASME Section VIII, Div. I, Section UG-99, to
minimize the risk of brittle fracture.
Generally, one should consider the lowest one-day mean temperature when setting
the MDMT for a design. However, a heavy wall or alloy vessel may have an
MDMT that is considerably higher than the lowest one-day mean temperature, if it
makes sense from an economic and operational point of view.
For any alloy (non-carbon steel) vessel, always consult a materials engineer in
setting the MDMT. Alloy materials can embrittle in some services, and the MDMT
set by the fabricator may not be adequate for the long term.
Low alloy steels (1Cr, 1.25Cr, 2.25Cr, and 5Cr) are subject to “temper
embrittlement” if they are exposed to temperatures above about (conservatively)
650°F (343°C). The net effect is that the “ductile-to-brittle transition temperature”
increases after being exposed to such temperatures. That is, the equipment will be
brittle at higher temperatures than when it was new; the MDMT should include an
appropriate margin above impact test temperature to account for “temper
embrittlement.” For more information, see Section 345 of the CES Corrosion
Prevention and Metallurgy Manual.
One additional note on MDMT—prior to 1987, the ASME Code default MDMT
was −20°F (−29°C), unless a lower temperature was specified. For equipment built
to a pre-1987 Code, an MDMT (MPT) should be established to prevent brittle
fracture based on Impact Testing Exemption Curves in the current ASME
Section VIII-I, UCS-66, using the methodology described in Sections 343 to 346 of
the CES Corrosion Prevention and Metallurgy Manual.

513 Relief Valves

Pressure Relief
The ASME Boiler and Pressure Vessel Code requires that all pressure vessels be
provided with protection against overpressure by use of pressure relief devices. The
protective devices need not be directly on the pressure vessel when the source of
pressure is external to the pressure vessel and the piping does not include any valves
between the relief device and the vessel, unless such valves are locked open if
allowed by Code.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-5
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Consequently, many heat exchangers do not have pressure relief valves directly on
the vessel, but are rather part of an overall hydraulic system which does have
protection from overpressure. In many cases, the source of pressure is a pump or
compressor external to the exchanger. The source of overpressure can also be
internal to the heat exchanger: the high pressure side (tube side or shell side) of a
heat exchanger can over pressure the low pressure side in the event of a tube failure,
or either side of a heat exchanger can be overpressured by heat input from the other
side (e.g.: If cooling water is blocked in at a heat exchanger, the cooling water side
can be overpressured).
The CES specification FRS-DU-5056 and FRS-DU-5057 discuss relief valve sizing
and overpressure protection in more detail.

Thermal Pressure Relief

Thermal pressure relief valves (TRVs) are required by the ASME Code by the
following conditions:
1. Either shell or tube side component can be overpressured by heat input from the
other side, and
2. The low pressure side (shell side or tube side) can be overpressured by the high
pressure side in the event of a tube failure, or
3. The component can be isolated from the main pressure safety valve (PSV) by
valves other than PSV maintenance valves, or
4. The component is not protected by a PSV.
To minimize the possibility of a TRV releasing during a PSV relief, the TRV can be
set at 110 percent of MAWP as allowed by ASME Code. One thermal relief valve
can serve as the protective device for multiple exchangers in series if there are no
block valves between them.

514 Rupture Surge Pressure

All exchangers with liquid or vapor-liquid mixtures on the low pressure side should
be designed for tube rupture safety. This is accomplished by setting the design
pressure on the low pressure side equal to the maximum normal operating pressure
plus the initial surge pressure due to the complete break of one tube. Long term
(e.g., 2 + seconds) pressure transients should be prevented with relief devices in the
piping. Tubesheets, shells, shell covers, body flanges, and channels should be
designed to this surge pressure. Tube rupture is defined as the sudden sharp break of
one heat exchanger tube, such as might be caused by a guillotine cut, that allows
high pressure fluid to flow through an opening equal to twice the inside cross
sectional area of one tube.
API Standard 521, 5th Edition, Section 5.19, “Heat-transfer equipment failure,”
provides requirements and general methodology for performing a tube rupture
analysis and mitigation. The “Exchanger Tube Rupture” section in CES
specification FRS-DU-5057 expands on API’s requirements and methodology and
states the criteria for when a steady state tube rupture analysis or a transient tube
rupture analysis is required.

500-6 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

An analysis of steady state tube rupture conditions shall be made for a heat
exchanger if EITHER of the following conditions exists:
a. MAWP of low pressure side is less than 0.77 of the MAWP of high
pressure side. For heat exchangers designed to the 1998 and earlier ASME
Code, the value is 2/3 the MAWP.
b. High pressure side operates at 1000 psig (68.9 barg) or more.
A computer analysis of transient tube rupture conditions shall be made for a heat
exchanger if ALL of the following conditions exist:
a. Low pressure side is liquid full; i.e., no vapor is present at either inlet or
outlet.
b. High pressure stream is gas, vapor, two phase, or liquid that will partially
vaporize if depressured to 130 percent MAWP of low pressure side.
c. MAWP of low pressure side and/or short term pressure rating of piping
connected to low pressure side is less than 0.77 MAWP of high pressure
side. For exchangers designed to the 1998 and earlier ASME Code, a value
of 2/3 of the MAWP of the high pressure side shall be used.
d. High pressure side operates at 500 psig (34.5 barg) or more.
If an analysis of transient tube rupture conditions is made, it shall be necessary to
also make an analysis of steady state tube rupture conditions.
Appendix F of the CES Heat Exchanger and Cooling Tower Manual details a
simpler methodology to calculate tube rupture flow rates and shock pressures for
shell and tube exchangers containing gases, two-phase mixtures, and liquids.
Example calculations are also provided in Appendix F. ETC is currently considering
under what conditions Appendix F can be used in lieu of the methods outlined in
FRS-DU-5057.
Note that in addition to the heat exchanger, the effects of tube rupture will impact
piping and equipment attached to the low-pressure side of the heat exchanger.
Steady-state analysis results are mitigated with appropriately placed relief devices.
If a tube rupture analysis is required by the criteria stated above, consult with ETC’s
Heat Exchanger Specialists and Relief Systems and Flares Specialists.

520 Bundle Design

521 Tubesheet Design

This section covers the applicable codes and industry practices for establishing
tubesheet design and tubesheet thicknesses. Tubesheets separate the shell side and
tube side fluids and provide the anchor point for tube ends. Historically, Tubular
Exchanger Manufacturers Association (TEMA) Standards for calculating tubesheet
thickness were used in the industry extensively. Beginning with the 2003 Addenda
to ASME Section VIII, Div. 1, Part UHX contains the requirements for tubesheet

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-7
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

design. ASME Boiler and Pressure Vessel Case 2429 allowed the use of TEMA
design rules for tubesheets until January 1, 2005, at which time Part UHX became
mandatory.
Repairs to tubesheets fabricated prior to the issuance of Part UHX may be made
such that the tubesheet is restored to meet TEMA design rules. Local jurisdictions
may require that replacement bundles for any heat exchanger, regardless of Code
edition of construction, be designed to Part UHX. The State of California requires
all replacement bundles to be designed to Part UHX.

TEMA Standards
Beginning with the 9th edition, the TEMA Standards include procedures to
establish tubesheet thickness for U-tube bundles, floating head bundles, and fixed
tubesheet construction as part of non-mandatory Appendix A of TEMA Standards.
This appendix may be consulted for repairs on existing tubesheets designed prior to
2005. The design and construction of new tubesheets shall comply with ASME
Code.

ASME Code, Section VIII, Division 1

ASME Code, Section VIII, Division 1, Part UHX, covers the design, fabrication and
inspection of shell-and-tube heat exchangers. It includes rules for:
• Determination of flange extension minimum thickness.
• Determination of tubesheet characteristics including: ligament efficiencies,
effective depth of pass partition grooves, and effective elastic constants to be
used in design of U-tube, fixed, and floating tubesheets.
• Design of U-tube, fixed, and floating tubesheets.
Tubesheet thickness may be determined by taking the largest thickness calculated
using the TEMA Non-Mandatory Appendix A method or the ASME Section VIII,
Division 1, Section UHX method. However, design calculations per Section UHX
will be required as well. A comparison of the TEMA method and the UHX method
indicates that for:
• Low pressure, large diameter shells, the TEMA method gives a thicker
tubesheet.
• Low pressure or high pressure small diameter shells, the UHX method gives a
thicker tubesheet.
• High pressure, large or small diameter shells, the UHX method gives a thicker
tubesheet.
If a new bundle is being designed for an existing heat exchanger, the stationary
tubesheet flange extension thickness (grip thickness at gasket surfaces) can usually
remain unchanged and any addition required tubesheet thickness can be added to the
shell-side of the tubesheet without impacting the distance between channel and shell
nozzles. For TEMA type S floating head bundles, this may cause a change in
thickness of the floating head and the split ring could require modification if the
tubesheet thickness changes.

500-8 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

In the case of fixed tubesheet heat exchangers such as surface condensers,

ASME Code, Section VIII, Division 1, Section UW-20 and non-mandatory
Appendix A, describe how to calculate allowable loads for tube-to-tubesheet joints.
This calculation may have an effect on tubesheet thickness, the method of joining
tube to tubesheet, or the need to provide an expansion joint in the shell of an
exchanger. Appendix A of Section VIII is non-mandatory and therefore does not
have to be followed by a vendor unless required by the Company.
The maximum temperature difference between the tube-side inlet and outlet
nozzles, or shell-side nozzles in the case of a TEMA F shell configuration, should
be less than 200°F (111°C) to prevent distortion of the tubesheet which can result in
loss of seal between the pass partition and tubesheet; in the case of a TEMA F shell,
a loss of seal between the tubesheet and longitudinal baffle resulting in flow
bypassing.

Unfired Waste Heat Boiler–Fixed Tubesheet Exchanger Type

Waste heat boiler tubesheets are designed in accordance with the ASME Boiler and
Pressure Vessel Code, Section I, Paragraphs PG-46, PG-49.1, PW-19.1, PFT-26 and
PFT-27, which account for staying capacity of the tubes. Tubesheet thickness is
governed by the largest unstayed area, which is usually the annular space between
the bundle and the shell.
Unstayed fixed tubesheets designed by ASME Code Section VIII, Division 1, Part
UHX rules would be much thicker and are unacceptable for high temperature steam
generators because of high thermal stresses. With cold liquid on one side and hot
fluid on the other side, the resulting thermal stresses and stresses from differential
expansion could be high enough to cause damage to the tubesheet, if designed to
UHX rules.

Tubesheet Thicknesses and Tolerances: Chevron Practice

• It is Company practice to use TEMA, Paragraph F-3, tolerances for thickness
and API 660, Table 3 for flatness tolerance on new tubesheets, although this is
generally not a problem. Though covered in CES specifications, this
requirement should be highlighted to manufacturers that are not familiar with
Chevron’s practice.
• Once a tubesheet thickness is determined, it is Chevron’s practice to add
1/8 inch (3.2 mm) of thickness to each gasket surface as a machining allowance
for future maintenance activities. This additional thickness is not to be used in
the calculation of tubesheet thickness.

Clad Tubesheets
For clad tubesheets with rolled tube-to-tubesheet joints, the nominal cladding
thickness should be 3/8 inch (9.5 mm) minimum tube-side and/or shell-side. For
shell-side clad tubesheets, the tube should be expanded to seal against the cladding
material for a distance of 1/4 inch (6.4 mm) minimum. In no case should the
expansion encroach within 1/8 inch (3.2 mm) of the shell-side face of the tubesheet.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-9
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

For welded tube-to-tubesheet joints the tubesheet cladding thickness should be

3/16 inch (4.8 mm) minimum on the tube side and 3/8 inch (9.5 mm) minimum on
the shell side.
The maximum nominal cladding thickness should not exceed 1/2 inch (13 mm) for
tubesheets. To prevent crevice corrosion between the tube-side cladding and the
base material, as a minimum, tubes should be strength welded into the tubesheet.
Roll-bond clad can be used for cladding or overlay up to base metal thickness of
4 inches (100 mm). Weld overlay cladding shall be used for base metal thickness
over 4 inches (100 mm). Explosion cladding can be used to clad tubesheets thicker
than 4 inches (100 mm) with Company approval. Clad and weld overlay
construction is discussed in detail in PVM-SU-4749 and PVM-SU-1322, which can
be found on the CES Pressure Vessels webpage.

Bundle Pull Hole Design

Removable bundles which are 20 inches (510 mm) or more in diameter should have
four tapped holes in the channel side of the stationary tubesheet for pulling bundle
out of shell. Pull hole size and thread engagement should take into account
tubesheet material and be designed for a maximum pulling load equal to 1.5 times
the bundle weight.
Bundles smaller than 20 inches (510 mm) in diameter may have two tapped holes
for pulling eyes. Small pre-engineered exchangers are usually supplied without
pulling holes.
These guidelines may have to be modified or waived for special construction, such
as for thin clad tubesheets. Pull holes should be protected in service by threaded
plugs.

522 Tube-to-Tubesheet Connection

The main function of tube-to-tubesheet joint is to seal the tubes tightly to the
tubesheet, and for some exchangers, an additional function is to support the
tubesheet against pressure induced load. Tubes are sealed inside the tubesheet by the
following methods.
• Expanding tubes inside tubeholes
• Seal welding and expanding tubes inside tube hole (not recommended)
• Strength welding and contact expanding tubes inside tubehole

Expanding Tubes Inside Tubeholes

Expanded tube-to-tubesheet joints are industry standard. In this case tubes are
expanded inside tubeholes by such methods as rolling or applying hydraulic
pressure directly to the tube end. Properly rolled joints have uniform tightness to
minimize tube fractures, stress corrosion, tubesheet ligament enlargement, and
dishing of the tubesheet. Rolling to at least 95 percent of tubesheet thickness is
recommended; tubes shall not be rolled to within 1/8 inch (3.2 mm) of the shell-side
face of the tubesheet. Rolling at or beyond the tubesheet thickness is not
recommended—for it may damage the tubes.

500-10 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

For all services with expanded tubes at least two grooves are machined (1/8 inch
wide by 1/64 inch deep [3.2 mm wide by 0.40 mm deep]) in each tube hole. See
Figure 500-1.

Fig. 500-1 Rolled Tube—Tubesheet Connection

Expanding the tubes into the grooved tube holes provides greater axial strength but
results in greater difficulties during tube removal. The tube hole grooves do not
improve the sealing ability of the joint; the seal between the tubesheet and tube is
achieved in the length of the tube expansion between the two smooth surfaces.
The following steps must be taken when tubes are rolled inside tube holes:
1. Tubes are positioned in the tubesheet.
2. Tubes are mechanically rolled into the tubesheet. Maximum tube wall thickness
reduction of 5 to 8 percent is used as an indicator of adequate rolling; however,
this value varies by tube material. See API 660, Section 9.10, Table 4.
3. Hydrotest the shell side after rolling.

Welding the Tubes to Tubesheet

Additional leak tightness and/or axial strength beyond that of the tube rolling is
sometimes required in the following areas:
• Exchangers where the design differential pressure across the tubesheet is
750 psi (51.7 bar) or higher
• Waste Heat Boilers where the tubesheets are thin and stayed by the tubes per
ASME Section I
• Boiler feedwater heaters

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-11
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

• Feed/effluent heat exchangers in hydroprocessing plants when any intermixing

of shell and tube side fluids results in off-test product
• Any service where cross-contamination of streams must be scrupulously
avoided
• Exchangers in hydrogen service
• Lethal service
In these cases, tubes are to be strength welded (not seal welded) to the tubesheet
and then expanded (rolled) into the tubesheet. Weld joint design and weld sizing
information is given in ASME Code Section VIII, Div. 1, Section UW-20.
Figure UW-20.1 shows acceptable types of tube-to-tubesheet strength weld details.
Details (c) and (d) are preferred and the minimum allowable tube projection is 
1/16 inch (1.6 mm) before welding. Tube ends are contact expanded after welding
or post weld heat treatment (PWHT) if required.
In the case of High Temperature Waste Heat Steam Generators with fixed
tubesheets, the tubes are strength welded to the tubesheets. The tubesheets are
designed thin to reduce the effects of thermal stresses and are stayed by the tubes,
meaning that the tubes act to resist hydrostatic end forces acting on the tubesheets.
Design considerations for stayed tubesheets and High Temperature Waste Heat
Steam Generators are discussed in Sections 350 and 450 of the CES Heat
Exchanger and Cooling Tower Manual.
When welded tube joints are specified, prior to inserting tubes, the tube holes are
prepared for welding; the weld preparation should contain sufficient depth or
contour such that the deposited metal meets the required dimension of the joint
design. The tubes are inserted into the tube holes such that a minimum projection of
1/16 inch (1.6 mm) is achieved. The tubes are either tack welded into position or
given a light contact expansion into the tube holes. The expansion must be light
enough such that gases generated during welding may escape out the shell-side of
the tube hole. If gases cannot escape, a pinhole can be created as the weld freezes at
the completion of the weld. The tube-to-tubesheet joint is welded per the design
detail. Following welding, the tubes are expanded, starting 1/4 inch (6.4 mm)
behind the weld. The welds are dye penetrant tested upon completion and the
bundle is hydrotested.
Seal welding is defined as a very small bead of weld around the tubes where no
credit can be taken for strength of that weld for calculation of tube-to-tubesheet
joint load (CES specification EXH-SC-2583, Shell and Tube Heat Exchangers
for General Refinery Service, does not allow seal welded and expanded
tube-to-tubesheet joints unless requested by the Purchaser). A strength weld is
designed to provide all the necessary axial strength and sealing ability of the
tube-to-tubesheet joint. Following welding, the tubes are expanded into the
tubesheet.
Following welding and prior to tube expansion, it is good practice to require the
fabricator to perform a “soap test.” A “soap test” is performed by installing the
bundle into the shell and applying a nominal pressure to the shell and while
applying a soap and water solution to the face of the tubesheet—soap bubbles are
generated if there are leaks in the welds. This test can identify pin-hole leaks not
found by dye penetrant testing.

500-12 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

523 Transverse Shell Baffles

Most heat exchanger bundles have transverse baffles that serve two general
purposes: transverse baffles establish the shell side flow path to enhance heat
transfer in the bundle and support the tubes. Baffles typically consist of disks with
windows cut from them—the size of the window cut from the baffle is a function of
the shell internal diameter (ID) and is expressed as a percent of the shell ID. A
20 percent cut baffle is a baffle with a vertical or horizontal cut on one side—the
height of the circular segment cut from the baffle is 20 percent of the shell ID,
measured diametrically.
The most commonly used baffle types are:
• Single Segmental
• Double Segmental
• Segmental No-Tubes-in-Window (NTIW)
The above common types of baffles are discussed in detail below. Other baffle types
include Rod-Type, Helical, and Disk and Donut. See Figure 500-2.
The required thickness of Transverse Baffles is that given in TEMA Table R-4.41
plus shell-side corrosion allowance. The thickness given in TEMA is a function of
nominal shell diameter and unsupported tube length between central baffles.
The distance between baffles determines the unsupported length of the tubes. Baffle
spacing is typically 20 to 50 percent of the shell ID. The maximum allowable
unsupported tube length (baffle spacing) is determined by flow-induced tube
vibration analysis as described on standard drawing GC-E1048, which is available
on the Chevron Engineering Standards website under Heat Exchangers.

Single Segmental
Single segmental baffles force the shell-side fluid to flow across the entire tube
field. This type of baffle arrangement is the simplest, least expensive arrangement
and gives the highest heat transfer rate (under normal conditions). It also results in
the highest shell-side pressure drop. Single segmental baffles are used when
sufficient pressure drop is available to permit the required maximum baffle spacing
as determined through vibration analysis using Standard Drawing GC-E1048. If
there is insufficient pressure drop available, double segmental baffles may be used.
In the case of very viscous fluids, high viscosity produces poor flow distribution;
bypass and leakage streams can be so large that the resulting heat transfer efficiency
is poor. Double segmental baffles can give better results because of a more uniform
flow distribution with less bypassing and leakage.

Double Segmental
Double segmental baffles split the shell-side flow into two parallel streams; each
stream is forced to cross only half the tube field, resulting in a lower pressure drop
when compared to single segmental baffles. Double segmental baffle arrangement is
used when the pressure drop available, based on system hydraulics, is insufficient
for a single segmental arrangement.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-13
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Fig. 500-2 Transverse Shell Baffle Types (Images reproduced with permission from Heat
Transfer Research, Inc.)
Shell (Typ.)

Single-Segment al Double-Segment al

Out erm ost Tubes

(Typ.)

Segment al
No Tubes in Window (NTIW)

Rod Suppor t Onl y, TEMA X Shell

Doughnut

Disk

Heli cal Disk & Doughnut

With this arrangement, the baffle spacing may need to be reduced, compared to
single segmental baffles, to provide the needed tube support.
Flow distribution is also improved when compared to single segmental baffles due
to less driving force (differential pressure) for leakage and bypass flows. However,
turbulent flow heat transfer can be significantly less than for single segmental
baffles at the same flowrate.

Segmental No-Tubes-in-Window (NTIW)

When the pressure drop resulting from a double segmental arrangement exceeds the
available pressure drop based on system hydraulics and tubes are inadequately
supported, a NTIW baffle arrangement can be used to further reduce the shell-side
pressure drop. In a NTIW arrangement, each baffle is cut on one side like single
segmental arrangements, however no tubes are present in the area of the baffle cuts
(windows). With this arrangement, tube support requirements are decoupled from

500-14 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

pressure drop considerations—with a NTIW arrangement, baffle spacing can be set

to achieve desired pressure gradient and intermediate tube supports can be added
without impact on the shell side flow path or pressure gradient.
Because the baffle spacing is maximized, the velocity is reduced and pressure drop
is reduced when compared to single and double segmental designs. For optimal flow
distribution, the cross-flow pressure drop (across tube bundle) should be greater
than the momentum pressure drop in the windows. The velocity through the window
is directly related to baffle cut. Window velocity should be set such that the velocity
head is less than, or equal to, 10 percent of the pressure drop across the bundle.
Velocity head is defined as:

Velocity Head = V2/(2 gc)

(Eq. 500-1)
where:
 = fluid density, lbs/ft3 (kg/m3)
V = window velocity, ft/s (m/s)
gc = gravitational constant 32.2 ft/s2 (set term equal to 1 for SI units)
Two disadvantages of using a NTIW design are that the heat transfer will be less
than that with single segmental baffles and the shell diameter will be larger.
A variation of the NTIW design is used in cross-flow TEMA Style X shells. The
windows are on top and bottom of the baffle plates and the baffles act as full support
plates, supporting every tube.

Rod-Type Baffles
Rod-type baffles are not really baffles at all—they do little to direct the flow on the
shell-side of a heat exchanger. They are simply arrays of rods or lattice work
positioned across the tube bundle. They only provide support for the tubes and
allow axial flow through the bundle with good flow distribution and low pressure
drop. The heat transfer with rod-type baffles is low compared to cross-flow
configurations discussed above. The most common version is patented by Phillips
Petroleum under the trade name RODbaffle®.

Helical Baffles
Helical baffles are quadrant shaped segment baffle plates that are positioned at an
angle to the longitudinal axis in a sequential arrangement to guide the shell-side
flow into a helical pattern. This is a licensed technology provided by Lummus Heat
Transfer (Helixchanger®) and Koch Heat Transfer Group (Heli-Baffle®).
Chevron does not have much experience with this type of baffle arrangement. One
particular application, a Crude Unit preheat train heat exchanger, resulted in a
dramatically higher than expected shell-side pressure drop.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-15
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Disk-and-Doughnut Baffles
As the name implies, Disk-and-Doughnut baffle arrangements consist of an
alternating pattern of disks and doughnuts; the outer diameter of the disks is much
less than the ID of the shell; the outer diameter (OD) of the doughnut nearly
matches the shell ID and the ID of the doughnut is somewhat smaller than the OD
of the disk. This arrangement was originally introduced to permit lower pressure
drops and better bundle penetration than that of single segmental baffles. The
double segmental baffles serve the same function.

524 Longitudinal Shell Baffles

In the design of heat exchangers, it is sometimes advantageous to use a TEMA
Type “F” two-pass shells. F-shells require a longitudinal baffle to control the shell
side flow. To prevent bypassing on removable bundles, the seal between the
longitudinal baffle and the shell is most commonly a “Lamiflex” type.

Longitudinal Baffle Thickness

Longitudinal baffles are used to direct shell-side flow in TEMA F, G, and H type
shell. Specification EXH-SU-2583 states that the longitudinal baffle thickness be
the largest of these three: (1) 1/4 inch (6.4 mm), or (2) the thickness required by
TEMA for transverse baffles, or (3) the thickness for a design differential pressure
loading of 20 psi (1.4 bar).

Attachment to Tubesheet
In the case of TEMA F shells, the longitudinal baffle should be welded to the
tubesheet with a fillet weld. The longitudinal baffle may be bolted to the tubesheet
using clips or some other means in cases where the baffle and tubesheet materials
are not readily weldable, (e.g.: require post weld heat treatment [PWHT]).
In the case of TEMA G and H shells, the longitudinal baffle spans from the
tubesheet toward the tube u-bends of the bundle and has windows cut in it to direct
flow. It too should be welded or attached to the tubesheet. Alternatively, it can be
welded to one transverse baffle using a continuous fillet weld. In no case should the
longitudinal baffle be welded to more than one transverse baffle due to differential
thermal expansion issues.

Lamiflex Baffle Seal

A stack of flexible strips is most commonly attached to the edge of the longitudinal
baffle by sandwiching them between the longitudinal baffle and a bolting strip as
shown in Figure 500-3.
A typical seal consists of a long stack of eight strips, each 0.004 inch (0.10 mm)
thick. This thickness represents a compromise: A thin strip is fragile and vulnerable
to mechanical damage when the tube bundle is handled and inserted into the shell,
and is also susceptible to damage by corrosion. However, only thin strips can flex
adequately to seal.

500-16 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Fig. 500-3 Lamiflex Baffle

Material used for the strips is typically soft or annealed Type 304 stainless steel, or
Type 302 stainless steel 1/2 or 3/4 hard. In exchangers with shell-side temperatures
of 650°F (343°C) or higher such as hydroprocessing unit feed-effluent exchangers,
Alloy 625 or Alloy C-276 strip is used.
The angle of contact between shell and flexible strips should be small so that
friction during installation is minimized and the differential pressure has the greatest
effect in causing the strips to seal. To this end, it is recommended that dimensions
“A” and “B” in Figure 500-4 should be about equal, with both in the range of 1/2 to
3/4 inch (13 to 19 mm).
Protection during installation. The lamiflex baffle must be protected with cribbing
to avoid damage during rigging operations. The lamiflex baffle strips should be
replaced every time the bundle is removed and reinstalled. The strips are typically
damaged beyond serviceability after removing and reinstalling the bundle.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-17
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Fig. 500-4 Lamiflex Baffle Dimensions

525 Impingement Devices

This section compares Chevron criteria against TEMA Recommended Guidelines
for impingement devices. It also discusses the types of impingement devices used in
the Company.
The tubes directly underneath the shell inlet nozzle may need to be protected against
impinging fluid. Lack of proper impingement devices can cause tube failure by
corrosion, erosion, or vibration. However, use of an impingement device when it is
not needed increases exchanger diameter and cost.

Chevron Practices
TEMA recommends impingement plates for most services. Impingement plates
have been a chronic cause of both erosion and vibration problems. Removing
impingement plates has been a common solution.
Chevron’s normal practice is to put an array of two staggered rows of rods in the
projection of the inlet nozzle to serve as an impingement device and also to
distribute flow in the bundle. The impingement rod arrays are recommended for all
exchangers (regardless of service) where shell diameter is 20 inches (510 mm) or
larger. Impingement rods are not practical in small exchangers (shell diameter less
than 20 inches [510 mm]) and impingement plates are used when required by
TEMA.

Impingement Rods
Impingement rods are preferred to an impingement plate for several reasons. First,
the plate creates a dead space directly beneath it, lowering the heat transfer in those
tubes. Also, if the plate blocks too much of the inlet area, then the fluid may
accelerate into the remaining gap causing serious erosion of the tubes in that area.
The design the impingement rod array is detailed as follows:
• The rods should be the same diameter as the active tubes.
• The two rows of rods replace the first two tube rows which extend past the
nozzle projection.

500-18 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

• The distance between the center-lines of the outermost rods in the first row is at
least equal to the 120 percent of the inside diameter of the shell inlet nozzle.
• The effective length of the rods is at least equal to 120 percent of the diameter
of the shell inlet nozzle. The actual length of the rods may extend beyond the
effective length as required for construction.
• For staggered tube layouts (30 and 45 degrees), the impingement rods should
be of the same layout as the active tubes. For inline tube layout (90 degrees),
the impingement rods should have a 45 degrees staggered layout.
• When replacing an impingement plate with impingement rods on an existing
bundle (or when replacing an existing bundle with a new bundle designed with
impingement rods), the thermal performance of the heat exchanger should be
evaluated using HTRI design software if rows of tubes were removed to make
room for the impingement rod assembly.

TEMA Guideline
The TEMA Standards provide a minimum guideline for determining when an
impingement device should be used. This guideline is appropriate for Company use
also. Impingement protection underneath the shell inlet nozzle is recommended for
the following:
• All noncorrosive, nonabrasive, single phase fluids with V2 > 1500 (2230) in
the shell nozzle.
• All other liquids, including liquids at their boiling point V2 > 500 (740) in the
shell nozzle.
• All gases and vapors, including all nominally saturated vapors, and for
liquid/vapor mixtures regardless of the magnitude of V2.
The TEMA Standard also recommends that in no case should the shell or bundle
entrance or exit area produce a value of V2 in excess of 4000 (6000). Refer to
TEMA Section RCB-4.62 for the definition of shell and bundle entrance and exit
areas. The shell entrance area and the bundle entrance area should not be less than
the inlet nozzle flow area.
In the expression V2, “V” is the linear velocity of the fluid in feet per second
(meters per second) and “” is its density in pounds per cubic foot (kilograms per
cubic meter).

Other Types of Impingement Devices

Impingement Plate. A circular or rectangular plate is placed directly underneath
the inlet nozzle perpendicular to nozzle flow. This plate could be welded to the
shell, bolted to clips which are welded to the shell, or bolted to baffles on either side
of shell inlet nozzle. The preferred construction is to attach the plate to the bundle.
If an impingement plate is used, it must be at least 1/4 inch (6.4 mm) thick and
extend a minimum of 1 inch (25 mm) (or 10 percent nozzle diameter, whichever is
greater) on each side of the projected nozzle bore. Also the flow area off the

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-19
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

impingement plate should be more than the inlet nozzle flow area. Impingement
plates may be used on small diameter shells (less than 20 inches [510 mm]), and in
special circumstances such as cross flow TEMA type K and X shells. Impingement
plates, however, are not recommended for larger diameter shells because of the
problems stated above.
Distribution Belt. A distribution belt consists of a collar that fits around the shell at
the inlet and/or the outlet. The shell nozzle attaches to this collar. The fluid enters
through the nozzle and flows through the annulus between the belt and the shell.
The fluid enters the tube bundle through windows cut in the shell, with a reduced
velocity.
Distribution belts are typically poorly designed by contractors and fabricators; belts
are poorly designed mechanically and can result in poor distribution of the shell-
side fluid into the bundle. Distribution belts are not widely used in Chevron; if used,
they should be reviewed by a heat exchanger specialist. They are expensive and
have maintenance problems.

526 Retrofitting Bundles for Ease of Maintenance or Enhanced Performance

Bundles can be retrofitted for ease of maintenance and to enhance performance and
reliability. Some examples of such retrofits are as follows:
• Converting floating head bundles to U-tube bundles
• Transverse baffle modifications for shell-side differential pressure reduction
and/or tube vibration mitigation
• Helical baffles
• Twisted Tube® bundles
• Retrofitting bundles for vibration control
Each of these topics will be discussed below.

Floating Head to U-Tube Conversion

It is often advantageous and cost effective to change a floating head bundle to a
U-tube bundle. This change may be warranted because of concern about the
possibility of leaks between floating head flange and the tubesheet, however, when a
graphite faced gasket is used in this joint, leaks are rare. U-tube bundles can be
more economical to fabricate than a floating head bundle. One disadvantage is that a
U-tube bundle is more difficult to clean if the tubes are heavily fouled—the U-bends
themselves can be difficult hydroblast.
Retrofitting a floating head to a U-tube bundle requires thermal, hydraulic, and
vibration redesign. Once the need for retrofitting has been established, the following
steps can be taken.
1. Obtain existing exchanger data sheet and fabrication drawings.
2. Complete a new datasheet for the heat exchanger which includes the following
information:
– Current performance requirement of the exchanger. In case of split ring
type floating head where possibly less heat transfer area will be available,
the re-evaluation of performance data may be required.
500-20 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

– Maximum allowable length of bundle. Allow minimum of 2 inches

(51 mm) clearance between end of U-bends and inside of rear shell cover.
– Location of shell nozzles relative to the face of the shell body flange.
– Location and thickness of existing channel pass partition plates.
– Material of construction for the bundle. Review inspection history for
corrosion issues and consult with a materials engineer regarding the suit-
ability of existing materials.
– Tubesheet thicknesses.
– Tube sizes, pitch, and layout preference.
– Baffles type, cut, and spacing preference.
– Impingement device requirement.
3. Note that all the above data are subject to re-evaluation for the new bundle. The
only criterion is that the new bundle must fit in the existing shell, rear shell, and
channel.
4. Design a U-tube bundle based on the new data sheet. This should be done by
using the HTRI design software.
– Compare cost of retrofit to cost of new exchanger: extensive modifications
to channel or shell may justify purchase of a complete new exchanger.
– Consider the possibility that it may be necessary to remove the channel
pass partition plates on the existing unit and install new ones. This is not
considered extensive modification.
– Consider the effects of excessive vibration and its prevention (see standard
drawing GC-E1048).
– Consider the location of the U-bends relative to the location of the rear
shell nozzle. Placing the U-bends in the shell entrance or exit region should
be avoided—the shell nozzle may need to be moved. U-bends represent
long unsupported spans that are susceptible to vibration and vibration
failure. A full support plate should be provided near the U-bends tangent.
When converting from a pull-through floating head to a U-tube bundle, the number
a tubes can usually be increased, providing additional surface area. In the case of
split ring floating head designs, the number of tubes may need to be reduced when
converting to a U-tube design.

Transverse Baffle Design Modifications

Modification to the transverse baffle design and layout can reduce the shell-side
pressure drop and mitigate tube flow induced vibration which can lead to the
severing of tubes. The reduction of pressure drop can be accomplished by changing
from single segmental baffles to double segmental, or no tube in window (NTIW)
baffles. Moving to a NTIW design can also reduce pressure drop, but the number of
tubes must be reduced in order to fit the new bundle into the existing shell.
Changes to the baffle design and layout must be weighed against the allowable
unsupported span of the tubes—see standard drawing GC-E1048 for the
determination of allowable unsupported tube spans. Changing the baffle design and
layout could result in an increase in the unsupported tube spans. Additional partial
support baffles may be required in the entrance and exit regions.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-21
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Changes to the baffle design and layout will impact the thermal performance of the
heat exchanger and should be evaluated using HTRI design software.

Helical Baffles
Helical baffles are quadrant shaped segment baffle plates that are positioned at an
angle to the longitudinal axis in a sequential arrangement to guide the shell-side
flow into a helical pattern. This is a licensed technology provided by Lummus Heat
Transfer (Helixchanger®) and Koch Heat Transfer Group (Heli-Baffle®).
Retrofitting an existing bundle with a helical baffle arrangement can result in
improved thermal effectiveness, enhanced heat transfer, reduced pressure drop shell-
side, lower fouling tendency shell-side due to more uniform velocity profile, and a
lower tube vibration risk.
Chevron has limited experience with this technology. One particular application, a
Crude Unit preheat train heat exchanger, resulted in a dramatically higher than
expected shell-side pressure drop.

Twisted Tube® Bundles

Twisted Tube® is a proprietary tube design provided by Koch Heat Transfer. A
sample of a Twisted Tube is pictured below in Figure 500-5. At the terminal ends of
the Twisted Tube are lengths of tube with a round cross-section such that the tube
can be rolled into a tubesheet.

Fig. 500-5 Twisted Tube®

Twisted Tubes are assembled into a bundle on a tight triangular pitch such that there
is a point of contact between adjacent tubes at approximately 1 inch (25 mm)
intervals. The arrangement of these points of contact forms the shell-side pass lanes
through the bundle, eliminating the need for baffles. Since the tubes contact each
other at 1 inch (25 mm) intervals, flow induced tube vibration is not an issue.
However, a major limitation is that the tube- and shell-side flow areas are dependent
on each other. The shell-side flow path cannot be adjusted without changing the
tube-side flow path.
Even with the tighter triangular tube pitch, shell-side cleaning can accomplished
with high pressure water spray through the pass-lanes that are formed by the tube
arrangement in the bundle. Tube-side cleaning is possible using conventional
methods.

500-22 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Prior to retrofitting an existing heat exchanger bundle with Twisted Tube bundle,
consult with a heat exchanger specialist. Twisted Tube technology should not be
used in some fouling services due to plugging concerns shell-side.

Retrofitting Bundles for Vibration Control

When plants are modified for increased throughput or change in service, existing
exchanger bundles can become susceptible to flow-induced vibration problems and
tube failures, tube spans should be checked for vibration using standard drawing
GC-E1048.
Bundles that become susceptible to vibration can be retrofitted with vibration
control “slats” which are lengths of flat bar stock inserted between every other row.
The bars are typically 1 inch (25 mm) wide and 1/8 to 1/4 inch thick (3.2 to 6.4 mm
thick), depending on the spacing between the tubes. When inserted, the bars should
be an “interference fit” with the tubes, meaning that each tube along the bar should
be in contact with the bar and pushed slightly out of position by the bar. This may
require alternating the thicknesses of bar stock in each slot.
At the perimeter of the bundle, another bar is rolled to match the tube limit circle
diameter and installed as a perimeter band. Both ends of the flat bars inserted
between the tubes are welded to this perimeter band so that they are held in position
within the bundle. The perimeter band is then tack welded to the tie-rods in the
bundle. Figure 500-6 depicts this arrangement.

Fig. 500-6 Vibration Control Slats Arrangement Example

Note If trying to control vortex shedding in the inlet and outlet regions, slats need
only encompass top or bottom four to five rows of tubes.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-23
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

530 Channel and Shell Design

This section covers mechanical design of the channel and shell for a shell and tube
exchanger. Refer to EXH-SU-2583 or EXH-SU-4764 for more details on channel
and shell construction for general service and high-pressure/alloy heat exchangers,
respectively.

531 General
Channel and Shell Thickness
• The channel and shell contain the two separated fluids in the exchanger. They
are almost always cylindrical in shape and follow rules and regulations of
ASME Code for structural integrity. ASME Code, Section VIII, establishes
minimum metal thickness of cylindrical channels or shells.

Stacking Restrictions
• Avoid stacking more than three exchangers. More than three can cause
maintenance, handling, and shipping difficulties.
• Piping and shell stresses in stacked exchangers should be within acceptable
limits per applicable Code (e.g.: ASME Boiler and Pressure Vessel Code,
Section VIII-I or B31.3 Piping Code).
• The lower shells of stacked removable-bundle heat exchangers should be
designed to withstand the superimposed loads due to exchanger operating
weight or bundle pull-out, without suffering distortion that could cause binding
of the tube bundles.

532 Body Flanges

Body flanges are used to permit disassembly and removal or cleaning of internal
parts of a heat exchanger. Integral flanges (hub or weld neck) are flanges that are
integral with the exchanger wall or neck. This type of flange is required on all
services. Use of loose flanges (slip on) is not permitted; any deviation should be
reviewed and approved by a heat exchanger specialist. See Figure 500-7.
For pressures over 1000 psig (68.9 barg), special closures should be considered,
such as bolted and gasketed integral channel and tubesheet, or screw plug closures.
Screw plug closures are cost effective when the product of shell ID in inches (mm)
and design pressure in psig (barg) exceeds 80,000 (140,000) and are discussed in
Section 533.

500-24 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Fig. 500-7 Body Flange Configuration

Flange
Bolt Hole

Shell

Gasket

Slip-On Body Flange (Avoid)

Flange Bolt Hole

Shell

Gasket

Int egral (Weld Neck) Flange (Required)

The ASME Boiler and Pressure Vessel Code establishes the minimum requirement
for a flange design and provides a method of calculation in Section VIII, Division 1,
Mandatory Appendix 2. Non-mandatory Appendix S provides some general
discussion of bolted closure design and practice. Historically, heat exchanger body
flanges designed by the Code rules have been prone to leakage. Chevron’s U.S.
refineries have found that leaking body flanges can be eliminated through the use of
controlled bolting practices that produce predetermined consistent gasket loading,
coupled with the use of graphite coated metal core gaskets. Section 1000,
Maintenance, of the CES Heat Exchanger and Cooling Tower Manual details
bolting practices. Stud load determination and resulting gasket loading requirements
for new flanges are discussed below in the ASME Code Flange Design and
Chevron’s Additional Requirements section.
Chevron practice is to design flanges in accordance with ASME Section VIII,
Division 1, Appendix 2 rules. Once the flange is designed, a check is made to
determine if sufficient gasket loading can be achieved with a bolt loading within a
specified range.

ASME Code Flange Design and Chevron’s Additional Requirements

Body flanges for heat exchangers are designed to the requirements of ASME
Section VIII, Division 1, Appendix 2. Also, the rigidity index “J” is calculated per
Section 2.14 of Appendix 2. The following criteria should be met:
– Rigidity Index “J” less than or equal to 1.0, except,
– Rigidity Index “J” less than or equal to 0.7 if flange ID is less than
24 inches (610 mm) and the design pressure is less than 400 psig
(27.6 barg).

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-25
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

The allowable stresses used in body flange Code calculations are reduced to
88 percent of the tabulated values given in ASME Section II, Part D, Table 1A. This
reduction is intended to reverse the change in design margin that ASME made in the
1999 addenda to the 1998 ASME Code. Chevron’s additional requirements assume
an allowable stress that was in effect prior to the change in design margin.
All calculations are made with dimensions in the corroded condition. No corrosion
allowance is applied to the face of the flange.
The ASME Code defines W1 as the bolt load required to balance the sum of gasket
reaction and the hydrostatic end force due to design pressure and maintain a seal in
service.

W1 = 0.785 G2 Pd + (2b 3.14 GmPd)

(Eq. 500-2)
Company practice defines W1 as the bolt load required to balance the sum of gasket
reaction and the hydrostatic end force in service, but modifies the equation to
include the load to seat the pass partition gasketing.

W1 = 0.785 G2 Pd + (2 b 3.14 G + PLPW) m Pd

(Eq. 500-3)
The ASME Code defines W2 as the bolt load required to seat the gasket at zero
pressure and ambient temperature.

W2 = 3.14 b G y
(Eq. 500-4)
Company practice defines W2 as bolt the load required to seat the gasket, but
modifies the equation to include the load to seat the pass partition gasketing.

W2 = (3.14 G b + PLPW) y
(Eq. 500-5)
where:
b = Effective gasket seating width, inches (mm), from Table 2-5.2,
ASME Section VIII-I, Appendix 2
G = Diameter at location of gasket load reaction, inches (mm),
Table 2-5.2, ASME Section VIII-I, Appendix 2
PL = Total length of gasket pass partition rib(s), inches (mm)
PW = Actual partition gasket width, inches (mm)
m = Gasket residual stress factor
Pd = Design pressure, psig (barg)
W1 = Bolt load required at operating conditions in lbf (N)

500-26 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

W2 = Bolt load required to seat gasket in lbf (N)

y = Required gasket seating stress, psi (MPa)
For graphite coated metal core gaskets (CGG and KAG type per specifications
EXH-SU-5150 and EXH-SU-5151), use m and y values of m = 3.75 and
y = 7,600 psi (52.4 MPa).
Channel covers for single pass heat exchangers are designed per ASME Code using
88 percent of the Code allowable stresses for the material. For multi-pass heat
exchangers, the channel cover plate thickness is calculated using the ASME Code
and TEMA guidelines using 88 percent of the Code allowable stress; the method
resulting in the thickest cover plate is used for the cover plate thickness. The TEMA
method is based on a cover deflection criteria to limit internal bypassing of fluid
across the pass partition plate(s) in the channel (see Figure 500-8).
Fig. 500-8 Channel Cover Dimensions

Once a flange has been designed per Code or TEMA requirements as modified
above, an additional set of calculations is done to determine if the target gasket
stress of 20,000 psi (139.7 MPa) (including ribs) can be achieved with a stud load
between 50 and 70 percent of yield stress for ASME SA-193-B7 studs.
The radial gasket width should be 0.5 inches (13 mm) unless the gasket stress
exceeds 20,000 psi (137.9 MPa) at a stud load equal to 50 percent of bolt yield
stress. The radial width is increased to reduce the gasket stress produced at this bolt
load. When the gasket radial width exceeds 1 inch (25 mm) to satisfy the criteria,
the allowable gasket stress may be increased to 30,000 psi (206.8 MPa).
If the gasket stress is under 20,000 psi (139.7 MPa) at a bolt load equal to
70 percent of bolt yield stress, then the number of studs, or size of the studs is
increased to meet the criteria.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-27
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Gasket stresses on both sides of the tubesheet must be calculated giving

consideration of pressure from both sides of the tubesheet.
Gasket stress is calculated using the following formula:

2
  S S   N S   S RA  –  0.7854 G ID  P 1 – P 2  
G STR = --------------------------------------------------------------------------------------------------------
-
2 2
  0.7854   G OD – G ID  + P L P W 
(Eq. 500-6)
where:
GID = Gasket ID, inch (mm)
GOD = Gasket OD, inch (mm)
GSTR = Gasket stress, psi (bar)
NS = Number of Studs
P1 = Pressure on the gasket side under consideration, psi (bar)
P2 = Pressure on opposite side of the flange, psi (bar), (0 for Shell or
Channel covers)
PL = Total length of partition gasket ribs, inch (mm)
PW = Width of Channel Partition Plate, inch (mm)
SS = Assembly Stud Stress, psi (bar)
SRA = Stud Root Area, square inch (square mm)
The gasket width determined by the above calculations supersedes traditional
industry calculations for minimum gasket width requirement based on “y” and “b”
values.
Once the final gasket width, number of studs and size of studs have been
determined, the flange is rechecked using ASME Section VIII, Division I,
Appendix 2 criteria with modifications described above.
A machining allowance of 0.125 inch (3.2 mm) is added to the final flange
thickness for field maintenance. This added thickness is not used in any of the
calculations and should be clearly indicated on the flange detail drawing.

Designing and Evaluating Body Flanges

Chevron personnel seldom design heat exchanger body flanges from scratch, but
often evaluate vendor designs or existing flanges. Two software packages are
commonly used to evaluate body flanges: COMPRESS pressure vessel design
software available from Codeware, and Chevron developed PCFLANGE program,
which is available on the Heat Exchange website. Both programs perform flange
design in accordance with ASME Section VIII-I Code, Appendix 2. When
evaluating existing flanges, Finite Element Analysis (FEA) can be used, however,
doing so goes beyond the scope of the Code. Such analysis should be done in
accordance with API Standard 579-1/ASME FFS-1, Fitness for Service
requirements.
500-28 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

COMPRESS is a complete vessel design package with a heat exchanger option;

PCFLANGE automates only the ASME Section VIII, Division 1 Appendix 2
calculations necessary for the evaluation of flanges. Appendix H of the CES Heat
Exchanger and Cooling Tower Manual describes the operation of the PCFLANGE
program.
When using PCFLANGE, it is important to note that for flanges which sandwich a
tubesheet with common bolting, bolt loads are not properly handled by PCFLANCE
and user intervention is required. Analyze each flange individually and determine
the required bolt load for each flange. The bolt loads will not be equal since the
gasket dimensions and configuration are likely different for each flange. Since the
flanges share the same bolting, the higher bolt load must be manually entered for
both flanges. PCFLANGE does allow the user to input arbitrary bolt stresses to
investigate flange stresses at high bolt stresses.
Flange design requires decisions regarding geometry, materials, gaskets, and bolts.
The design of a flange may be iterative, as the required bolting may dictate an
increase in flange OD, which may, in turn, increase the bolt size or number.
The flange ID is set by the shell ID, which is set by the process and thermal design
of the heat exchanger. The materials are dictated by the operating temperature and
the corrosive nature of the fluids. The gasket type is dictated by the anticipated
movement at the gasket surface due to thermal stresses and piping stresses. The bolt
size and number and the flange thickness are dictated by the pressures and stresses.
The flange OD is dictated by the bolt size and number.
Experience in domestic refining has found that graphite coated, metal core gaskets
per specifications EXH-SU-5150, Graphite-Covered Corrugated Metal CGG-Style
Gaskets, and EXH-SU-5151, KamProfile KAG-Style Gaskets, tolerate differential
movement at the gasket surface and performed well in heat exchanger body flanges.
Materials. The choice of flange and bolt materials is based on design temperature
and the corrosive nature of the process fluid. The stress should be below creep stress
limits at design fluid temperature. Flange creep is not a problem in low alloy steels
below 700°F (371°C). At temperatures above 700°F (371°C) creep may be a
problem. When designing flanges in this temperature range, consult the Materials
and Corrosion Team of ETC.
Gaskets. Selection of the proper gasket is essential in flange design. Gasket
resilience, the ability of a gasket to maintain a seal when the two mating flanges
move relative to each other, is an important gasket parameter. Gasket parameters
required for a flange analysis include the gasket ID, OD, and the Code values of
minimum seating stress and gasket factor. See Section 540 for recommended gasket
materials. The preferred gaskets for heat exchanger body flanges, in order of
preference, are:
1. Graphite-Covered Corrugated Metal (CGG-Style) Gaskets per specification
EXH-SU-5150
2. KamProfile (KAG-Style) Gaskets per specification EXH-SU-5151
Both of the above gaskets are discussed further in Section 540.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-29
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Spiral Wound (SWG-style) gaskets are also discussed further in Section 540.
Double jacketed and solid core gaskets are not acceptable for heat exchanger service
as these types of gaskets do not tolerate differential movement at the gasket surface
and have demonstrated poor sealing performance.
Specify a 125 to 250 micro-inches (3.2 to 6.3 micrometers) finish on flange gasket
surfaces for all gaskets. Nubbins are not to be used on gasket surfaces.
All body flange joints with graphite coated gaskets are required to be hot-torqued
when the units are placed in service. All studs are re-torqued to the initial specified
torque once the flange surface is between 250 and 400°F (121 and 204°C). Since
floating head flanges cannot be hot torqued, the initial bolt load is set high to
achieve an initial gasket sealing stress of 20,000 to 25,000 psi (137.9 to 172.4 MPa)
to allow for relaxation.
Flanges, Bolts, and Washers. The flanges and bolts should be of material with
similar coefficients of thermal expansion, i.e., B-7 studs for low alloy flanges. If the
materials of the flanges and bolts are not similar, an analysis should be done to
confirm that differential thermal expansion at design fluid temperature will not
unseat the gasket or yield the bolts or flanges.
Bolt relaxation (creep) is a function of both temperature and actual bolt stress.
Given the typical bolt loads applied to flanges, the following temperature limits
should not be exceeded for ASME SA-193-B7 and ASME SA-193-B16 studs.
• For B7 Bolting: 700°F (371°C) process temperature
• For B16 Bolting: 950°F (538°C) process temperature
Note The above limits apply to uninsulated body flanges.
Floating head flange bolting is wetted by the process directly. Typically, B7M
bolting is used and the same process temperature limit given above for B7 bolting
applies. The difference is that in service, the floating head bolting will be at a lower
state of stress compared to body flange bolting because floating head bolting is not
hot torqued.
Hardened washers, per ASTM F-436, should be installed under each nut, or at least
each nut to be torqued. Hardened washers are not used on floating head joints due to
concern of environmental cracking of the washer.
Code rules, as indicated in Appendix S of the ASME Pressure Vessel Code,
recognize that normal bolt-up practices are not precise. Actual loadings often
significantly exceed design loads. For example, in order to hydrotest a code
designed flange, bolt stress must exceed Code allowable by about 50 percent.
Section VIII, Division 1, rules are intended to permit this practice.
Use the smallest bolts that will satisfy the flange design requirements; spacing
requirements and flange dimensions are shown on TEMA Standards, Table D-5.
The number of bolts should be divisible by 4 to conform to symmetrically oriented
bolting equipment. Bolt area should be calculated based on the thread root area
shown on TEMA Standards, Table D-5. Bolt hole diameter should be 1/8 inch
(3.2 mm) larger than bolt diameter.

500-30 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Flange Geometry and Stresses. Flange thickness and hub dimensions are the main
variables that control the magnitude of the stresses in the flange. The COMPRESS
and PCFLANGE programs print out the stresses in the various parts of the flange
and the corresponding code allowable limits. The programs can be run with various
flange thicknesses and hub dimensions until all the stresses are at or below code
allowables. In PCFLANGE, arbitrary bolt stresses can be specified in the program
to investigate the resulting flange stress at high bolt stresses.
Flange Rotation. As flanges are stressed by forces at the bolt circle, gasket, and
shell, they pivot, or bend, about the bolt circle and gasket. This bending is called
rotation. All flanges rotate to some degree, even at low stresses. The rotation is
usually not significant for small flanges, however, it can become significant for
large diameter flanges and for high stresses. In cases of extreme high stress, the
flanges can rotate until metal to metal contact exists between mating flanges at the
flange OD. Depending on the flange geometry and location of bolt stop rings,
rotation can sometimes unload gaskets and cause leaks.
To approximate flange rotation, the flange is considered a free body, disregarding
metal in the hub and the restraint of the nozzle neck or shell. These assumptions
result in the following equation, which slightly over-estimates the actual rotation.

 = 1.91 M R / (E b t3)
where:
 = Angle of rotation, radians
M = Total moment (in lbf [N m])
R = Mean radius of flange (in [m])
b = Radial width of flange (in [m])
t = Thickness of flange (in [m])
E = Modulus of Elasticity of flange at temperature (psi [MPa])
When calculating the total moment, M, acting on the flange, consider the bolt load
to be acting at the bolt circle, the hydrostatic load at the inner edge of the flange (if
pressured conditions are being considered), and the gasket reaction at the mean
gasket diameter. Then calculate the total moment on the flange relative to the mean
flange radius. See Figure 500-9.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-31
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Fig. 500-9 Calculating Moment for Flange Rotation Equation

With the rotation and the flange dimensions, the deflections at any point of interest
can be calculated. For example, deflection at the flange OD for rotation about the
gasket is shown below:

d = l
where:
d = Deflection at OD of flange (inches [m])
 = Angle of rotation (radians)
l = Radial distance from center of gasket to flange OD (inches [m]).
Flange rotation can cause problems if deflection at the flange OD
approaches 1/2 the gasket thickness.

500-32 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Flange Assembly and Bolting Practice

Bolt-up. Section 1000 of the CES Heat Exchanger and Cooling Tower Manual
discusses proper assembly of flanges and bolting requirements. Detailed assembly
requirements are also covered in GEN-SU-5209. Following is a summary of the
information presented in Section 1000 and GEN-SU-5209:
1. Inspect the gasket seating surfaces for tool marks, cracks, scratches or pitting.
Radial tool marks on a gasket seating surface are virtually impossible to seal.
2. Inspect the gasket for defects or damage. If the graphite coating on a gasket is
damaged to the point that the core is exposed, it can be repaired using adhesive
backed APX-2 graphite.
3. Inspect bolts, nuts, and washers for obvious defects. New studs, nuts and
hardened washers should be used each time the joint is assembled after being in
service.
4. Liberally lubricate all threads and nut facings with a high temperature
molybdenum disulfide based anti-seize thread lubricant such as Jet-Lube 550.
“Jet-Lube Riser” should be used in lieu of Jet-Lube 550 for studs in blind holes
or when service temperature exceeds 700°F (371°C). “Jet-Lube Riser” bolt
lubricant requires special hot torque instructions. The coefficient of friction for
“Jet-Lube Riser” lubricant varies with temperature, resulting in “torque
coefficient” or “nut factor” used in the torque equation that is not linear with
temperature. The torque value used when hot torquing a flange set with “Jet-
Lube Riser” lubricant may need to be adjusted accordingly. Consult with a
gasket and sealing specialist prior to using “Jet-Lube Riser” bolt lubricant.
5. Install the graphite coated gasket in place using a light coat of 3M Super 77
spray adhesive. A few dabs of heavy grease may be used as an alternative.
6. Assemble the flanges, install all the studs and washers and hand tighten nuts
(hand wrench tight). As a minimum, hardened washer should be placed under
each nut to be torqued. No hardened washers are used on floating head flanges
due to concern of environmental cracking of the washer and space limitations.
7. For flanges with 24 studs or more, select four studs equally spaced around the
flange and tighten to 100 percent of the specified torque with a clicker type
torque wrench. The required torque is that torque that gives stud stress required
to provide a gasket sealing stress of 20,000 psi (137.9 MPa) based on full
gasket width and design pressure(s) as discussed in Section 532 under the
heading “ASME Code Flange Design and Chevron’s Additional
Requirements.” For a properly designed flange, this stud stress will be between
52 and 74 ksi (360 and 510 MPa). See the Torque equation below to determine
torque based on stud stress.
8. Torque all studs to 100 percent of the required torque value following a circular
pattern. See Figure 1000-8 in Section 1000 of the CES Heat Exchanger and
Cooling Tower Manual.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-33
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

9. Torque all studs a second time to 100 percent of the required torque value to
compensate for relaxation in the previous pass.
10. Hot-Torque: When the unit is placed in service, re-torque all studs to
100 percent of the required torque when the flange temperature (service
temperature) is between 250 and 400°F (121 and 204°C). If the flange
temperature never reaches 250°F (121°C), re-torque all the studs three days
after operating temperature is reached. The graphite on graphite facing gaskets
will “flow” as it heats up, resulting in a reduced gasket stress and reduce bolt
load. Above 400°F (204°C), the thread lubricant begins to burn off, increasing
the friction between the nut and stud, which limits the stud stress achieved at a
given torque.
Floating head flange assemblies are not hot-torqued; the initial gasket stress is set
between 20 and 25 ksi (138 and 172 MPa) to compensate for effects of temperature
on the graphite and joint relaxation.
Torque. Stud stress or “stud load” in psi is a parameter in flange design. This stress
value can be converted to bolt torque using the following equation.

T = K x (Snd / 12) x Srd x SS

(Eq. 500-7)
where:
T = Torque, ft-lbf (N-m)
K = Torque Coefficient or Nut Factor (K = 0.17 when Jet Lube 550 is
used)
Snd = Stud nominal diameter, inches (m)
Srd = Stud root area, inches2 (m2)
SS = Stud Stress, psi (Pa)
Note For SI units, delete the division by 12 from the equation.
The equation for torque assumes a friction coefficient of 0.17. The torque
coefficient K depends on the type of lubricant and condition of the threads. Use a
torque coefficient of 0.17 for the recommended molybdenum disulfide based Jet-
Lube® 550 lubricant and new studs and nuts. This value is based on field data on
heat exchanger body flange studs with load cells under the nuts. Proper lubrication
of the stud and use of a calibrated torque wrench or tensioning head is a critical step
to ensure that the target gasket load is achieved.
Regularly calibrated torque wrenches are adequate for bolting well designed
flanges. Many other devices and methods have been used for precision bolting, such
as bolt elongation measurements and control of nut methods. Mechanical tensioning
devices may be required for large bolts that are beyond hand torque wrench
capabilities.

500-34 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Dealing With Leaking Body Flanges

With proper flange design, gasket selection and proper bolting practice, heat
exchanger body flange leaks can be eliminated. Section 1000 of the CES Heat
Exchanger and Cooling Tower Manual and the “Bolting Up” section of
GEN-SU-5209, as referenced in Section 1000, provide information and steps to take
to prevent flange leaks.

533 High Pressure Closures

The use of heat exchanger closures other that TEMA Type A, B, and C type
gasketed should be considered when the product of design pressure in psig (barg)
times diameter in inches (mm) is greater than 80,000 (140,000). In these cases,
TEMA type D Screw Plug closures are recommended.
In the past, for heat exchangers over 20 inches (508 mm) in diameter, with design
pressures in excess of 1,000 psig (68.9 barg), TEMA type C Integral Tubesheet and
Channel with Seal-Welded Diaphragm Closures were recommended when the
product of design pressure (psig [barg]) and diameter (inches [mm]) was greater
than 20,000 (35,000) and less than 80,000 (140,000). To access the tube-side, the
seal-welded diaphragm must be cut from the channel. When used with a high
pressure shell, the bundle is removable, however it is welded in place and the weld
must be machined out. Many closures of this type are in service in Refineries and
elsewhere.
However, the cost associated with opening and inspecting these types of closures
has been estimated to be as much as 250 thousand dollars (US) on a seven
day/week, 24 hr/day Turnaround work schedule basis.
With advancement in gasket sealing technology and bolting practice, Chevron has
moved away from recommending seal-welded diaphragm closures in favor of
integral tubesheet and channel designs with gasketed bolted closures for this
intermediate range of diameters and pressures.

Recommended High Pressure Closure Configurations—Screw Plug

There are two types of Screw Plug Heat Exchanger Closures:
• Type “A” is a high pressure tube-side, high pressure shell-side configuration
with a removable bundle.
• Type “B” is a high pressure tube-side, low pressure shell-side configuration
with an integral tubesheet and channel and removable shell. Figure 500-10 and
Figure 500-11 depict the two configurations.
Both the Type A and Type B designs employ an ACME threaded Lock Ring and
Channel Cover assembly behind a Diaphragm Plate (Seal Plate) and gasket that
contains the hydrostatic end force. In some cases, the Lock Ring and Channel Cover
may be a single piece. All components inboard of the Diaphragm Plate are wetted
by the process fluid; all components outboard of the Diaphragm Plate are not
wetted.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-35
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

The Type A Design is configured as an integral shell and channel with removable
bundle. There are two sets of Jack Screws—an inner set and an outer set. The inner
set of Jack Screws loads the Tubesheet to Shell Gasket through a load path
consisting of the Jack Screws bearing against the Inner Push Rods, Inner
Compression Ring, Diaphragm Plate, Internal Sleeve, Internal Flange, and Internal
Partition Assembly which bears against the Tubesheet, and loads the gasket. This
gasket can also be loaded by the Internal Flange Jack Screws—this is typically done
to seal the Tubesheet against the Shell for a shell-side hydrotest.
The outer set of Jack Screws loads the Diaphragm Plate Gasket which seals the
process from the environment. The load path consisted of the Jack Screws bearing
against the Outer Push Rods, Outer Compression Ring, and Diaphragm Plate which
loads the gasket.
The tubesheets in Type A units are designed for shell-to-tube side differential
pressure only. The differential pressure is of the magnitude of 500 psi (34.5 bar) or
less.
The Type B Design is configured as an integral channel and tubesheet with a
removable shell. The shell is bolted with a gasket to the back side of the channel
assembly. In the Lock Ring there is a single set of Jack Screws which serve the
same purpose as the outer Jack Screws in the Type A Design. The Jack Screws load
the gasket behind the Diaphragm Plate through a load path consisting of the Jack
Screw, Push Rod, Compression Ring, and Diaphragm.

500-36 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Fig. 500-10 Type “A” Screw Plug Heat Exchanger Closure: High Pressure Tube-Side, High
Pressure Shell-Side Configuration with Removable Bundle.

1. Tubesheet to Shell Gasket 2. Tubesheet

3. Internal Partition Assembly 4. Packing Gland Assembly
5. Packing Gland Assembly Cover 6. Partition Cover Gasket
7. Partition Cover Plate 8. Internal Flange
9. 3 Piece Split Ring (Key Ring Assembly) 10. Internal Flange Jack Screw
11. Internal Sleeve 12. Diaphragm Plate Gasket
13. Diaphragm Plate 14. Outer Compression Ring
15. Inner Compression Ring 16. Channel Cover
17. Inner Jack Screws 18. Threaded Lock Ring
19. Outer Jack Screw 20. Inner Push Rod
21. Outer Push Rod

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-37
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Fig. 500-11 Type “B” Screw Plug Heat Exchanger Closure: High Pressure Tube-Side, Low
Pressure Shell-Side Configuration with Internal Tubesheet and Channel with
Removable Shell.

Parts Gaskets Required

1. Shell Flange to Channel 2. Channel to Shell A. Diaphragm
3. Push Rod 4. Jack Screws B Tubesheet
5. Channel to Shell Bolts 6. Threaded Lock Ring C Partition Box
7. Channel Plug 8. Diaphragm (Seal Plate)
9. Partition Box 10. Tubesheet
11. Compression Ring 12. Partition Cover Plate
13. Tie Rods (Not Shown) 14. Vent

Other High Pressure Closure Configurations—Welded Diaphragm and

Tubesheet
Though no longer recommended, Figure 500-12 shows a welded-diaphragm closure
and welded-in tubesheet used for feed-effluent exchangers with high-pressure on
both shell and tube sides. The tubesheet is welded to an internal sleeve/partition box
assembly which is welded to the channel. The sleeve-to-channel strength weld is
located near the channel end to facilitate access for removal and re-welding. The
inconel overlay at this location permits rewelding without heat treatment.

500-38 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

The channel closure consists of a thin diaphragm seal plate welded to the front of
the channel and a bolted cover plate. The cover plate and bolting withstand the
hydrostatic end force and the diaphragm seal weld provides the seal for the process
fluid.

Fig. 500-12 Removable Tube Bundle with Welded Tubesheet and Diaphragm Closure

Though no longer recommended, Figure 500-13 shows a typical welded-diaphragm

closure for a high-pressure channel and low-pressure shell. This closure consists of
an integral channel and tubesheet, gasketed and bolted-on shell, and a welded
diaphragm seal plate backed by a bolted cover plate. The bolted cover plate
withstands the hydrostatic end force and the seal welded diaphragm seals the
process fluid.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-39
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Fig. 500-13 Integral Tubesheet and Channel with Seal-Welded Diaphragm Closure
Channel

Closure Weld
See Det ail
Tube
Pass Par tition

Cover Pla t e
Tube
Sheet
Diaphragm

Channel Flange

Diaphram Seal Weld

Inconel
Overl ay Cover Pla t e
Diaphram

Det ail

Designs described above in Figure 500-12 and Figure 500-13 are no longer
recommended for the following reasons:
• Improvements in gasket sealing technology, gasket selection, assembly and
bolting practices allow high pressure applications to be sealed as gasketed
closures.
• The diaphragm and tubesheet retaining welds have to be cut from the unit using
in-field machining tools operated by specialty field machining contractors, or
removed by grinding or arc-gouge as space permits. Some machining will be
required to clean up the surfaces after arc-gouging prior to re-welding.
• Typically, the shell and channel on these units are fabricated from Chrome-
Moly material; an Inconel or stainless steel overlay is provided in the channel
where the diaphragm and tubesheet skirt are welded into the shell (or, the entire
channel is overlaid for corrosion protection); the overlay, diaphragm, and
tubesheet skirt are typically of the same material such that no post weld heat

500-40 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

treatment is required when re-welding the diaphragm and tubesheet skirt to the
channel. However, if excess overlay material is inadvertently or accidently
removed such that the base metal is nearly or completely exposed, a difficult
field post weld heat treatment will be required on the thick channel section and
shell after additional overlay material is applied.
• Prior to machining out the diaphragm plate, the channel studs must be removed
for access to the diaphragm weld. These studs can be difficult to remove and at
times, have had to be machined out, or removed by stud disintegration specialty
contractors. If the female threads are damaged in the process, oversized studs or
“step-studs” may have to be installed, resulting in additional assembly time and
cost.
• Re-welding a process contaminated material can be difficult due to weld
cracking or porosity in the base material.

Other Considerations—High Pressure Closures

High pressure closures are not commodity items. Industry standards and appropriate
codes for their design do not exist. Few manufacturers can design safe, fully
serviceable high pressure exchangers. Some of the common problems and solutions
are discussed below. Chevron has written specification EXH-SU-4764, Appendix E,
that outlines requirements for the design and construction of screw plug closures.
Standard bolting practices that are appropriate for hardened studs and nuts should
not be used in high pressure closures. Bolt stresses should be less than yield stress
and relaxation stress of the soft base metal of the female threads.
Stainless steel internals with clad low chrome channels have caused many problems
for high pressure closure designs. Welded stainless steel, pass-partition plates in low
chrome channels have caused fatigue cracking and catastrophic channel failure.
Stainless steel pass partition plates in low chrome channels should be made bolted
with adequate clearances to accommodate differential thermal expansion.
Screwed covers in Figure 500-10 and Figure 500-11 and have jammed due to
inappropriate maintenance practice, thread corrosion and severe thermal transients.
Large threads and clearances mitigate this problem. Specification EHX-SU-4764
calls for 1-inch pitch ACME Threads (minimum).
One non-Company exchanger (similar to Figure 500-10) failed catastrophically due
to an event that resulted in expanding the threaded portion of the channel and
allowing the cover to disengage. Some similarly designed Company exchangers
have been modified with larger threads, new covers and externally stiffened channel
ends to prevent this possibility.
High pressures closures should be carefully specified including:
• Materials of all components
• Applicable Chevron specifications
• Steady state design conditions
• Any transient conditions that may be encountered and their frequency
• The design life of the exchanger

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-41
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Normal plant startup and shutdown transients have no impact on design. Feed pump
failure transient with continuing recycle gas and full reactor effluent flow should
probably be considered for all feed/effluent exchangers in hydroprocessing plants.
Reactor temperature excursions followed by rapid depressuring should be
considered in plants where excursions are possible. Transient thermal stress analysis
effects clearances and minor design details and has minimal effect on equipment
cost.

534 Connections
This section discusses recommendations for piping and instrument connections on
the body of shell and tube exchangers. In general the number of connections should
be minimized to minimize sources of leaks.

Inlet and Outlet Nozzles

Inlet and outlet nozzles for shell side and tube side fluids are required and they
should conform to Chevron Piping Standards governing the attached piping as a
minimum. Nozzle size should always match the diameter of the piping attached to
the nozzle. There should be no expansion or contraction directly upstream or
downstream of a nozzle.
Connections NPS 1-1/2 and larger should be flanged; flanges are to be in
accordance with ASME B16.5 and pressure class is to match or exceed mating
piping flange pressure class. Slip-on flanges are not to be used. Projection of nozzle
should be such that the flange bolts can be easily removed from either side of the
flange without removing insulation from the heat exchanger.

Inlet and Outlet Nozzle Sizing

Inlet and outlet nozzles should be sized such that the pressure drop through each
nozzle, whether on the shell or channel, is 5 to 10 percent of the overall pressure
drop across the exchanger. Nozzle pressure drop is calculated using a velocity head
loss type equation as follows.

delPn = Kn Vn2 / (2 x G)
(Eq. 500-8)
where:
delPn = pressure drop across nozzle, lb/ft2 (N/m2)
Kn = velocity head loss term for a single nozzle
= 1.1 for tubeside inlet nozzle (radial position)
= 0.7 for tubeside outlet nozzle (radial position)
= approximately 1 + [Dn / (4 HE + 0.15 Dn)]2 for a shell-side
nozzle. Typical range is 1.3 (outlet) to 3 (inlet) for a bundle with
impingement rods.
Dn = nozzle inside diameter, inches (mm)

500-42 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

HE = height under nozzle open for flow, inches (mm)

 = fluid density, lb/ft3 (kg/m3)
Vn = velocity through nozzle, ft/s (m/s)
G = gravitational constant, 32.17 ft/s2
Note For SI units, set “G” term equal to 1.

Vent and Drain Connections

All high and low points on the shell and tube sides of an exchanger not otherwise
vented or drained by nozzles or piping should be provided with NPS 1-1/2 (min)
flanged connections for vent and drain. Consider where maintenance blinds will be
installed in determining the need for, or location of, vent and drains. Condensers and
steam heated exchangers can require dedicated vents which operate continuously to
vent non-condensable gases from the unit. Threaded connections should not be
used.

Pressure Gage Connections

All inlet and outlet nozzles NPS 3 or larger should be provided with a NPS 1-1/2
flanged horizontal connection for a pressure instrument unless special
considerations require it to be omitted. Threaded connections should not be used.

Thermometer Connections
All inlet and outlet nozzles NPS 4 or larger should be provided with a NPS 1-1/2
(min) horizontal connection for a thermowell unless special considerations require it
to be omitted. Threaded connections/thermowells should not be used.

Chevron Practices: Construction

1. Chemical cleaning connections, when required, should consist of a pair of
flanged and blinded nozzles, one at the inlet and the other at the outlet of each
heat exchanger (or each series-connected group of heat exchangers). Waste heat
boilers with stayed fixed tubesheets should always have shell-side chemical
cleaning nozzles. The chemical cleaning connection should be made as
branches on the heat exchanger nozzles. The size of the chemical cleaning
connections should depend on the size of the heat exchanger nozzle to which
they are attached, as shown in Figure 500-14.
2. To double as inspection openings, the following nozzles should be NPS 4 or
larger: (a) blowdown nozzle on steam generators, and (b) steam condensate
outlet nozzle on vertical reboilers.
3. Steam heated vertical reboilers should be provided with a shell side vent as
close to the top tubesheet as possible to relieve air binding and corrosion due to
noncondensibles. Steam heated horizontal reboilers should be provided with a
channel side (steam side) vents on top of the channel and just below the
partition plate to relieve air binding and corrosion due to noncondensibles.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-43
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

4. Kettle-type steam generators should have one manhole located either in the
shell above the tube bundle or in the shell cover.
5. Threaded connections should not be used.
6. For steam heated heat exchangers, condensate lines should be free
draining.

Fig. 500-14 Nozzle Sizes for Chemical Cleaning

Chemical Cleaning
Nozzle, NPS Shell Side Nozzle, NPS Tube Side Nozzle, NPS
2 2 to 4 2 to 6
3 6 8
4 8 and larger 10 and larger

540 Gaskets
Selection of the proper gasket is essential to achieve a leak-free flanged joint. The
ability of a gasket to maintain a seal when the two mating flanges move relative to
each other, is an important gasket parameter. The preferred gaskets for heat
exchanger body flanges, in order of preference, are:
1. Graphite-Covered Corrugated Metal CGG-Style Gaskets per specification
EXH-SU-5150
2. KamProfile KAG-Style Gaskets per specification EXH-SU-5151
Spiral Wound Gaskets (SWG) are also discussed below; KamProfile gaskets are
general used in place of spiral wound gaskets. Double jacketed and solid core
gaskets are not acceptable for heat exchanger service as these types of gaskets do
not tolerate differential movement at the gasket surface and have demonstrated poor
sealing performance.

Corrugated Metal Graphite Covered Gaskets

Corrugated Metal Graphite Coated Gaskets (CGG) have been found to perform the
best in terms of leak free operation, can tolerate at least 0.050 inch (1.3 mm) of
radial movement between flange faces, and are cost effective. This type of gasket
should be used in all body and floating head flanges in non-oxidizing services up to
800°F (427°C). A CGG gasket consists of a 0.030 inch (0.76 mm) thick corrugated
304L or 316L core coated both sides with 0.015 to 0.020 inch (0.38 to 0.51 mm)
thick Polycarbon APX-2 graphite. The minimum ring width for a CGG gasket is
1/2 inch (13 mm); the minimum width for pass partition ribs is 3/8 inch (9.5 mm)
(ribs are not corrugated). CGG gaskets are readily available from gasket
manufacturers. Figure 500-15 shows a picture of a typical CGG gasket. The
graphite is removed from the right portion to show the corrugated core.

500-44 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Fig. 500-15 CGG Gasket with Half the Graphite Removed

KamProfile Graphite Covered Gaskets

KamProfile gaskets have similar sealing characteristics as CGG gaskets and can be
used if a narrower gasket than a CGG is required due to space constraints.
KamProfile gaskets consist of an 1/8-inch (3.2-mm) minimum thick serrated 304L
or 316L core coated both sides with 0.020-inch (0.51-mm) thick Polycarbon APX-2
graphite. The minimum width is 3/8 inch (9.5 mm). Figure 500-16 shows a picture
of a typical KamProfile gasket. The graphite is removed from the right portion to
show the serrated core.

Fig. 500-16 KamProfile Gasket with Half the Graphite Removed

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-45
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

For services with design temperatures exceeding 800°F (427°C), KamProfile

gaskets can be used with a strip of mica (COGEBI, Novus High Temp, or
Thermiculite), 1/8 inch (3.2 mm) wide minimum, laid on the gasket core adjacent to
the outer edge of graphite. The mica acts as an oxygen barrier, protecting the
graphite from oxidation at high temperature. A 3/8 inch (9.5 mm) minimum width
of graphite is required to achieve a reliable seal. A strip of mica may also be applied
to the core adjacent to the inner edge of graphite to protect against oxidation in
some cases. Use of mica on KamProfile gaskets should be discussed with a bolting,
gasket, and sealing specialist.

Spiral Wound Gaskets

Spiral wound gaskets are a sound technology and are used in heat exchangers in
some special circumstances such as heat exchangers that are designed for brief high
temperature conditions such as reactor regenerations. Chevron is moving away from
spiral would gaskets; they require higher m and y values than graphite-coated
corrugated metal (CGG) and KamProfile gaskets which can result in thicker flange
designs. If specified, spiral wound gaskets must be constrained on the inner and
outer diameters by a tongue and groove joint design or inner and outer rings as
spiral would gaskets are prone to inward buckling. A KamProfile gasket is usually
specified instead.
They are custom-designed to meet the compression requirement of body flange
bolting. Spiral-wound gaskets are fully seated when the flanges are pulled up snugly
against the compressing guide ring, or brought into metal to metal contact in the
case flanges with a recess gasket groove. This prevents gasket crushing by over-
tightening of bolts. The large diameters make them hard to handle—the windings
tend to pop out of the guide rings. A spiral-wound gasket is shown in Figure 500-17
below.

Fig. 500-17 Spiral-Wound Gasket

500-46 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

Spiral wound gaskets with PTFE (Teflon) filler material are recommended in some
acid services where the acid will oxidize graphite materials. Figure 500-18 lists
suitable gasket materials for acid service.

Fig. 500-18 Suitability of Gasket Materials in Acid Service(1) (2)

Graphite
Service Gasket Type Material PTFE Suitable?
Suitable?
H2SO4 Spiral Wound Alloy 20 Yes No
HNO3 Spiral Wound 304L Stainless Steel Yes No
HF KamProfile or Monel Yes Yes
Spiral Wound
H3PO4 Spiral Wound 316L Stainless Steel Yes No
HCI KamProfile or Hastealloy C Yes Yes
Spiral Wound
(1) This table gives conservative recommendations for materials resistant up to at least 200°F.
Please consult with the Materials and Corrosion Team when selecting gaskets for a new chem-
ical service.
(2) Applies to all concentrations of acids.

Recommended Gasket Seating Stresses:

The recommended full width sealing stresses for CGG, KamProfile, and spiral
wound gaskets are as shown in Figure 500-19:

Fig. 500-19 Recommended Gasket Full Width Sealing Stress(1)

Minimum Recom-
mended Sealing
Stress for Optimum Maximum Gasket
Minimum Sealing Performance in psi Sealing Stress in psi
Gasket Type Stress in psi (MPa) (MPa) (MPa)
Graphite Covered 8,000 (55.2) 10,000 (68.9) 40,000 (276)
Corrugated Metal
Gasket
KamProfile Gaskets 8,000 (55.2) 10,000 (68.9) 40,000 (276)
Spiral Wound 10,000 (68.9) 20,000 (138) 66,000 (455)
Gaskets
(1) Sealing Stresses are based on full width of gasket.

Hot Torquing
All body flange joints with graphite coated gaskets are required to be hot-torqued
when the units are placed in service. All studs are re-torqued to the initial specified
torque once the flange surface is between 250 and 400°F (121 and 204°C), or three
days after being placed in service if the flange surfaces do not reach 250°F (121°C).
The graphite on graphite facing gaskets will “flow” as it heats up, resulting in a
reduced gasket stress and reduce bolt load. Above 400°F (204°C), the thread
lubricant begins to burn off, increasing the friction between the nut and stud, which
limits the stud stress achieved at a given torque.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-47
500 Shell and Tube Exchanger Component Design Considerations Heat Exchanger and Cooling Tower Manual

Floating head flanges cannot be hot-torqued. Since floating head flanges cannot be
hot torqued, the initial bolt load is set high to achieve an initial gasket sealing stress
of 20,000 to 25,000 psi (138 to 172 MPa) to allow for relaxation.

550 Insulation

551 Reasons for Insulating

Exchanger shells, and channels are insulated for the following reasons:
• To minimize heat loss and consequently save fuel. Obviously, insulating may
not be appropriate in “cooling services,” such as for cooling water exchangers.
• To protect personnel working where surfaces are over 140°F (60°C).
Exchangers which are not readily accessible need not necessarily be insulated
to protect personnel. Exchanger shells which are accessible but should not be
insulated may use alternative means to protect personnel, such as expanded
metal guards, guard posts and signs.

552 Types of Insulation

Calcium silicate, Perlite, and mineral wool are the common types of insulation used
on exchangers. Calcium silicate is generally preferred, especially in areas of high-
foot traffic or where flammability is a concern. See Section 100 of the CES
Insulation and Refractory Manual for more information on the types of insulation
available. See IRM-SU-1381 for installation requirements.

553 Weatherjacketing
In general, 3/16-inch (4.8-mm) pitch cross-crimped aluminum weather jacketing
should be used on exchanger shells, and flat aluminum or mastic weather jacketing
used on exchanger heads. See Section 100 of the CES Insulation and Refractory
Manual, IRM-SU-1381, and standard drawing GD-N99785 for more detailed
information.

554 Flange Insulation

In general, heat exchanger body flanges should not be insulated. Insulated flange
studs cool down slower than the flanges following process upsets because the studs
are not in direct contact with the fluid in the exchanger. Consequently, the flanges
thermally contract at a faster rate than the studs which can lead to a loss of stud load
and gasket stress, resulting in a flange leak. Flange insulation is more common in
cryogenic services. Heat exchangers in cryogenic service may be excluded from this
rule and insulating of flanges should be considered on a case by case basis.
If personnel protection is required, apply an expanded metal enclosure instead of
insulation. Weather shields can also be used to keep rain from impinging directly on
the flange.

500-48 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. November 2009
Heat Exchanger and Cooling Tower Manual 500 Shell and Tube Exchanger Component Design Considerations

If a body flange must be insulated, consult with local gasket, bolting, and sealing
specialists or heat exchanger specialists. Insulating flanges should be handled on a
case by case basis. Flange stud loads will run hotter and be prone to further
relaxation, resulting in lower gasket loading and increased risk of leakage.
If the flanges and insulation are improperly designed, the following problems may
occur:
• The bolts will relax and stretch if the internal operating temperature is above
the creep stress limit of the bolts. This problem can cause leakage and possibly
auto-ignition because of the high temperatures.
• Insulation can “soak up” leakage and may lead to auto-ignition.
To prevent these problems, the following criteria should be used for design and
insulation of flanges:
• The flanges and bolts should be of similar material, i.e., B7 or B16 studs for
carbon steel or low alloy flanges.
• Flange and bolt materials should be designed for the maximum internal design
temperature and corrosive nature of the process fluid. In other words, the stress
in flange and bolt material must be kept below the creep stress limits at the
maximum internal design temperature.
• Use insulation covers designed for safe leakage. See Section 100 of the CES
Insulation and Refractory Manual and IRM-SU-4197 for the design of leak-
safe, removable insulation covers.
• Apply insulation when the flange is cold (after hydrotest and before startup) to
minimize startup stresses. Insulation may be temporarily removed after startup
to inspect for leaks. Adding and removing insulation from flanges in service
can result in flange leaks due to the impact on stud loads as the studs heat up or
cool down.

November 2009 © 1994–2009 Chevron U.S.A. Inc. All rights reserved. 500-49
600 Air Cooled Heat Exchangers

Abstract
This section provides general design guidelines for air cooled exchangers (also
called air coolers). See API Standard 661 for air cooler nomenclature and illustra-
tions and for mechanical design requirements.

Contents Page
610 General Design Considerations 600-2
611 Design Air Temperature
612 Tube Side Design
613 Viscosity Plugging
614 Headers
615 Layout
620 Draft 600-3
630 Fans 600-4
640 Fins 600-4
650 Winterization 600-4
660 Estimating Rules of Thumb 600-4

December 1989  1989 Chevron USA Inc. All rights reserved. 600-1
600 Air Cooled Heat Exchangers Heat Exchanger and Cooling Tower Manual

610 General Design Considerations

Air cooled exchangers are used to cool or condense low viscosity fluids. They are
usually designed by the manufacturer. The HTRI ACE computer program may be
used to rate air coolers and condensers, if air rate is known. Fan design and rating is
beyond the scope of the ACE program.

611 Design Air Temperature

Design air temperature is usually selected as the dry bulb temperature that is
exceeded about 150 hours per year during the hottest 4 consecutive months. This
corresponds to meeting design performance 98+% of the year, but not for a few
hours a day during the hottest days.
Evaluated weather data can be obtained from the following sources:
• “Evaluated Weather Data for Cooling Equipment Design,” by Fluor Products
Company, 1958.
• “Army, Navy and Air Force Manual Engineering Data,” by Departments of the
Army, the Navy, and the Air Force, April 1, 1963.
• “Evaluated Weather Data for Cooling Equipment Design, Addendum No. 1,
Winter and Summer,” by Fluor Products Company, 1964.
Consult ETD Engineering Analysis Division if these or other weather data sources
are not readily available.
Air cooling process streams to about 20F over the design ambient temperature, or
higher, is economic. Water cooling or refrigeration is used if lower process tempera-
tures are required.

612 Tube Side Design

Tube side design for air coolers and shell and tube exchangers involves the same
principles. Economic tube side velocities given in Section 220 and process side heat
transfer coefficient given in Section 481 apply to air coolers. Estimating rules of
thumb specific to air coolers are given in Section 660 below.

613 Viscosity Plugging

Cooling viscous fluids is usually not practical in air coolers because of viscosity
plugging. The viscosity of viscous oil usually increases rapidly with decreasing
temperature. Air side maldistribution inevitably causes some tubes to run cooler
than others. This increases the viscosity in the cooler tube and therefore reduces
flow through that tube, which causes further cooling and more flow reduction. The
result of this temperature-viscosity-flow effect is to virtually stop flow in some
tubes and force flowing tubes into the turbulent flow regime, if the viscous oil
pumps can develop the needed head. Otherwise, flow essentially stops.

600-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 600 Air Cooled Heat Exchangers

Air coolers should not be used when the tube side outlet Reynolds number (defined
in Appendix B) is less than 2000.
Air recirculation described in Section 650 may be used to air cool fluids that
become very viscous or freeze near ambient temperatures. Tempered water cooling
described in Section 330 is an alternative.

614 Headers
Most air coolers have header plugs opposite each tube on each end to facilitate
inspection and cleaning. Removable cover plates on each header is an alternative
but is not economic and rarely used.

615 Layout
Air-cooled exchangers usually have four rows of staggered tubes. Six rows are
sometimes used for large process temperature range. Aligned tube rows are never
advantageous because of significant air side channeling and temperature maldistri-
bution.
Air-cooled exchangers are often mounted over pipeways to minimize plant real
estate. Pipeway mounted units are usually 30 feet long. Shorter bundles are common
for grade mounted units.
More than one bundle (service) may be in the same bay with a common pair of fans,
if overcooling some of the services is not a concern. If overcooling is a potential
problem, adjustable louvers for that bundle can be used, or the service can be built
as a separate unit with dedicated fans.

620 Draft
Forced draft is required when the maximum outlet air temperature is higher than the
manufacturer’s rated temperature of the fans or the auto-variable fan pitch control
hubs. Operation with the fans off and during steam out should be considered. The
air outlet temperature is approximately the same as the process inlet temperature
when fans are off, and same as the steam temperature during steam out. The plenum
of induced draft units acts like a stack and produces a significant “fans-off” duty.
The “fans-off” performance of forced draft units is negligible.
Induced draft is sometimes specified for a column overhead condenser where
temperatures permit. Induced draft is normally considered if overcooling during rain
storms is a problem or if design relief loads would be reduced by the superior fans
off performance of induced draft units. Rain affects the full bundle in forced draft
units and only about half the bundle in induced draft units.
When the above considerations do not govern draft, the manufacturer usually
decides. Induced draft provides better air flow distribution than forced draft;
however, induced draft fans must move a greater volume of (heated) air than forced
draft. These effects are not large.

December 1989  1989 Chevron USA Inc. All rights reserved. 600-3
600 Air Cooled Heat Exchangers Heat Exchanger and Cooling Tower Manual

630 Fans
There should be at least two fans per bundle with the fans covering at least 40
percent of the bundle. Half of the fans are usually autovariable pitch fans. During
the summer, the manual fan is on and the autovariable fan adjusts air flow to control
duty. During the winter, the manual fan is usually shut off. Auto-variable fan pitch
control is usually justified by fan power savings rather than control needs, but
serves both functions. Fan blades are usually aluminum or plastic. The best plastic
fan blades are good to about 350F. Auto-variable fan pitch control hubs may have
rubber seals that limit temperature to about 250F.

640 Fins
Fins are almost always aluminum. Three fin types are commonly used: footed, inte-
gral (extruded) and imbedded. Maximum process inlet temperatures for these fin
types are 250F, 550F, and 750F, respectively. API 661, Section 5.1, describes
these fin types in more detail.

650 Winterization
Winterization is required for air coolers with fluids that may become immobile
under expected winter conditions due to freezing, pour point problems or the
viscosity plugging phenomenon described in Section 610. Winterization usually
means building an enclosure around the air cooler with provisions to recirculate a
portion of the outlet air back to the inlet. A steam coil under the bundle is also
normally provided to heat up the bundle if the fluid in the bundle ever becomes
immobile. Winterization of air coolers is discussed in detail in API Recommended
Practice 632 (not included in this manual).

660 Estimating Rules of Thumb

Air coolers usually have four layers of 1-inch O.D. tubes with 10, 5/8-inch high fins
per inch, arranged in a 2-1/2-inch pitch equilateral triangular layout. The ratio of fin
surface to bare outside surface (bos) is about 20:1; the required plot space is about
200 ft2 per 1000 ft2bos; the air rate is about 600 lb/hr/ft2bos; the installed fan power
is about 20 hp per 1000 ft2bos; and the air side heat transfer coefficient is about
175 Btu/hrF  ft2bos .
Note that the installed fan power (“pumping power”) is four times higher than that
normally used for other types of heat exchangers. This is because air cooler design
is limited by warm weather operation. Annual average fan power is usually about
one fourth of the design value.

600-4  1989 Chevron USA Inc. All rights reserved. December 1989
700 Plate and Frame Exchangers

Abstract
This section briefly describes the advantages and disadvantages of plate and frame
exchangers.

Contents Page
710 General Information 700-2

December 1989  1989 Chevron USA Inc. All rights reserved. 700-1
700 Plate and Frame Exchangers Heat Exchanger and Cooling Tower Manual

710 General Information

Plate and frame exchangers consist of thin-rippled alloy plates spaced about
1/4 inch apart. Gasketing (usually rubber) seals the edges of the plates and directs
the flow into and out of appropriate headers. The streams exchanging heat usually
flow in every other channel in pure counter-current flow. The general arrangement
is illustrated in Figure 700-1.

Fig. 700-1 Plate and Frame Exchanger

The advantages of plate and frame exchangers are:

1. Pure counter-current flow.
2. Compact.
3. Light weight.
4. Easily expandable (assuming excess pumping capacity).
The disadvantages of plate and frame exchangers are:
1. Less efficient than in-tube turbulent flow.
2. Less efficient than in-shell viscous flow over a staggered tube layout.
3. Limited to moderate temperatures (<250F) and pressures (<200 psig).
4. Superior metallurgy (noncorroding) required.

700-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 700 Plate and Frame Exchangers

5. Gasket deterioration is a problem with hydrocarbon mixtures.

6. Poor fire resistance because of large exposed bolts and normal gasket material
(usually rubber).
For turbulent flow, the best conversion of pressure drop to heat transfer is in a
straight tube where all friction pressure drop thins the laminar sublayer near the heat
transfer surface. Plates are less efficient as a portion of the pressure drop is
expended changing fluid direction around the plate ripples. Turbulent-flow heat
transfer in plates and in shells is equally efficient.
In viscous flow, thermal boundary layers limit heat transfer. Flow across a stag-
gered-tube layout is best with viscous flow because thermal boundary layers are
discontinuous from row to row. Viscous flow in plates exhibits the same rapid
decrease in heat transfer with increasing viscosity as in-tube flow (see Figure 200-4
in Section 213).
HTRI tests have shown that fouling in plates, in tubes, and in shells is about the
same for the same expenditure of pumping power (pressure drop) per unit of heat
transfer surface.
Plate and frame exchangers may not be cost-competitive when compared to a well-
designed shell and tube exchanger unless the following applies: weight, space,
and/or expandability are highly valued; the same alloy is required for both types of
exchangers; pressures and temperatures are moderate; fire hazard is nonexistent or
minimal; and fluids are compatible with rubber gaskets. At the same time, some
upstream experience suggests that plate and frame exchangers are cost competitive,
and they are worth investigating in most cases.
The use of titanium plate and frame exchangers on offshore platforms in sea
water/fresh water service is an appropriate application.

December 1989  1989 Chevron USA Inc. All rights reserved. 700-3
700 Plate and Frame Exchangers Heat Exchanger and Cooling Tower Manual

700-4  1989 Chevron USA Inc. All rights reserved. December 1989
800 Materials Considerations

Abstract
This section provides general background on exchanger materials. It summarizes
major factors that must be considered in selection of materials for exchanger
components and in the exchanger design.

Contents Page
810 Major Component Materials 800-2
811 Tubes
812 Tubesheets
813 Baffles
814 Shell
820 Minimum Pressurizing Temperature 800-6
830 Sacrificial Anodes 800-7
840 Insulation 800-7

December 1989  1989 Chevron USA Inc. All rights reserved. 800-1
800 Materials Considerations Heat Exchanger and Cooling Tower Manual

810 Major Component Materials

This section suggests materials for components of shell and tube heat exchangers,
including tubes, tubesheets, baffles, and shell. The table at the end of this section
(Figure 800-3) provides a list of ASME materials commonly used for these compo-
nents.

811 Tubes
Tube Material
There often is no single correct material for a given service. Although the choice of
tube material is generally dictated by temperature and corrosion conditions, how
well a material performs is greatly influenced by actual service conditions and the
corrosion control measures in use. Keep this in mind when reading Figure 800-1,
which lists common exchanger tube materials. The information is not meant for
materials selection. For information about specific corrosives or specific types of
process plants refer to the Corrosion Prevention and Metallurgy Manual.
Selecting the right tube material is only one way to ensure good performance. Often
one can control conditions in the exchanger by altering the nature of the process
fluid or by controlling exchanger design. Also, corrosion-inhibiting chemicals may
be added to the process fluid.

Fig. 800-1 Common Exchanger Tube Materials (1 of 2) (This table is illustrative only. It is not suitable for materials
selection. See Section 811.)
Service Typical Materials Comments
Fresh Water Carbon Steel Short life unless water is good quality and
 chemical treatment is carefully controlled
 
Admiralty(1) Very few problems
 
304 Stainless Steel Only for low chloride waters under
nonscaling conditions
Sea Water Admiralty(1) Suffers end impingement at high velocity
 
 More resistant to end impingement than
70—30 Cupro-Nickel Admiralty
 
 Essentially corrosion-proof to 250°F. Special
Titanium grades OK to 450°F
Hydrocarbons-Sweet Carbon Steel Very sensitive to trace H2S over 500°F
Hydrocarbons-Sour Carbon Steel Limited to about 550°F maximum
 
5 Chrome—1/2 Moly Where too hot for carbon steel
Hydrocarbons-Naphthenic 316 Stainless Steel Above 1.5 neutralization number
Hydrogen-Sweet Carbon Steel, C-1/2 Mo, 1-1/4  Choice depends on temperature and
CR-1/2 Mo, 2-1/4 CR-1 Mo hydrogen partial pressure. See API Publica-
tion 941

800-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 800 Materials Considerations

Fig. 800-1 Common Exchanger Tube Materials (2 of 2) (This table is illustrative only. It is not suitable for materials
selection. See Section 811.)
Service Typical Materials Comments
Hydrogen-Sour Same as above; also 321 Stain- Materials choice depends on stainless steel
less Steel temperature and on hydrogen and H2S
partial pressure
Steam Carbon Steel CO2 corrosion may demand better mate-
rials in a condensing environment
(1) Inhibited grades only. Use ASTM B111 Grades C44300, C44400.

Tube Wall Thickness

Tube wall thickness is chosen for temperature-pressure and corrosion consider-
ations. Except at high pressures (usually over 1000 psi) where the strength of
thicker tubes is needed, anticipated corrosion rates will determine wall thickness.
The tube thicknesses given below fulfill 95% of typical requirements, although
thicker tubes may be used where high corrosion rates are expected. We try to stan-
dardize sizes to simplify maintaining a stock of materials for maintenance.
Carbon steel tubes: 14 gage minimum (13 gage is average) 12 or even 
10 gage is sometimes called for
Alloy tubes: 16 gage (except titanium)
Titanium tubes: 18 or 20 gage

Effect of Exchanger Design on Corrosion

Maximum Allowable Tube Velocity. In most common hydrocarbon services,
exchanger design sets the maximum velocities to accommodate conditions other
than corrosion, such as pressure drop. However, there are some services for which a
velocity beyond some critical threshold may initiate rapid corrosion. One example
of this is the “end impingement” failures of copper alloy tubes in sea water. Some
other examples are given in Figure 800-2.

Fig. 800-2 Some Typical Tube Velocity Limits (1 of 2)

Service Material Maximum Velocity
Sea Water (Avoid velocities below Admiralty 5 fps
3 fps)  
70—30 Cupro-Nickel 6 fps
 
Titanium Effectively no limit
Fresh Water All Materials No hard data available; seldom limits
exchanger design
Hydrocarbons All Materials Depends on service but seldom limits
exchanger design

December 1989  1989 Chevron USA Inc. All rights reserved. 800-3
800 Materials Considerations Heat Exchanger and Cooling Tower Manual

Fig. 800-2 Some Typical Tube Velocity Limits (2 of 2)

Service Material Maximum Velocity
Concentrated H2SO4 Carbon Steel 3 fps
Ammonium Bisulfide Solutions  Carbon Steel 20 fps in air-cooled exchangers;
(3% +); in wastewater treatment and  generally avoid water cooling
hydroprocessing plants  
Stainless Steel, 30 fps
Incoloy 800 
 
Titanium Effectively no limit; do not use 
titanium if high-pressure hydrogen is
also present

Effect of Shell-side Water on Tube Material. In cooling water service, tube life is
greatly affected by where the water is put. Cooling water on the shell side creates
corrosion problems that may be hard to overcome. For example, in well-treated
cooling water, carbon steel exchanger tubes have good life when the water is tube
side. However, it is virtually impossible to obtain good tube life on carbon steel with
water on the shell side, no matter how well the water is treated. If one tries to
compensate by upgrading tube material to Admiralty, for example, a galvanic corro-
sion problem is created where the alloy tubes join the carbon steel baffles. And if
the baffles are upgraded to brass, a new galvanic cell is made where brass baffles
touch the steel shell.
Shell-side water is also a poor choice for stainless steels. Local boiling may occur in
low flow areas, especially adjacent to tubesheets, with resultant concentration of
chlorides and stress corrosion cracking of the tubes.

Tube Quality and ASTM Specifications

While specifying the proper alloy to get the required corrosion resistance, tube
quality is ensured by ordering tubes to the appropriate ASTM specification. An
ASTM specification not only covers chemical composition, but also the method of
tube manufacture and quality control tests. ASTM specifications cover a variety of
end uses.
Some of the ASTM specifications commonly used for tubes are:
• A179 Seamless carbon steel
• A214 Welded carbon steel
• A199 Chrome-moly steel
• A249 Welded stainless steels
• A213 Seamless stainless steels
• B111 Copper alloys

Except for the carbon steel, all of these specifications cover a number of related
alloys. For example, B111 covers four kinds of Admiralty, several cupro-nickels,
aluminum brass, and aluminum bronze. When specifying materials, cite both the

800-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 800 Materials Considerations

ASTM specification and the grade. For example, a welded Type 304 stainless tube
would be specified as ASTM A240-TP304.

Welded vs. Seamless Tubes

In general, seamless alloy tubes are ordered instead of welded, but carbon steel
tubes may be either seamless or welded. With carbon steel, there is little or no sacri-
fice in life by using welded tubes, and the cost is much lower. However, welded
tubes should be purchased only from Company-approved suppliers who have a
proven track record on quality control.
One exception to the above generality is the use of welded titanium tubes; the cost
differential between welded and seamless here is significant, and tube manufac-
turers have demonstrated their ability to produce defect-free tubing.

812 Tubesheets
Materials
Tubesheets are usually made of the same material as the tubes. One major excep-
tion is with copper alloys. General practice is to use naval-rolled brass (NRB)
tubesheets with Admiralty tubes and Monel or 70-30 cupro-nickel tubesheets with
70-30 cupro-nickel tubes.

Cladding
When constructing an exchanger using alloy tubes in which the corrosive fluid is on
the tubeside, it may be economical to use alloy-clad rather than solid alloy
tubesheets. If clad tubesheets are used, cladding thickness should be 0.5 inch, so that
the first serration is entirely within the cladding. When the tubes are rolled in place,
this will allow an alloy-to-alloy seal at the first serration. Such a seal prevents corro-
sive fluid from entering the crevice between tube and tubesheet to cause galvanic
corrosion where alloy and carbon steel are in contact. See Section 520 and Specifi-
cation EXH-MS-2583 for more information on cladding.

Galvanic Attack
In places where tubesheet and tube materials differ take the following precautions:
• Consider differential thermal expansion that can loosen rolled joints.
• Consider the galvanic relationship between tube and tubesheet when handling a
corrosive aqueous (i.e., electrically conductive) fluid such as sea water. Do not
use materials that are far apart in the galvanic series, and especially do not have
the tubesheet be the more noble metal. (See the Corrosion Prevention and
Metallurgy Manual, Section 200.)
• Using the wrong materials combination may result in accelerated corrosion or
end impingement. For example, in sea water service use of a Monel rather than
naval brass tubesheet with Admiralty tubes can reduce tube life by an order of
magnitude.

December 1989  1989 Chevron USA Inc. All rights reserved. 800-5
800 Materials Considerations Heat Exchanger and Cooling Tower Manual

• Know that there are many fluids in which the probability of galvanic attack
may not be obvious. The answer here is to search for relevant experience with
that fluid.
Note that galvanic attack is not a problem in hydrocarbons and usually is not severe
in fresh water. Galvanic attack can sometimes be prevented through the use of sacri-
ficial anodes.

Tube Rolling
There is no problem in rolling soft tubes into a hard tubesheet. But rolling hard
tubes into a soft tubesheet can result in the enlargement of tubesheet holes, without
the joint becoming tight. See Section 520 for information on the tube-to-tubesheet
joint.

813 Baffles
Baffles are usually made of the same material as the exchanger shell, carbon steel
being the most common. The most important consideration in choosing baffle mate-
rial is corrosion resistance; it is poor economy to have to rebuild an exchanger
because the baffles have corroded while the tubes are still in good condition. Baffles
should be designed to last at least as long as the tubes. One refinery had to replace
four expensive titanium bundles after 4 years because the steel carcass had
corroded.
Seal strips on longitudinal baffles (so-called “lamiflex baffles”) have some special
problems. They are very thin and have little tolerance for corrosion. In addition,
bending stresses render them susceptible to stress corrosion cracking in certain
services. The 300 Series stainless steels are the most common seal strip materials,
but special alloys are required for some services such as hydroprocessing plants.
Seek the advice of corrosion or materials engineers before choosing seal strip mate-
rials for new services.

814 Shell
An exchanger shell is nothing more than a pressure vessel and is designed according
to the same criteria (typically ASME Code, Section VIII, Division 1). Materials
suitable for pressure vessels are also acceptable for exchanger shells. One important
materials limitation is that it is seldom practical to use more than 1/4-inch corrosion
allowance on an exchanger shell. If corrosion is deeper than this, by-passing around
the baffles will cause a major degradation in exchanger performance.

820 Minimum Pressurizing Temperature

Minimum pressurizing temperature (minimum design metal temperature) is a crit-
ical design factor for pressure vessels. Exchanger shells and channels are pressure
vessels, and must be designed accordingly. This subject is covered in detail in the
Pressure Vessel Manual.

800-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 800 Materials Considerations

In brief, we establish a minimum pressurizing temperature to avoid a catastrophic

brittle fracture. Ordinary carbon steels, for example, become brittle at low tempera-
tures. The ductile-to-brittle transition temperature may range from well above
ambient to well below ambient, depending on the grade and thickness of steel used.
A material must be chosen that will not suffer brittle fracture under the conditions in
which an exchanger is expected to operate. This includes hydro-test, which must be
done at a temperature above the minimum pressurizing temperature.

830 Sacrificial Anodes

The purpose of sacrificial anodes is to extend the life of critical heat exchanger parts
by the application of cathodic protection. The subject is discussed in detail in
Section 1600 of the Corrosion Prevention and Metallurgy Manual.
The Company does not often use cathodic protection for heat exchangers. Although
sacrificial anodes can be installed in exchanger channels or water boxes to protect
tubesheets, tube ends, and the channel section itself, anode life usually is not long
enough to justify their installation. While such anodes can protect tube ends from
corrosion (as long as the anodes last) protection generally does not extend more than
one or two tube diameters down the inside of the tube.
One application that did prove to be reasonably successful was the installation of
carbon steel sacrificial anodes in water boxes in the Borco Refinery sea water desal-
ination plant. This was to protect the 90-10 cupro-nickel tubesheets from galvanic
corrosion caused by contact with the titanium tubes.
Another successful anode installation was the use of aluminum anodes in exchanger
channels at Richmond. They prevented galvanic attack of Monel tubesheets caused
by dissimilar metal contact with titanium tubes in a sea water environment.

840 Insulation
Heat exchangers are basically pressure vessels and are insulated as such. This
subject is covered in the Insulation and Refractory Manual.
The large flanged connections on heat exchangers cause the major problem with
insulation. For more information on the criteria for insulating large flanges and the
design of the flanges and bolting, see Section 550 of this manual.

Removable Insulation
Removable covers can be removed for exchanger maintenance, or to look for flange
leaks after startup, and then reinstalled after inspection. Large covers, however, are
hard to handle, particularly those used on very hot equipment. For design informa-
tion on removable covers, see the Insulation and Refractory Manual, Section 100,
and IRM-SU-4197.

December 1989  1989 Chevron USA Inc. All rights reserved. 800-7
800-8

800 Materials Considerations

Fig. 800-3 Commonly Used Materials for Components of Shell-and-Tube Heat Exchangers (1 of 2)
Low Alloy Steels High Alloy Steels

18Cr-8Ni
Components Carbon Steel C-1/2Mo 1-1/4Cr-1/2Mo 2-1/4Cr-1/2Mo 5Cr-1/2Mo 12 Cr 18Cr-8Ni-3Mo 18Cr-8Ni Stabilized

PLATES:         
 SA-285-C SA-204-A SA-387-11 SA-387-22 SA-387-5 Do not use 12Cr SA-240-TP316 SA-240-TP304 SA-240-TP321
(For rolled and SA-515 and 516 SA-204-B or Class 1 or 2 Class 1 or 2 Class 1 or 2 for pressure or or or
welded shells, (All grades) SA-204-C (Tubesheets containing parts SA-240-TP316L SA-240-TP304L SA-240-TP347
shell covers, and baffles only. (except tubes).
channels and Use 12Cr clad Use 5Cr-1/2Mo
nozzle necks, for shells and tubesheets with
heads flat channels.) 12Cr tubes.
 1989 Chevron USA Inc. All rights reserved.

covers,
tubesheets and
baffles)

PIPE:         
 SA-106-B or SA-335-P1 SA-335-P11 SA-335-P22 Not used Do not use SA-312-TP316 SA-312-TP304 SA-312-TP321
(For pipe sized SA-53-B or or or
shells and SA-312-TP316L SA-312-TP304L SA-312-TP347
nozzle necks) (Seamless or (Seamless or (Seamless or
welded) welded) welded)

FORGINGS:         
 SA-105 or SA-182-F1 SA-182-F11 SA-182-F22 SA-182-F5 or Do not use SA-182-F316 SA-182-F304 SA-182-F321
(For body and SA-181 SA-182-F5a or or or
nozzle flanges, (Class 60 or 70) (For forged SA-182-F316L SA-182-F304L SA-182-F347
blind flanges, tubesheets and

Heat Exchanger and Cooling Tower Manual

couplings and covers only)
forged flat
covers and
tubesheets)

TUBES: WELDED:      WELDED: WELDED: WELDED:

SA-214 SA-209-T1 SA-199-T11 SA-199-T22 SA-199-T5 SA-268-TP405 SA-249-TP316 SA-249-TP304 SA-249-TP321
SEAMLESS: or or or or
SA-179 SA-258-TP410 SA-249-TP316L SA-249-TP304L SA-249-TP347
(Seamless or SEAMLESS: SEAMLESS: SEAMLESS:
welded) SA-213-TP316 SA-213-TP304 SA-213-TP321
or or or
December 1989

SA-213-TP316L SA-213-TP304L SA-213-TP347

 Fig. 800-3 Commonly Used Materials for Components of Shell-and-Tube Heat Exchangers (2 of 2)
December 1989

Heat Exchanger and Cooling Tower Manual

Low Alloy Steels High Alloy Steels
18Cr-8Ni
Components Carbon Steel C-1/2Mo 1-1/4Cr-1/2Mo 2-1/4Cr-1/2Mo 5Cr-1/2Mo 12 Cr 18Cr-8Ni-3Mo 18Cr-8Ni Stabilized

BOLTS:         
SA-193-B7 SA-193-B7 SA-193-B7 or SA-193-B7 or SA-193-B5 Do not use SA-193-B8M SA-193-B8 SA-193-B8T
SA-193-B16 SA-194-B16 or
SA-193-B8C
Caution: For hydrogen service,
verify that Cr and Mo content are
high enough to resist H2 attack.

NUTS:         
 1989 Chevron USA Inc. All rights reserved.

SA-194-2H SA-194-2H SA-194-2H SA-194-2H SA-194-3 Do not use SA-194-8M SA-194-8 SA-194-8T
SA-194-8C

800 Materials Considerations

800-9
800 Materials Considerations Heat Exchanger and Cooling Tower Manual

800-10  1989 Chevron USA Inc. All rights reserved. December 1989
900 Inspection and Testing

Abstract
This section covers both shop inspection of new equipment and inspection of
existing units during maintenance turnarounds. The section on shop inspection
covers information on shop testing and inspection to verify that new units are built
to specification and with good workmanship.
Heat exchanger failures cause many plant shutdowns and slowdowns. The cost of
these failures is high due to lost production and replacement of the exchangers.
Inspection’s responsibilities are to: (1) inspect existing exchangers when they are
taken out of service to determine extent of repairs and remaining life; (2) verify that
repairs are completed to the tolerances specified; and (3) maintain accurate long
term records that facilitate accurate forecasting.
This section covers inspection and testing requirements for new and existing shell
and tube and air cooled heat exchangers. The type of damage found in exchangers is
discussed, along with the primary inspection and testing techniques used to deter-
mine the condition of an exchanger. Detailed shutdown inspection checklists are
included. Acceptable dimensional tolerances are discussed, and guidance on Safety
Instruction Sheets (SIS) for shell and tube exchangers and air cooled heat
exchangers is provided.

Contents Page
910 Shop Inspection of New Exchangers 900-3
911 General Comments
912 Degree of Inspection
913 Shop Inspection of Shell and Tube Exchangers
914 Shop Inspection of Air Cooled Heat Exchangers
915 Inspection Tasks
920 Inspection of Existing Shell and Tube Exchangers and 
Air Cooled Exchangers 900-9
921 The Inspection, Engineering, Operation, and Maintenance Team
922 Inspection Sequence
923 Double Pipe Exchangers
924 Inspection of State Registered Unfired Steam Generators

December 1989  1989 Chevron USA Inc. All rights reserved. 900-1
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

930 Types of Damage Found In Exchangers 900-10

940 Inspection and Testing Techniques 900-12
941 Inspection Techniques
942 Pressure Testing Techniques
943 Safety
950 Dimensional Tolerances 900-14
960 Safety Instruction Sheets 900-15
970 Maintenance Inspection Checklists 900-15

900-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

910 Shop Inspection of New Exchangers

911 General Comments

Shop inspection is warranted for virtually all heat exchangers except “off-the-shelf”
lube oil coolers. Inspection at the fabrication plant is normally done by the Quality
Assurance (QA) section of Purchasing, or by an inspector contracted by Purchasing
QA to inspect the exchanger. The purpose of shop inspection is to provide assur-
ance that the equipment meets specification and order requirements, displays good
workmanship, and is free of significant damage or defects.
ASME Code uses the term “Authorized Inspector” to describe the inspector who
works for a state, municipality, or insurance company and must be used by the fabri-
cator to inspect the exchanger and sign the Manufacturer’s Data Report. The Autho-
rized Inspector is not the same shop inspector that Chevron uses. The Authorized
Inspector’s responsibility is to verify that the exchanger is designed and fabricated
in accordance with ASME Code, but he has no interest in and does not check such
things as:
1. Dimensions or orientations (except for diameter and thickness).
2. Presence of all required nozzles, baffles, or impingement plates; plumb-
ness/squareness of nozzles.
3. Limitations or restrictions in Company specifications on materials or welding
processes.
4. Special construction details in Company specifications that exceed ASME
Code requirements.
5. Inspection or testing requirements in Company specifications that exceed
ASME Code minimums.

912 Degree of Inspection

Figures 900-1 and 900-2 show the degrees of inspection for shell and tube heat
exchangers and for air cooled exchangers, respectively. The degrees of inspection
ranging from one or two inspection visits to resident inspection. The visits them-
selves are listed in chronological sequence. Degrees of inspection are listed gener-
ally in accordance with the importance of the visit.
Purchasing QA usually establishes the visits required based on the specifics of an
order, but the job engineer can always participate in this decision. Figures 900-1 and
900-2 define the basic purpose for each visit but do not list all of the tasks that the
Company inspector is required to complete during these visits (see Section 915).

December 1989  1989 Chevron USA Inc. All rights reserved. 900-3
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-1 Degrees of Inspection for Shell and Tube Heat Exchangers

Visit Notes:
(1) This visit is required for any parts of exchanger that are required to be PWHT’D. This does not apply to stress relief of
tube bends.
(2) This visit is made prior to any hydrostatic test and after PWHT if any. For exchangers with fixed tubesheets, internal
inspection of the shell prior to attachment of closing tubesheet is required.
(3) This visit is required for all stacked exchangers with interconnecting nozzles.
(4) This visit is required for Degree 9, 10, and 11 inspection if finish paint is shop applied.

General Notes:
(1) Add supplementary visits A, C, and D as applicable.
(2) Visits for purposes other than listed may be warranted and added (for example, to witness penetrant test of seal-
welded tubes/tubesheets).
(3) This matrix does not include all of the inspection tasks that an inspector completes during the required plant visits
shown. A list of detailed inspection tasks can be several pages long. This list is normally part of the inspection plan for
a particular exchanger. Inspection plans are usually prepared by Purchasing QA.
(4) Each visit is not necessarily a full 8-hour day.

Problems with quality, specification compliance, or vendor cooperation/scheduling could increase the number of visits beyond
the number shown.

900-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Fig. 900-2 Degrees of Inspection for Air Cooled Exchangers

Visit Notes:
(1) Outside face of tubesheet must be painted prior to tubing bundle if painting outside face is important; not normally
done.
(2) Test erection and shop run are normally required. This visit can usually be included with visit for final inspection.

General Notes:
(1) Add supplementary visits A and B as applicable.
(2) Visits for purposes other than listed may be warranted and added.
(3) This matrix does not include all of the inspection tasks that an inspector completes during the required plant visits that
are shown. A list of detailed inspection tasks can be several pages long. This list is normally part of the inspection plan
for a particular exchanger. Inspection plans are usually prepared by Purchasing QA.
(4) Each visit is not necessarily a full 8-hour day.
(5) Problems with quality, specification compliance, or vendor cooperation/scheduling could increase the number of visits
beyond the number shown.

December 1989  1989 Chevron USA Inc. All rights reserved. 900-5
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

913 Shop Inspection of Shell and Tube Exchangers

Figure 900-1 identifies the different degrees of inspection for shell and tube
exchangers. Figure 900-3 summarizes when to use the various degrees. Degree 4 is
about the minimum for any shell and tube exchanger in hydrocarbon or critical
service. Degree 10 is appropriate where any of the following apply.
• The exchanger has a fabrication time over 6 weeks.
• The exchanger is very large (over 4 feet diameter).
• The exchanger wall thickness is over 1-1/4 inch.
• The exchanger has a high pressure rating (flanges 600 pounds or more).
• The exchanger is fabricated from low alloy or high alloy steels.
• The exchanger has substantial cladding or weld overlay.
Degree 11 (resident inspection) is rarely applied to heat exchangers but could be
considered during fabrication of heavy wall shells or channels (over about 4 inches
thick).
Visits for purposes in addition to those listed can be added to cover any special
concerns.
Fig. 900-3 Application of the Degrees of Inspection for Shell and Tube Exchangers (1 of 3)
GUIDELINES FOR CHOOSING THE APPROPRIATE DEGREE OF INSPECTION
FOR SHELL AND TUBE HEAT EXCHANGERS
(NOTE THAT THESE ARE GUIDELINES ONLY)
Degree 1 when:
1. No Company specification applies, and
2. Exchanger:
a. is small and simple (under 18 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has low pressure rating (under 250 psig), and
d. is not in flammable or hazardous services.
Degree 2 when:
1. No Company specification applies, and
2. Exchanger:
a. is of moderate size (18 inch to 30 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has low pressure rating (under 250 psig).
OR
3. Only one low dollar value hair-pin type exchanger is on a purchase order (an attempt is made to accomplish
Degree 2 inspection in one visit to include inspecting parts, witnessing assembly, and witnessing hydrostatic
tests).

900-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Fig. 900-3 Application of the Degrees of Inspection for Shell and Tube Exchangers (2 of 3)
Degree 3 when:
1. Company specifications and/or standard drawings apply, and
2. Exchanger:
a. is small and simple (under 18 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has low pressure rating (under 250 psig), and
d. is in relatively noncritical service (choose at least Degree 4 for critical services).

Note “Critical Service” is difficult to define; judgment is required; factors to consider are type of
service (contents of exchanger), consequences of failure with regard to the process, consequences of
failure with regard to personnel and surrounding facilities.
Degree 4 (this is the normal minimum for exchangers in hydrocarbon service) when:
1. Company specifications and/or standard drawings apply, and
2. Exchanger:
a. is of moderate size (18 inch to 30 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has moderate pressure rating (150 pound or 300 pound flanges), and
d. is in relatively critical service.
Degree 5 when:
1. Company specifications and/or standard drawings apply, and
2. Exchanger:
a. is of large size (over 30 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has moderate pressure rating (150 pound or 300 pound flanges), and
d. is in relatively critical service
Degree 6 when:
1. Company specifications and/or standard drawings apply, and
2. Exchanger:
a. is of large size (over 30 inch diameter), and
b. is carbon steel (tubes can be another material), and
c. has shell thickness 3/4 inch or more, and
d. has moderate pressure rating (150 pound or 300 pound flanges), and
e. is in relatively critical service.
Degree 7 (same as Degree 6 but when diameter, thickness of tubesheet, or tubesheet material warrants a visit for
tubesheet/baffle inspection before installing tubes; this visit should always be made for tubesheets that are clad,
weld overlayed, or will have tube/tubesheet welds, and can be considered for all moderate size and large size
process exchangers.)
Degree 8 (same as Degree 7 but when surface preparation required for painting is “near-white” or “White Metal.”)
Degree 9 (same as Degree 7 but when surface preparation required for painting is “near-white” or “White Metal” and
inspection of dry primer is judged to be warranted.)

December 1989  1989 Chevron USA Inc. All rights reserved. 900-7
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-3 Application of the Degrees of Inspection for Shell and Tube Exchangers (3 of 3)
Degree 10 when:
1. Exchanger:
a. fabrication time is over 6 weeks, or
b. is very large (over 4 feet diameter), or
c. has wall thickness over 1-1/4 inch, or
d. has a high pressure rating (flanges 600 pound or more), or
e. is fabricated from low alloy or high alloy steels.
f. is fabricated from clad or weld overlayed material.
Degree 11 (Resident inspection is rarely applied to heat exchangers but could be considered during fabrication of
heavy wall shells or channels; heavy wall means over 4 inches thick.)

914 Shop Inspection of Air Cooled Heat Exchangers

Figure 900-2 identifies the degree of inspection for air cooled heat exchangers.
Degree 5 is about the minimum for any air cooled exchanger in hydrocarbon or
critical service. Factors similar to those listed in Section 913 above are used to
decide when more inspection visits are warranted; consult Purchasing QA for addi-
tional guidance.

915 Inspection Tasks

Figures 900-1 and 900-2 do not list all of the tasks the Company inspector is
required to complete during the plant visits. These tasks are normally listed in an
inspection plan prepared by Purchasing QA. The tasks listed in the inspection plan
can be several pages long. Some of the principal tasks are:
1. Verification that welding procedures and welders are qualified per ASME
Code.
2. Review of material test reports for principal exchanger parts and verification
that exchanger parts are traceable to material test reports.
3. Verification that welding procedures and preheat requirements are being
followed.
4. Visual inspection of all welds for flaws, contour, size, and reinforcement.
5. Reading all radiographs (“x-rays”) and witnessing other nondestructive exami-
nation, if required.
6. Complete dimensional and orientation check against Company reviewed
vendor drawings.
7. Verification that all nozzles are present, and that they are correct size, rating,
and material.
8. Verification that bundle has all required baffles and support plates, and that
baffle and tubesheet holes have been deburred; verification that tubes have been
properly expanded into tubesheet holes.
9. Review of Manufacturer’s Data Report to verify all entries are correct.

900-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

920 Inspection of Existing Shell and Tube Exchangers and Air Cooled
Exchangers

921 The Inspection, Engineering, Operation, and Maintenance Team

While much of the standard maintenance work on exchangers is initiated by inspec-
tion, the major decisions must be made with input from inspection, engineering,
operations and maintenance. These decisions include:
• Retubing or rebundling
• Plugging tubes
• Nozzle replacement
• Major shell repairs or replacement
• Gasket surface repairs
The above organizations also must work together to establish the objectives of
exchanger repairs well in advance of the shutdown. These objectives include:
• Exchanger or bundle design life
• Time between internal inspections (for some cases, this is determined by state
regulations; i.e., waste heat boilers)
• Degree of internal or external leakage which is acceptable over the life of the
bundle
• Degree of cleaning required (often dictated by process economics)
Section 1000 contains a Heat Exchanger Bundle Repair Worksheet for helping the
engineer coordinate the decision making process.

922 Inspection Sequence

Inspection of existing exchangers and air coolers involves the following steps:
1. Set objectives as discussed above.
2. Examine inspection and maintenance records to note any trends and make
initial recommendations.
3. Perform preshutdown inspection to note leaks and any areas which must be
inspected in more detail when the unit is shut down.
4. Determine need for bleeder tests prior to opening.
5. Inspect and photograph shell and bundle after exchanger has been opened but
before cleaning to look for patterns and types of fouling and evidence of
plugged tubes.
6. Inspect shell and bundle after cleaning and make recommendations for repairs.
7. Inspect repairs and witness testing.

December 1989  1989 Chevron USA Inc. All rights reserved. 900-9
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

8. Record gagings and repairs in the inspection file to anticipate what to expect in
the future.
9. Process paperwork for any code repairs or operating permits for state regis-
tered equipment.
Section 970 contains a detailed inspection checklist for exchangers and air coolers.

923 Double Pipe Exchangers

Double pipe or “hairpin” exchangers can be one pipe or several pipes inside another
pipe. In either type of exchanger the corrosion is usually found in the area of the
inlet nozzle and at the return bend.
All these units are usually small enough in diameter to radiograph. Radiography can
reveal the inner and outer pipe wall thickness, tube-wall thickness, and any fouling
problem, without taking the unit out of service. Most repair work can therefore be
planned before the plant is actually shut down.
In one double pipe exchanger, called a “chiller,” the stock in the inner pipe is
scraped using a scroll, and corrosion is usually found externally because of low
temperature conditions. The scrolls may also break or become disconnected.

924 Inspection of State Registered Unfired Steam Generators

Like fired boilers, unfired steam generators require periodic inspection and state
certification by qualified boiler inspectors. Generally, unfired steam generators are
of tubular heat exchanger design. Consequently, inspection and repair follow the
same general pattern as that used for process heat exchangers.
1. The tube bundles are removed, cleaned, and inspected externally and internally.
2. The shells and flash drums are internally inspected and ultrasonically gaged.
3. The gage glasses, pressure gages and alarms are cleaned and serviced. The
safety valves are removed, cleaned, tested and reset.
4. Hydrostatic tests, when applied, should include the flash drum, if any, and all
interconnected piping and appurtenances.
The work copies of the inspection report and operating permit sheet are completed
and processed as required by the local inspection organization.

930 Types of Damage Found In Exchangers

The various types of damage found in exchangers are discussed below. Refer to the
Corrosion Prevention Manual for more information.
End impingement—where high velocity stock enters the tube and erodes the end.
This can involve the tubesheet as well.

900-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Galvanic corrosion—occurs when the tube material and tubesheet material are
apart on the galvanic scale, often with salt water as the electrolyte; e.g., for a Monel
tubesheet with titanium tubes, the Monel tubesheet will corrode.
Lodgement attack—occurs when an obstruction in the tube increases velocities at
that point and erodes the tube.
External corrosion—wall thinning on the outside of the tube.
Internal corrosion—wall thinning on the inside of the tube.
Dezincification—usually occurs on the I.D. of Admiralty brass tubes in saltwater
service operating above 120°F or when an inhibited grade of Admiralty was not
used.
Denickelification—usually occurs on the I.D. of copper-nickel tubes in saltwater
service which have operated well above their normal tube-wall temperature.
CO2 attack—occurs in condensing steam service and is common at the floating
tubesheet end of the tube O.D. on vertical reboilers.
Stress corrosion cracking—usually occurs at a stressed point of tubes, such as at
the U-bends. Can occur on any metal. Chlorides with stainless steels or NH3 with
brass are examples where this can be a problem.
Mechanical deformation—caused by improper sling placement when handling the
bundle, for example.
Vibration wear—occurs when the area between the tube and transverse baffle is
too great or when the baffle spacing is too great, allowing the tube to vibrate in the
hole.
Seal weld cracking—cracks in the tube-to-tubesheet seal weld. (Usually detected
during a hydrotest.)
Roll leaks—when the tube-to-tubesheet roll joint leaks. (Usually detected during a
hydrotest.)
Thermal shock—usually occurs during an operation upset. Improper starting up or
shutting down of equipment or loss of cooling water can cause shock. This
commonly causes roll leaks.
External impingement—usually occurs on the tube O.D. in the area of the shell
inlet nozzle. Can be remedied by installing an impingement device on the bundle,
usually impingement rods or tubes.
Necking down—occurs on the tube O.D. at the transverse baffle. It is caused by too
large an area between the tube O.D. and the transverse baffle, allowing bypassing.
Usually the baffle will erode as well as the tube, requiring replacement at the next
retubing.

December 1989  1989 Chevron USA Inc. All rights reserved. 900-11
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

940 Inspection and Testing Techniques

This section discusses the primary inspection and testing techniques used to deter-
mine the condition of an exchanger before, during, and after repairs. It does not
include visual inspection, which is very important.

941 Inspection Techniques

• Ultrasonic (UT) gaging can be used to determine remaining wall thickness of
the shell as well as defects in the wall.
• Micrometer measurements are used to determine tube wall thickness, depth of
pitting, and metal loss on body flanges or tubesheets.
• Radiography can be used to check wall thickness and fouling in tubes. Radiog-
raphy can be used on large bundles by sliding the film under the outer row of
tubes. On large U-tube bundles the U-bends can be radiographed for wall thick-
ness and fouling.
On bundles with tubes 2-inch O.D. and larger, tube wall thickness and O.D.
fouling can be identified using radiography by sliding a narrow film cassette in
one tube and the source in another tube. Place the film in Tube A as shown in
Figure 900-4. Put the source in Tube D. The two walls projected on the film
will be the inside walls of Tubes B and C. Any fouling between Tubes B and C
will show on the film. Several films can be shot with one exposure, depending
on the tube layout and space available.

Fig. 900-4 Film Layout for X-Raying Bundles with Tube 2 inches and Larger

• Eddy current testing is an effective inspection tool for determining tube wall
thickness. It is very useful for determining the tube wall in the center of tube
bundles when specimen tubes are unattainable, such as in U-tube bundles.
Ferrous as well as nonferrous tubes can be inspected; however, the tubes must
be clean, and specimen tubes to confirm eddy current results are advisable
before major repairs.
• Specimen tubes are tubes which have been removed from various locations
throughout the bundle to determine the extent of corrosion in the bundle.

900-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Choosing the proper location of the specimen tubes is critical in developing an

accurate profile of the corrosion. Often, eddy current is used to determine the
worst areas of corrosion before pulling specimen tubes.
• Boroscopes connected to a TV can be used to visually inspect the tube interior.
This technique will not measure wall thickness or pitting depth but can be a
good visual tool for determining tube condition.

942 Pressure Testing Techniques

In general, exchangers out of service for repairs are pressure tested before the
repairs to determine where work is required and after repairs to determine if the
exchanger meets code requirements.

Test Pressures
The code hydrostatic test pressures are calculated using the code under which the
exchanger was built (usually ASME). The exchanger fabricator generally supplies
these pressures. The exchanger should be tested to full code test pressure after any
repair work affecting the strength of the vessel.

Minimum Pressurizing Temperature (MPT)

The temperature of the pressurizing fluid must be above the MPT as discussed in
Section 820 of this manual and in the Pressure Vessel Manual.

Tube Side Tests

Pressuring the tubes permits inspection of the channel cover gasket, channel,
channel-to-tubesheet gasket, tubesheets, floating head gasket, floating head, and the
tubes. Leaks can be identified by leakage at the joints; however, individual tube
leaks are not identifiable because the tubesheets are inaccessible.

Shell Side Tests

Pressuring the shell side tests the shell, shell-to-tubesheet gasket, channel-to-
tubesheet gasket, tubesheets, and tubes. If the channel has a removable cover this
test allows individual leaking tubes to be plugged off because the tubesheet is acces-
sible. Roll leaks also can be individually identified and repaired. U-tube bundles are
easily tested with the shell enclosed. Floating head bundles require the use of a test
ring to observe the floating tubesheet during the shell test. Figure 900-5 is a sche-
matic of a test ring.

Leakage Tests
Tests for leakage can be conducted at the operating pressure for repairs not affecting
the vessel “strength,” such as plugging tubes.

Bleeder Tests
Shell and tube bleeder tests can be applied quickly; however, they will indicate only
that the unit is leaking. The type of leak will be unknown. When the shell or tubes
are pressured, inspect for leaks at the channel or shell drains.

December 1989  1989 Chevron USA Inc. All rights reserved. 900-13
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-5 Test Ring - Floating Tubesheet (Courtesy of TEMA)

Testing Medium
The testing medium should be oil or water. Gas tests are more sensitive, but are
extremely dangerous. Such tests over 15 psig require the approval of local manage-
ment.

943 Safety
The inspector must remember that pressure tests involve certain hazards. In addi-
tion to gaskets blowing out, previously driven plugs can blow out as well. While
witnessing pressure tests, the inspector should stand to one side. Unauthorized
personnel should not be in the vicinity of the exchanger on test. The test pump must
have a safety valve set at the test pressure, and a gage with a range of roughly
double the test pressure (i.e., if test pressure is 150 psi, gage range should be 0 to
300 psi).

950 Dimensional Tolerances

Dimensional tolerances for standard shell and tube exchangers are tabulated in
Section 2 of the Standards of Tubular Exchanger Manufactures Association
(TEMA).
Copies of TEMA Tables F-1 and F-2 are included for your quick reference.
(See Figures 900-6 and 900-7.)
Dimensional tolerances for air cooled exchangers are covered by API Standard 661,
also included in this manual.

900-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

960 Safety Instruction Sheets

Safety Instruction Sheets (SISs) are used by several organizations as a single-
source summary document for important data on operating equipment. These data
are important to operations, technical groups, inspection and maintenance, and they
include:
• Safe operating limits (temperatures, pressures, rotational speeds, etc.)
• Particular precautions for safe operation
• Specified routine tests required for safe operation
• References necessary to operate and/or maintain the equipment
• Inspection/technical data (minimum thickness, corrosion allowances, test pres-
sures, relief valve settings, minimum pressurizing temperatures, etc.)
SI sheets should be completed for (1) new exchangers and (2) existing exchangers
that have been altered in any way that affects the SI sheet content. This includes:
material changes, derating, minimum thickness changes.
SI sheets are routinely filled out by a project engineer and countersigned by an oper-
ating representative. They become permanent records in the inspection groups and
in the equipment manuals for the plants. Many technical groups also maintain
compiled copies. Some organizations are now shifting this information to a central-
ized information database, PEIS (Plant Equipment Information System). Regardless
of the format, the SI sheet is a good checklist of important information to have
readily available.
A significant problem with SI sheets has been getting them accurately and
completely filled in prior to startup. One reason for this is that vendors are not
always as responsive during the latter stages of a project, after the equipment has
been purchased and installed. If you are committed to having SI sheets, you should
consider methods to motivate vendors to complete them (or provide data) after all
design and operating data are confirmed.
Instructions for completing the following SI sheets are included in the Safety
Instruction Sheet Section:
• EXH-EF-787 Shell and Tube and Double Pipe Heat Exchangers SI Sheet
• EXH-EF-788 Air Cooled Heat Exchangers SI Sheet

970 Maintenance Inspection Checklists

Attached are inspection checklists for the following:
• Shell and Tube Exchangers (Figure 900-8)
• Air Cooled Exchangers (Figure 900-9)
This information was excerpted from the CUSA Manufacturing Inspection Manual.

December 1989  1989 Chevron USA Inc. All rights reserved. 900-15
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-6 Standard Tolerances for External Dimensions, Nozzle and Support Locations of Heat Exchanger (Courtesy
of TEMA)

900-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Fig. 900-7 Standard Clearances and Tolerances for Tubesheets, Partitions, Covers, and Flanges (Courtesy of TEMA)

December 1989  1989 Chevron USA Inc. All rights reserved. 900-17
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-8 Inspection Checklist—Shell and Tube Exchangers (1 of 3)

Date: ____________________
Exchanger: ____________________
Service: Tube side __________________
Shell side __________________
Initial: _______

Preshutdown Inspection:
Discuss and agree on the “shutdown objectives with Operations, Engineering, and Maintenance.” _______
Records Check:
Review records and become familiar with corrosion history. Prepare a sketch indicating type and loca- _______
tion of previous corrosion.
External Visual Inspection:
Inspect exchanger supports for proper bearing and anchoring. _______
Look for corrosion on the shell where it is in contact with the concrete fireproofing. _______
Check slide plates or slotted bolt holes for freedom of movement at the exchanger supports. _______
Inspect support leg fireproofing for cracks and spalling and corrosion underneath. _______
Inspect the anchor bolts for necking down at the baseplate. _______
Where exchangers are saddled into a concrete support, inspect the shell for external corrosion. _______
Inspect the insulation and retainers. Corrosion often occurs where moisture penetrates the insulation _______
and the shell is not hot enough to dry it out. Insulated exchangers which operate below ambient
temperature are susceptible to this type of corrosion. Representative sections, or even all of the insula-
tion, should be removed to permit a complete external inspection and ultrasonic gaging.
Inspect all small piping for leaks and external corrosion, including vents, drains, relief valves, plugs, _______
and other small screwed fittings.
Look for gasket leaks on the inlet and outlet piping flanges and at the channel or shell cover flanges. _______
Evidence of leaks can be seen as deposits on the ground or platform under the unit.
Check the packing gland on the bottom leg of the reboilers for leaks. _______
Check gaskets and gasket surfaces for evidence of leakage. _______
If the unit is a water cooler or condenser, note the water outlet temperature either by dial thermometer _______
or by feel. Excessively high water temperature (above 120°F for salt water) indicates possible fouling
and probable high corrosion rates in the tubes.

Ultrasonic Inspection:
Take ultrasonic gage measurements at locations where internal corrosion is suspected. A sufficient _______
number of measurements should be taken to provide reasonable assurance of the exchanger’s integ-
rity. It is important to recognize that external gage points can reflect general corrosion rates, but that
actual remaining wall thickness is probably lower.

Shutdown Inspection:
Shell and Covers:
Verify exchanger is properly cleaned for inspection. Cleaning can be done by water-washing, chemical _______
cleaning, or by sandblasting. The amount of cleaning required depends on the severity of the corro-
sion and fouling in the unit.

900-18  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Fig. 900-8 Inspection Checklist—Shell and Tube Exchangers (2 of 3)

Internally inspect shells 20 inches or greater in diameter. Look for evidence of corrosion, noting particu- _______
larly the nozzle, flange and shell welds, and the locations where the transverse baffles contact the
shell.
Take ultrasonic (UT) gage measurements in corroded areas and at representative locations on the _______
shell cover and channel section to establish an accurate picture of the corrosion patterns. UT the shell
at the inlet and outlet nozzles.
Prior to cleaning, check gasket surfaces for evidence of leakage. _______
After buffing, inspect gasket surfaces and flanges for evidence of corrosion or mechanical damage. _______
Severely corroded gasket surfaces can require weld buildup and machining. Lightly damaged gasket
surfaces can be repaired by machining or drawfiling.
Inspect the baffles in channel sections for warpage, distortion, corrosion, or evidence of bypassing. _______
Ensure that baffle stay rods are not cracked or broken. _______
Inspect lifting lugs on channel sections and covers for mechanical damage or corrosion to ensure they _______
are safe for continued use.
Hammer-test or UT gage all small piping on the channel, shell, or shell cover and inspect all threaded _______
fittings as necessary. On high pressure, or stainless steel piping, radiography should be used instead
of hammer-testing.
For services other than water, verify that all piping connections, including small piping, are welded up _______
to the first root value. Eliminate connections less than 3/4 inch in size.
Inspect longitudinal grooves or guides in the shell for the horizontal baffle in the bundle. _______
Inspect alloy lining for cracked attachment welds and pinhole leaks. Have leaking or bulged lining _______
removed for inspection under the lining.
If channel cover staybolts are used, observe the bolt holes behind the lining. Graphitization frequently _______
enlarges the holes and results in leaks. Repairs can be made by using oversize gaskets under the
nuts.
Inspect floating head bolts for evidence of corrosion, breakage or overstressing. _______
Inspect split rings for warpage which can cause leaks at the floating head gasket. _______
Bundles:
A summary of the various types of bundle wear is included in Section 930 of the Heat Exchanger and _______
Cooling Tower Manual.
Before Cleaning:
Record plugging and fouling conditions. _______
Record clean washed areas indicative of high velocity and subsequent corrosion. _______
Walk around the unclean bundle, take pictures, gather samples of fouling products and scale for anal- _______
ysis if necessary. Look for distortion, leaks, and any unusual condition. Much of this evidence is
washed away when the bundle is cleaned.
After Cleaning:
Gage O.D. of accessible tubes with calipers. _______
Use a mirror to visually inspect the full length of the outside of as many tubes as possible. Have the _______
bundle rotated if necessary.
Look for localized corrosion near the shell inlet nozzle in areas of impingement. _______
Use a boroscope or a mirror and reflected light to visually inspect tube I.D. _______

December 1989  1989 Chevron USA Inc. All rights reserved. 900-19
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

Fig. 900-8 Inspection Checklist—Shell and Tube Exchangers (3 of 3)

Look at tube ends for corrosion and thinning using a tube I.D. gage. _______
Consider eddy current inspection for tube wall thickness. _______
Pull specimen tubes in strategic areas if necessary. Split the tube and check the entire length for _______
remaining wall thickness. The eddy current inspection may give you an idea of where to pull specimen
tubes.
Inspect the front and backsides of tubesheets for corrosion. _______
Verify that all gasket surfaces are clean and free of defects and corrosion that could affect the tight- _______
ness of the gasketed joint.
Check the tubesheet baffle gasket surfaces for bypassing. Bypassing can corrode both the tubesheet _______
gasket surface and the baffles.
Check tubesheets with seal welded tubes for warpage which can cause excessive baffle bypassing. _______
Inspect the tubesheet which is to be reused as a bundle to ensure that it is retubed. Look for the tube _______
hole enlargement using a “go/no-go” gage. Look for pieces of thimbles left in the tubesheet serrations.
Check for cracked ligaments. Note the history of the bundle. If roll leaks have been a problem, the liga-
ments are probably spongy, unable to hold a tight roll, and the tubesheet should be replaced.
Inspect the segmental cross baffles for hole enlargement by moving the tubes and observing excess _______
movement.
Look for general thinning of all carcass components. _______
Inspect the carcass for possible reuse when the bundle is retubed. Ensure that the frame will last the _______
life of the bundle, especially when the tube material is being upgraded.

900-20  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 900 Inspection and Testing

Fig. 900-9 Inspection Checklist—Air Cooled Exchangers

Date: _____________________
Equipment #: _____________________
Service: _____________________
Initial _______
Tubes:
Inspect fin condition and fouling. _______
Have header plugs removed and measure tube I.D. with an I.D. gage. _______
Consider eddy current inspection to determine tube wall thickness. _______
Pull specimen tubes if the eddy current inspection indicates it is warranted. _______

Header boxes and tubesheet:

Inspect the header box and tubesheet for wall thickness. _______
Check the plugs for impingement on the outlet sides. _______
Check the plug and header hole threads for corrosion and galling. _______

Fan, driver, and housing:

Inspect the fan for cracking or distress. _______
Check the guards to ensure they are intact, for personnel protection. _______
Check the light gage steel housing for general condition. _______

December 1989  1989 Chevron USA Inc. All rights reserved. 900-21
900 Inspection and Testing Heat Exchanger and Cooling Tower Manual

900-22  1989 Chevron USA Inc. All rights reserved. December 1989
1000 Maintenance

Abstract
This section discusses the major maintenance requirements for heat exchangers.
This includes preventing leaks, repairing body flanges, body flange gaskets and
torquing considerations, bundle replacement and repair, tube leak repairs, shell
repairs, and on-line leak repairs.

Contents Page
1010 Preventing Flange Leaks 1000-2
1011 Introduction and Scope
1012 Evaluation and Inspection of Flanges Face and Gasket Surface Warpage
1013 Gasket Surface Repair and Replacing Body Flanges
1014 Gasket Selection
1015 Stud Load Selection and Torque Calculations
1016 Studs, Nuts, and Hardened Washers
1017 Bolting-Up and Hot Torquing Flanges
1018 Tensioned Joints
1019 Flange Insulation
1020 Bundle Re-Tubing, Replacement, or Repair 1000-15
1021 General Considerations for Opening Up an Exchanger
1022 Considerations for Re-Tubing, Repairing, or Replacing the Bundle
1030 Tube Leak Repairs 1000-18
1031 General
1032 Types of Tube Plugs
1040 Shell Repairs 1000-20
1050 Online Leak Repairs 1000-22
1051 Types of Online Repairs
1052 Temporary Online Leak Repair Procedures
1060 References 1000-24

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-1
1000 Maintenance Heat Exchanger and Cooling Tower Manual

1010 Preventing Flange Leaks

1011 Introduction and Scope

Exchanger flange leaks are a major environmental, safety, and economic problem.
In 1994 Chevron initiated a study in El Segundo to determine the root causes of
leaking flanges. The study developed recommendations to eliminate leaks in
existing exchangers, and also specified design practices for new heat exchangers.
The program came to be known as ETGIS (Evaluation, Tightening, Gaskets, 
Installation, and Studs). The ETGIS principals were further used to develop
Chevron Global Refining Instruction GR-800, Flange Gasket and Bolting 
Instruction. The recommendations with respect to gasket and flange design were
also incorporated into the Heat Exchanger specification EXH-SC-2583 for the
design and fabrication of new exchangers. The purpose of this Section is to 
summarize and provide the general background and technical justifications behind
the recommendations.
The ETGIS study found there is no single cure for the problem of leaking flanges.
The flange connection must be evaluated as a system: flange design, gasket type,
gasket stress, assembly procedure, bolting technique, and stud loads. Generally, the
most common reason for leaking flanges is:
• insufficient gasket load due to improper stud loading
• long term gasket relaxation
• wrong gasket selection
The following sections will address the main components of the system:
• inspection and repair of flanges
• gasket selection
• installation of gaskets
• bolting up
• torquing, or tensioning.
In addition to information provided in the subsequent Sections, ETC offers training
on bolting and gasket selection for flange connections for engineers and
maintenance staff.
Note Screw Plugs (also called Breech Lock Closures), and flanges which use sheet
gaskets are not covered in this section.

1000-2  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

1012 Evaluation and Inspection of Flanges Face and Gasket Surface Warpage
Exchanger body flanges should be visually inspected for leaks on a regular basis
during normal operation and prior to the shutdown. All leaks should be monitored
and recorded to document the connection improvement or deterioration.
 Caution Even very small leaks have the potential to get worse during process
upset conditions. They should not be ignored. The goal (which is very
attainable) is to have NO leaks between scheduled maintenance 
shutdowns.
Due to flange design and/or uneven or excessive bolt load, flange faces can roll
toward each other, resulting in uneven gasket loading and leakage. In some cases,
this rotation can result in permanent flange warpage. Flange warpage could be a
result of faulty fabrication, residual stress in the flange after fabrication or 
deformation due to overloading from excessive stud loads.
Figure 1000-1 illustrates the impact of flange rotation on gasket sealing. In the top
example a corrugated metal or kamprofile type gasket loses some effective seal
area. The joint is probably not leaking because an increase in bolt load resulted in
higher gasket stresses around the OD of the gasket. The bottom example in
Figure 1000-1 illustrates a spiral wound gasket in the rotated flange. The fulcrum of
rotation is the outer compression ring. The increase in stud stress can cause the
flanges to lift off the gasket, causing the flange to leak. This is one of the reasons
that Chevron has moved away from employing spiral wound gaskets in refinery
applications.
Flange warpage is limited to 0.05 inches (1.25 mm) in the radial direction. A
minimum gap of 1/8 inch (3.175 mm) is recommended between the mating flanges
when they are bolted together inclusive of a gasket. All new exchangers manufac-
tured according to the Shell and Tube Heat Exchanger Specification 
EXH-SC-2583 shall have the nominal clearance between flanges after assembly of
no less than 3/16 inch (4.76 mm). Kamprofile gaskets can be used when the 1/8 inch
(3.15 mm) gap cannot be maintained using a CMG gasket.
Metal clad gaskets, also called double jacketed or DJ’s have been eliminated from
use in heat exchangers due to poor performance between flanges operating at
different temperatures. This temperature difference leads to a differential thermal
expansion across the radial width of the gaskets, which will cause the rolled edge of
a clad gasket to split. Once the gasket splits, it will leak. It was found that 
40 percent of clad gaskets with graphite filler will fail over a plant’s operating cycle.
The amount of differential expansion varies from flange to flange depending on the
diameter and temperature difference, but generally will range from a few 
thousandths of an inch up to 1/4 inch (6.35 mm).
An axial misalignment between the flanges is shown Figure 1000-2. This
misalignment could cause a leak in the field. A graphite-faced gasket will
accommodate this problem much better than any other gasket, and remain leak-free.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-3
1000 Maintenance Heat Exchanger and Cooling Tower Manual

High quality gasket seating surfaces are essential for proper sealing. All flange
gasket surfaces shall be inspected for flatness, and proper surface finishes which are
free from dents, corrosion, scratches or other imperfections. The allowable 
circumferential warpage criteria are defined in the API 660, Table 3 (There are more
stringent requirements for special applications heat exchangers). The 
allowable radial warpage is –0 to +0.012 inches (-0 to +0.30 mm) and the surface
roughness should be 125–250 micro inches (3.175–6.350 micro meters.
There are some heat exchangers installed within Chevron facilities that employed
nubbins in conjunction with jacketed gaskets or solid metal gaskets. The ETGIS
study determined that nubbins do not reduce the chances of having a leak and in
many cases can be the cause of the leak. Nubbins should be removed if found in the
existing exchangers. Instead, a Graphite-Covered Corrugated Metal 
CMG-style gasket should be installed. The Shell and Tube Heat Exchanger 
EXH-SC-2583 does not allow nubbins for new equipment.
 Caution Do not install a Graphite-Covered Corrugated Metal CMG-Style or
Kamprofile Style gasket on a nubbin..

Fig. 1000-1 Impact of Flange Rotation on Gasket Sealing

1/8" MIN. REQUIRED

GASKET

Rotation reduces effective seating area of the gasket. The joint is probably not leaking because the fulcrum
for rotation is the gasket
1/8" MIN. REQUIRED

SPIRAL WOUND GASKET

The fulcrum of rotation is the compression in the spiral wound gasket. Increase in the bolt stress will
decrease the gasket stress.

1000-4  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

Fig. 1000-2 Impact of Flange Misalignment on Gasket Sealing

PRIMARY SEAL CAN BE LOST

A GRAPHITE FACE GASKET CAN TOLERATE

SOME MISALIGNMENT.

GASKET

1013 Gasket Surface Repair and Replacing Body Flanges

Flange gasket surfaces that do not meet the warpage and roughness criteria outlined
in the Section 1012 may be repaired by machining to within “plus or minus 0.002
inches (0.05 mm) from flat” in the radial and circumferential directions. Both gasket
surfaces should be machined, except where the mating flange has been measured
and is verified not to be warped.
The minimum required flange thickness per ASME should be verified prior to main-
tenance gasket surface machining activities. Depending on the flange thickness, the
flange may need to be weld built-up before machining to maintain the minimum
ASME Section VIII Div. 1 Appendix 2 required thickness. All welding repairs must
be performed using ASME Code weld procedures and an NBIC qualified contractor.
All new heat exchangers manufactured in accordance with the Shell and Tube Heat
Exchanger EXH-SC-2583 have a minimum of 1/8 inch (3.175 mm) additional thick-
ness intended for future machining.
 Caution Following gasket surface matching activities, a check of the gap
between the assembled flange set should be made to ensure that the
minimum

gap of 1/8" inclusive of the gasket is maintained.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-5
1000 Maintenance Heat Exchanger and Cooling Tower Manual

In cases when the heat exchanger flanges may not be repaired due to excessive
gasket surface deformation or flange rotation the flanges should be replaced with
new flanges that are designed per requirements listed the Shell and Tube Heat
Exchanger Specification EXH-SC-2583. Replacing the body flanges should be
considered an ASME Code alteration.

Fig. 1000-3 Typical Gasket Types

Metal Clad

Solid Metal

Corrugated Metal

Kamprofile

Spiral Wound

1000-6  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

1014 Gasket Selection

The ETGIS study evaluated performance of several gasket styles installed in heat
exchangers. Most of them were either metal clad, solid metal or spiral wound
gaskets (see Figure 1000-3 for gasket sketches).
The clad gaskets are most likely to fail as they can not tolerate the differential radial
expansion that takes place between the flanges. The other problems found with the
metal clad gaskets include over-compression breakage, graphoil blow-outs and
relaxation.
• The solid metal gaskets are also easily damaged by radial differential
expansion that results in scratching or galling in either the gasket or in the
flange gasket surface. Performance of the spiral wound gaskets was in general
better than the metal clad gaskets.
• The spiral wound gaskets experienced problems that include inward buckling
of windings and unwinding of the spirals. Inner and outer rings on the gasket
alleviate this problem.
The Corrugated Metal Gaskets (CMG) performed the best in terms of leak free
operation and were found to be cost effective solutions for gasket replacements.
All body and floating head flanges (in non-oxidizing services up to 800F [425C])
should use 316 L or 304 L stainless steel Corrugated Metal Gasket (CMG) coated
with Polycarbon APX-2 graphite. See Figure 1000-4 for an example of the CMG
gasket. The minimum CMG gasket width is 1/2 inch for all body and floating head
joints, and a minimum width of 3/8 inch (9.5 mm) for pass partitions. The specific
requirements for CMG gaskets are specified in Specification EHX-SU-5150.
A Kamprofile gasket can be used if a narrower gasket is required due to inadequate
gasket stress or excessive bolt loading. Minimum width for a Kamprofile gasket is
0.375 inch (9.53 mm). For services with design temperatures exceeding 800F
(425C), the gaskets should be selected on a case by case basis. In many such
services, a Kamprofile type gasket can be used with mica on the OD (and ID too in
some cases) to protect the graphite from oxidation. The specific requirements for
Kamprofile gaskets are specified in Specification EXH-SU-5151.
All gaskets should be test fit prior to final assembly to make sure they will fit before
gluing the gasket in place. For example, measure the flange tongue and recess. If the
flange set is assembled without a gasket, there should be a 1/16 inch (1.6 mm) gap
between the flanges at the flange rim. If the gap is less than 1/16 inch (1.6 mm), a
Kamprofile gasket should be used in place of CMG type gaskets (to prevent the
flanges from touching after applying a proper gasket stress). In addition, a thicker
gasket inner core may be specified to provide proper clearance.
For graphite covered gaskets, small scratches or depressions in the soft graphite will
not pose a sealing problem as the graphite will flow together when it is compressed.
If the graphite is damaged to the point where the core is exposed, the area should be
repaired by covering it with adhesive-backed APX-2 graphite tape.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-7
1000 Maintenance Heat Exchanger and Cooling Tower Manual

Fig. 1000-4 Graphite Covered CMG Type Gasket with Pass Partition

Fig. 1000-5 Example of Stud Loads Changes due to Short Term and Long Term Gasket Relaxation

120

110

100
Stu Load (% of Initial Load)

90

80

70

60

50

40
d

9/ 30

11 999

12 999

1/ 99

11 000

12 000

1/ 00
g

10 999

2/ 00

3/ 00

4/ 00

5/ 00

6/ 00

7/ 00

8/ 00

9/ 00

10 000

2/ 01

3/ 01

4/ 01
01
27
fte Loa

7/

0
To ati

20

20

20

20

20

20

20

20

20

20

20

20
Se e 7/

/1

/1

/1

/2

/2

/2
1

2
e

1/

1/

1/

1/

1/

1/

1/

1/

1/

1/

1/

1/

1/

1/
al

/1

/1

/1

/1

/1

/1
rH

ic
qu
iti

rv
r
In
A
ot

n
rI
rH

te
te

Af
Af

1000-8  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

1015 Stud Load Selection and Torque Calculations

Chevron has studied gasket relaxation extensively in operating heat exchangers.
Typical changes in the heat exchanger stud load over 15 months are summarized in
Figure 1000-5. The stud load drops significantly, about 30 percent, after initial
heating due to graphite flow. The impact of initial gasket relaxation due to flow of
graphite is compensated with hot torquing the exchanger. About three days after the
hot torquing the stud load drops again about 30 percent as a result of additional
gasket relaxation. Long term, it takes about 15 months before the gasket stops
relaxing and the stud load stabilizes. In the example below the stud load declined to
about 50 percent of the initial load.
 Caution On CMG and Kamprofile gaskets, a graphite coating significantly
greater than 0.020 inch (0.508 mm) will result in excessive relaxation
and loss of bolt load

Stud Load Selection

Stud loads should be selected after considering several factors:
• gasket relaxation
• sudden changes in the process temperature (a drop in temperature can cause a
drop in stud loads)
• inaccuracy of tightening equipment
Directionally, accounting for all these factors means higher stud loads. At the same
time, the stud loads should be limited to prevent yielding of the studs and/or flanges.
Chevron developed procedures to determine the stud loads and required torque
values that are described in the refinery Instructions GR-800 and Specification
EXH-SC-2583. The purpose of these calculations is to assure a proper gasket
seating stress. Criteria for calculating stud loads include:
• The target gasket seating stress is 20,000 psi (1380 bar). The minimum and
maximum gasket seating stresses are 12,500 and 40,000 psi respectively (860
and 2760 bar).
• The stud loads to obtain the required gasket seating stresses should be between
50,000 and 80,000 psi (3450 and 5500 bar) for standard SA-193-B7 studs, and
between 50,00 and 70,000 psi (3450 and 4830 bar) SA-193-B7M studs. If it is
necessary to exceed 80,000 psi for B7 studs, or 70,000 psi for B7M studs,
consult with ETC’s Heat Exchanger Specialist. In no case should the stud load
exceed 90% of the minimum specified yield strength for the bolting material as
listed in ASME Section II, Part D.
• The ASME Code has not established the gasket factor “m” and the gasket
seating stress “y” for CMG type gaskets. In the flange calculations use the
gasket factor m =3.75 and the gasket seating stress y=7600.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-9
1000 Maintenance Heat Exchanger and Cooling Tower Manual

To determine the required torque, use the following formula:

T = K  S nd  12  S rd  S s
(Eq. 1000-1)
where:
T = Torque, ft.-lb (N-m)
K = Torque Coefficient (K=0.17 when Jet Lube 550 is used)
Snd = Stud nominal diameter, inches (m)
Srd = Stud root area, inches2 (m2)
Ss = Stud stress, psi (N/m2)
Note Note: for SI units delete division by 12 from the equation.
The torque coefficient K depends on the type of lubricant and condition of the
threads. Use a friction factor of 0.17 for the recommended molybdenum disulfide
based Jet-Lube 550 lubricant and new studs. A proper lubrication of the stud is a
critical step to ensure proper gasket loads. Jet -Lube 550 is currently the only
approved anti-seize and any substitution requires approval of the local bolting and
sealing Subject Matter Expert (SME) or ETC Heat Exchanger Specialist.

1016 Studs, Nuts, and Hardened Washers

Each time the exchanger flanges are opened, the studs and nuts should be replaced
with new material of the same type. Use of the old studs and/or nuts leads to 
unpredictability of the friction factor and will result in high variations in stud stress.
Studs should be long enough for full engagement through the nut, but should not
extend more than two threads beyond the nut, unless tensioning will be utilized. As
discussed in the previous section, Jet Lube 550 is the only approved anti-seize.
The charts shown in Figure 1000-6 and Figure 1000-7 compare a stud load
distribution for two bolting up procedures: re-use of wire brushed studs, and
torquing the new studs. High variation in the stud loads for the re-used studs
explains the requirement of installing new studs each time when the exchanger is
opened for maintenance.
The hardened washers should be ASTM F-436 and should be provided under the nut
that is to be torqued. Washer should be placed so manual torque wrenches are
always pulled downward to avoid back injuries. In some cases, it is appropriate to
place hardened washers under both nuts. The washer size should be compatible with
the stud size and the washer thickness should be at least 1/8 inch (3.175 mm).
 Caution The hardened washers should be omitted on floating head flanges due
to service-induced cracking potential.

1000-10  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

Fig. 1000-6 Stud Load Distribution for Wire Brushed Re-Used Studs(1)
60,000

Average Bolt Stress 28,000

50,000

Targeted Bolt Stress

33,500

40,000
Stud Load in PSI

30,000

20,000

10,000

0
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
61
63
Stud Number

(1) To obtain bars, multiply psi by 0.07

1017 Bolting-Up and Hot Torquing Flanges

ETGIS research concluded that a traditional star pattern for bolting can be
significantly simplified for body flanges with more than 24 studs (this will include
most exchangers) without compromising the stud stress distribution. The modified
star pattern (see Figure 1000-8) consists of four passes.
• In the first pass tighten all studs with a wrench for alignment
• In the second pass tighten the four studs as shown in Figure 1000-4 to 
100 percent of the final torque value
• In the third pass tighten all studs to the 100 percent required torque
• In the fourth pass go around the flange one more time in a circular pattern and
re-tighten the first four studs to compensate for any unloading of studs from the
previous pass.
The details on the modified star pattern bolting-up procedure are shown in the 
GR-800 Instructions, Appendix IV. For the exchanger body flanges with less than
24 studs use the traditional star pattern procedure that is described in of GR-800
Instructions, Appendix VIII.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-11
1000 Maintenance Heat Exchanger and Cooling Tower Manual

Fig. 1000-7 Stud Load Distribution for New Studs(1)

60,000

Average Bolt Stress 29,500

50,000

Targeted Bolt Stress

33,500

40,000
Stud Load in PSI

30,000

20,000

10,000

0
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
61
63
Stud Number

(1) To obtain bars, multiply psi by 0.07

Due to the characteristics of graphite faced gaskets, the flanged joint must be 
re-torqued or “hot torqued” as the unit is starting up. This is due to the fact that the
graphite facing will “flow” as it warms up, resulting in a reduction in gasket stress
and stud load.
During start-up, re-torque (hot torque) the connection when the cooler of the two
mating flanges is between 250 and 400F (120 and 204C). When the flange
temperature, measured at the rim of the flange, increases beyond 400F (204C) the
lubricant begins to burn off and the nut friction factor can increase significantly. Hot
torque the joints in a circular pattern. Make sure that each nut moves or is “broken
loose”. If the flange temperature does not reach 250F (120C) due to service 
conditions, hot torque the flanged joint once the operating temperature is reached.
The gasket seating stress is based on the stud load, and the stud load is a result of
the applied torque which depends strongly on lubrication and the accuracy of
torquing equipment. Experience shows that proper torquing is a critical factor in
assuring the leak free exchanger joints.

1000-12  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

Fig. 1000-8 Modified Star Pattern for Exchanger Body Flanges with 24 or More Studs

First Pass: Second

Hand Pass:
4
Tighten All Torque 4
Studs For 3
Studs To
Alignment 100%

Third Fourth
Pass: Pass:
Torque All Torque All
Studs To 100% 100% Studs To
100% in 100% in
Circular Circular
Pattern Pattern

1018 Tensioned Joints

Tensioning is another method of setting the stud loads. During the tensioning the
stud is pulled or stretched until the proper load is obtained, then the nut is manually
run down until it contacts the flange. The stud is released and the holding force is
transferred to the nut. With this technique the stud-to-nut friction factor is not an
issue. Tensioning is an option when very accurate stud loads are required If studs
are to be re-used, the studs should be tensioned to be sure accurate bolt load is
attained.
Generally, hot tensioning is not recommended because the O-ring seals in the
tensioning equipment may fail due to overheating. However, depending on the
flange temperature, techniques can be used to retension studs when the system is
hot. This includes using a few heads, moving quickly and cooling the heads in water
before working on the next set of studs.
The studs to be tensioned must be long enough to leave one stud diameter past the
nut for connection of the tensioning heads. Tensioning will set the most consistent
stud loads. This technique is seldom used for heat exchanger flanges because it
slower and more expensive.
Figure 1000-9 shows an example of tensioning equipment and Figure 1000-10
shows a distribution of the stud load after using the tensioning technique.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-13
1000 Maintenance Heat Exchanger and Cooling Tower Manual

Fig. 1000-9 Example of Tensioning Equipment

Fig. 1000-10 Stud Load Distribution after Tensioning(1)

60,000

Average Bolt Stress 47,300

Targeted Bolt
Stress 45,000

50,000

40,000
Stud Load in PSI

30,000

20,000

10,000

0
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
61
63
65
67
69

Stud Number

(1) To obtain bars, multiply psi by 0.07

1000-14  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

1019 Flange Insulation

In general, heat exchanger flanges should not be insulated. Studies determined that
insulated studs may cool down slower than the flanges during process upsets.
Consequently, the stud load may decrease and the gasket stress will decrease to a
point where the flange starts to leak. The heat exchangers in a cryogenic service
may be excluded from this rule and should be considered on a case by case basis.
If personnel protection is required, apply a perforated steel enclosure instead of
insulation.

1020 Bundle Re-Tubing, Replacement, or Repair

This section lists the major considerations for determining:
• whether to open an exchanger and
• whether to re-tube, replace, or repair a tube bundle.
It also includes a worksheet for use in making these decisions before a plant 
turnaround.

1021 General Considerations for Opening Up an Exchanger

It is expensive to open an exchanger for inspection and repairs. Also, for exchangers
that are operating well, opening them unnecessarily may cause future problems.
Therefore, it is very important to have an organized procedure for making deci-
sions. This is especially important going into a plant turnaround where a large
number of exchangers will need to be investigated.
The following list of considerations can be used to help determine when to open an
exchanger.
• Is the exchanger leaking internally or externally?
• Does the unit need recertification?
This is usually a concern only for steam generators, which require periodic state
inspection and certification. Some states, such as California, require internal inspec-
tions of any pressure vessel per API 510 every ten years.
• Is the exchanger performance satisfactory?
In other words, does the U-value or DP indicate fouling or internal damage?
• Based on this bundle’s history and the history of similar bundles in similar
services, will the bundle last until the next turnaround?
Accurate inspection records are important in making this decision.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-15
1000 Maintenance Heat Exchanger and Cooling Tower Manual

• Can exchanger be cleaned or repaired on the run?

Isolating an exchanger and cleaning it with the surrounding equipment oper-
ating entails significant safety and operating problems. This is usually done
only if the exchanger and piping were designed to allow for online cleaning.
• Can the unit be chemically cleaned?
Light uniform fouling may be chemically cleanable. Locally plugged
exchangers cannot be chemically cleaned. Chemical cleaning is expensive and
all of the environmental and safety implications should be considered first.
Section 700 of the Corrosion Prevention and Metallurgy Manual, Book 1,
discusses chemical cleaning in more detail.

1022 Considerations for Re-Tubing, Repairing, or Replacing the Bundle

Assuming the bundle is leaking or, based on past experience, it may not last until
the next turnaround, then some repairs will need to be made. The following areas
need to be considered before deciding on the extent of repairs.
• Is bundle nearing the end of its historical life?
If a bundle is leaking and it is not approaching its historical life, then a failure
analysis should be performed to determine if design changes are required. The
best way to determine the cause of failure is to pull the leaking tubes and
inspect them. In the case of a U-tube bundle, extracting tubes is not practical
unless the tube is on the periphery of the bundle.In this case, NDE techniques
such as Eddy Current Testing or Ultrasonic IRIS can be used to assess tube
condition.
• Can tubesheet be reused?
• Normally, a tubesheet can be used two or three times before it becomes diffi-
cult to seal tubes in the tubesheet by mechanical rolling. This is especially true
of copper alloy materials. Carbon steel materials are more tolerant to multiple
retubes. If a tubesheet cannot be reused, then a new bundle will be necessary.
Also, the time required during a shutdown to retube a bundle can often justify
the cost of purchasing a replacement bundle in advance of the shutdown.
• Can leaking tubes be plugged or replaced?
Individual leaking tubes can be the sign of a much larger problem. Simply plug-
ging or replacing leaking tubes may be setting yourself up for another failure
before the next shutdown.

1000-16  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

The leaking tube should be pulled and inspected to determine the cause of
failure. The position of the failed tube relative to baffles and nozzles should be
determined and recorded. This can help identify the cause of the failure. If it is
not obvious that repairing or replacing the tube will ensure a sound bundle, then
the bundle should be re-tubed or replaced. In the case of a U-tube bundle,
leaking tubes should be plugged because tube replacement is not practical
unless the subject tube is on the periphery of the bundle.
Sometimes, all of the leaks may be in one part of the bundle. In this case, a
design change (i.e., replacing tubes with solid rods, changing inlet
configuration, etc.) may eliminate the problem.
• If a new bundle is required, should it be redesigned?
Simple design changes may substantially improve a bundle’s operation and
service life. If a bundle needs to be re-tubed or replaced, the incremental cost
for making these design changes may not be very high. However, the cause of
the bundle deterioration should always be identified and process changes
considered along with design changes.
Some examples of bundle configuration changes that may improve the long term
heat transfer or bundle life are listed below. (Discuss these changes and
improvements with an ETC heat exchanger specialist or the ETC Process Design or
Process Consultation groups.)
• Change inlet impingement design to eliminate tube vibration wear.
• Plug tubes to increase tube velocity. (This improves heat transfer if fouling is
significant.). Plugging some tubes will increase the fluid velocity if the flow
rate through the exchanger is set by the process and there are no hydraulic
limitations due to an increase in pressure drop. In such services as cooling
water, plugging the tubes will have no impact on the velocity. In these services,
the cooling water flow is controlled by a set pressure drop between the supply
and return headers. Plugging the tubes in cooling water services will cause
deterioration in performance due to a lowering of the cooling water flow rate
and increase in the cooling water outlet temperature.
• Change shell side pass configuration and baffling to improve heat transfer and
decrease fouling.
• Use different materials that will increase the bundle reliability and life. (As a
general rule, even in corrosive services such as sea water, a bundle should last
at least two operating runs. Contact the ETC Materials and Equipment
Engineering Unit for help in this area.)
• Change from floating head to U-tube design.

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-17
1000 Maintenance Heat Exchanger and Cooling Tower Manual

If a tube bundle is known to be at end of life and is of carbon steel construction, it

can be much more economical to purchase a new tube bundle ahead of the turn-
around and have it available at the plant when the shut down begins. Trying to
determine the cause of failure and reworking an existing tube bundle in emergency
situations leads to bad decisions such as:
• Resorting to the available material in lieu of material of choice
• Reusing questionable parts due to lack of material availability
• Losing opportunity for design modifications due to lack of rework time.

1030 Tube Leak Repairs

1031 General
Roll leaks should be repaired by re-rolling (never by driving a drift pin or pin
wrench in the tube.) Caution shall be taken not to overexpand the tubes. Typically,
an expanded (rolled) tube to tubesheet joint is effective up to an eight percent wall
reduction, dependant on tube and tubesheet material. When re-rolling leaking tube
to tubesheet joints in the field, one should not expect to gain a reliable seal when the
tube wall reduction increases over 12 percent. When the tube wall reduction
required for an effective seal is excessive, the owner should require either welding
the tube end, or plugging the tube.
When plugging defective tubes, the tubes shall be vented (drilled hole or fly cut)
behind the tubesheet to prevent creating a dead space that can build pressure
between the tube plugs. The inside surface of the tube ends should be thoroughly
cleaned before installing plugs. Tube plugs shall be the same material as the tube
and tubesheet.

1032 Types of Tube Plugs

Tapered Tube Plug
• Advantages
– most common, versatile and economical type of tube plug
– can be machined by local machine shop on short notice
– can be machined from any material where round stock is readily available
– least expensive of all types of tube plugs
– can be welded to tube/tubesheet where reliable leak free plugs are required

1000-18  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

• Disadvantages
– not recommended with high pressure differentials with the higher
pressure on the shell side (greater than 150 psi for 1" OD tubes, or 280 psi
for 3/4" OD tubes) without welding in place
– can be a safety hazard if a high shell test pressure is employed, unless
welded to tubesheet
– can damage tube hole in tubesheet or even cause leaks in adjacent tubes if
installed with excessive force. Seal welding tube and/or tube plugs can also
result in leaks at adjacent tubes.

Two Piece Plugs

• Advantages
– can be installed with no damage or very minimal damage to the tube end
and/or tubesheet
– able to seal at slightly higher pressures than the single piece tapered tube
plugs (use good judgment when evaluating plug manufacturer’s pressure
capability)
– works well if tube ends at the tubesheet are damaged
• Disadvantages
– have to be ordered from plug manufacturer for specific material and tube
size, may not be readily available
– plug sets can not seal the range of tube sizes that single piece plugs can
– tubes that are to be plugged need to be round and free from corrosion
– are not easily welded when needed for leak tightness
– have two possible leak paths after installation (bushing-to-tube and 
plug-to-bushing)
– higher cost than single tapered tube plug

Pop A Plug
• Advantages
– internal seal and capable of sealing high differential pressures
– ability to plug both ends of a straight tube from one end of the bundle
– works well if tube ends at the tubesheet are damaged
• Disadvantages
– have to be ordered from plug manufacturer for specific material and tube
size, may not be readily available
– plug sets can not seal the range of tube sizes that single piece plugs can
– are not easily welded when needed for leak tightness or if leak persists
after attempted installation
– special installation tooling and training for operator(s) is required or must
be installed by a tube plugging contractor
– highest cost plug

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-19
1000 Maintenance Heat Exchanger and Cooling Tower Manual

Figure 1000-11 shows examples of the above tube plug types. Normally, tapered
plugs are used.

Fig. 1000-11 Types of Tube Plugs

Tapered Metal Two Piece Plug Pop-A-Plug™

Plug

When using the single tapered plugs, the design engineer should consider if the plug
needs to be welded to the tubesheet (see Figure 1000-12). The following are the
major considerations:
• For services with the shell side pressure significantly higher than the tube side
pressure, seal welding plugs to the tubesheet should be considered. A welding
specialist should be consulted for alloy tubesheets, tubesheets with weld
overlay and for exchangers in hydrogen or sour service. Many times, welding
on contaminated tubesheets creates more problems than a single tube leak.


Adjacent tubes can develop roll leaks and may need to be seal-welded as well.
• For services with a tube side pressure higher than the shell side pressure, seal
welding of plugs is not necessary. However, welding the plugs may be recom-
mended to assure safety during a hydro test of the shell side with the channel or
channel cover removed.
• In cases where the tubes are seal-welded or strength-welded into the tubesheet,
a tapered metal plug should be seal-welded into place.

1040 Shell Repairs

Thin areas of shells can be weld repaired. All weld repairs must be ground smooth
for bundle installation and removal. If extensive weld repairs are required, cutting
out the corroded area and installing a butt-welded patch may be necessary. In either
case, whether making weld repairs or replacing part of the shell, precautions must
be taken to prevent out-of-roundness.
If repairs are required, it must be remembered that exchanger shells are pressure
vessels and fall within the scope of the ASME Code. Refer to the Pressure Vessel
Manual for more information on ASME Code repair procedures.

1000-20  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

Fig. 1000-12 Tapered Metal Plug Chart

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-21
1000 Maintenance Heat Exchanger and Cooling Tower Manual

1050 Online Leak Repairs

There are times when repairs must be made to an exchanger while it is in operation.
Normally, this is only when the exchanger cannot be taken out of service without
risk to personnel or equipment or without major expense. These repairs are usually
external leak repairs and should only be undertaken with the utmost caution with the
appropriate safety, operating, engineering, and maintenance review.

1051 Types of Online Repairs

All of the repairs below are intended to resolve leaking flange connections. Flange
leaks should be viewed as very serious events because they will require that
modifications be made to operating equipment. If the flange set and gasket are prop-
erly designed and assembled, the expectation is that flange leaks should not occur.
Therefore if a flange leak does develop, if is very important that the root cause be
identified and plans put in place for a permanent corrective action the next time the
connection can be opened.
Online repairs that can be made include:
1. Re-torquing flanges
a. This usually involves tightening all of the bolts.
b. Follow appropriate torquing procedures to prevent worsening the problem.
c. Retorquing flanges where the studs have been exposed to temperature
exceeding 400F (204C)
2. Tapping the flange and pumping in sealant between the flange faces.
3. Installing a ring around two flanges and pumping in sealant.
Repairs 2 and 3 above are only temporary fixes and can be very expensive. 
Permanent repairs must be made during the next turnaround. The work required to
remove the leak sealing compound in order to repair the exchanger can also be time
consuming and costly. Section 1052 discusses these temporary repairs in more
detail.

1052 Temporary Online Leak Repair Procedures

Leak Sealing Contractors
Normally, special leak sealing contractors provide personnel who are trained and
qualified with regard to the equipment, material, and techniques commonly used for
online leak sealing. Sealing compounds must be resistant to the leaking material and
suitable for operating pressure and temperature.

1000-22  1998–2007 Chevron USA Inc. All rights reserved. November 2007
Heat Exchanger and Cooling Tower Manual 1000 Maintenance

Clamps, boxes, and other enclosures used for leak sealing are normally designed,
fabricated and installed by the leak sealing contractor (see Figure 1000-13). The
leak seal contractor’s clamp or box design should be reviewed carefully. The clamp
or box material should be the same as the equipment being sealed and subject the
same post weld heat treatment requirements. The clamp or box should be designed
using the allowable stresses from the ASME Code.

Safety Considerations
• The contractor must understand all local safety regulations applicable to the
work.
• The contractor must understand the nature of the leakage and why it is
occurring in order to use a safe procedure.
• A detailed written procedure must be developed with input from the contractor.
• All safety precautions and protective clothing required should be reviewed with
the contractor.
• Sealant injections should not be made into any pressure relief valve or bursting
disc if the sealant could obstruct their free and full discharge requirements.

Fig. 1000-13 Methods for Sealing Flange Leaks

November 2007  1998–2007 Chevron USA Inc. All rights reserved. 1000-23
1000 Maintenance Heat Exchanger and Cooling Tower Manual

1060 References
1. GR-800, “Flange Gasket and Bolting Practice”
2. ASME Code, Section VIII, Division 1. Latest ed. Appendix S and 2.
3. ASME Code, Section II
4. ASTM Standard F-436, Hardened Steel Washers
5. API 510, Pressure Vessel Inspection Code
6. API 660, Shell and Tube Heat Exchangers

1000-24  1998–2007 Chevron USA Inc. All rights reserved. November 2007
1100 Performance Monitoring

Abstract
This section discusses performance monitoring of operating heat exchangers.
Emphasis is placed on basic principles and available computer tools.

Contents Page
1110 Reasons for Monitoring 1100-2
1120 Measuring Performance 1100-2
1121 Heat Transfer Efficiency
1122 Pressure Drop
1130 Monitoring Frequency 1100-6
1140 Computer Programs 1100-7
1150 Example: Evaluating a Heat Exchanger 1100-8
1160 Monitoring Followup 1100-10

December 1989  1989 Chevron USA Inc. All rights reserved. 1100-1
1100 Performance Monitoring Heat Exchanger and Cooling Tower Manual

1110 Reasons for Monitoring

Understanding past and current performance of heat exchangers and being able to
predict future performance are important for:
• Maintenance planning
• Troubleshooting
• Process improvements
Before shutdown, Operations must decide which exchangers should be cleaned. A
good monitoring program will define the degree of fouling, the fouling trend with
time, and the economic impact of cleaning or not cleaning an exchanger.
Having a continuous record of heat exchanger performance is very valuable for
troubleshooting. Such a record may identify cause and effect when there is a change
in the historical fouling pattern of an exchanger. The effectiveness of the record
depends, however, on the extent that total unit operations are monitored routinely.
Monitoring information can also be used to define the economic incentive for
possible modifications or redesign of poor performers. Older exchangers could be
performing poorly because of outmoded design and/or changes in service since the
original design.

1120 Measuring Performance

Heat exchanger performance should be measured in terms of heat transfer effi-
ciency and hydraulic capacity. Performance in both areas can decline with time and
cannot usually be recovered without mechanical cleaning. The impact of declining
performance may be increased fuel usage, horsepower consumption, and/or
decreased throughput.
Monitoring involves comparing observed performance to expected performance.
Process monitoring systems usually gather the information needed to calculate heat
transfer efficiency. Field measurements are usually required to obtain hydraulic
data. Calculations can be completed manually or programmed into automatic
computer monitoring systems such as CHEVMON. See Section 1042 of the Instru-
mentation and Control Manual for a description of CHEVMON and its capabilities.

1121 Heat Transfer Efficiency

One measure of performance is the overall heat transfer coefficient, Us, as defined
by the equation:

Us = Q / ( A  MTD )
(Eq. 1100-1)
where:
Us = Heat transfer coefficient, Btu/hrft2
MTD = Mean temperature difference, F

1100-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 1100 Performance Monitoring

Q = Rate of heat transferred, Btu/hr

A = Area of tubes, ft2
See Section 211 for calculation of MTD.
Monitoring Us defines the change in heat transfer capacity with time. Economics
can be applied to networks and to single exchanger systems (non-networks) such as
condensers or feed bottoms exchangers.
Tracking Us alone can be misleading, as it may vary considerably as a result of
process changes. It is more reliable to monitor the fouling resistance, rf, directly by
the following equation:

rf = (1 / Us) - (1/Ucl)
= (MTD  A / Q ) - [(1 / hi) + (1 / ho) + rw ]
(Eq. 1100-2)
where:
rf = Fouling resistance, hrft2F/Btu
Us = Service heat transfer coefficient, Btu/hrft2F
Ucl = Clean heat transfer coefficient, Btu/hrft2F
hio = Inside heat transfer coefficient, Btu/hrft2F (referred to outside
area)
ho = Outside heat transfer coefficient, Btu/hrft2F
rw = Tube wall heat resistance, hrft2F/Btu
Figure 1100-1 is an example of how fouling resistance can vary with time. For the
exchanger in this example, the fouling rate was fairly constant until 15-Oct, at
which time something caused a rapid escalation of fouling. The time at which the
problem began is clearly pinpointed.

December 1989  1989 Chevron USA Inc. All rights reserved. 1100-3
1100 Performance Monitoring Heat Exchanger and Cooling Tower Manual

Fig. 1100-1 Heat Exchanger Fouling, rf vs. Time

Figure 1100-2 is a corresponding plot of Us. As can be seen, the graph does not
pinpoint the time of rapid escalation.

Fig. 1100-2 Heat Exchanger Fouling, Us vs. Time

1100-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 1100 Performance Monitoring

Figures 1100-3 and 1100-4 show some typical fouling patterns that may be discov-
ered when rf is tracked with time. The shape of the curve reflects the fouling mecha-
nism. Section 240 describes the principal fouling mechanisms.

Fig. 1100-3 Fouling Data Trends: Example 1

Fig. 1100-4 Fouling Data Trends: Example 2

December 1989  1989 Chevron USA Inc. All rights reserved. 1100-5
1100 Performance Monitoring Heat Exchanger and Cooling Tower Manual

Monitoring may sometimes produce data that do not appear to make sense, such as
the horizontal line in Figure 1100-4. Possible reasons for this horizontal line could
be poor maintenance (i.e., the exchanger was not cleaned), or a wrong correlation
was used to calculate heat transfer. The latter could occur if the process is operating
in the laminar flow regime when turbulent flow was assumed for the calculations.
If the fouling mechanism is known, it becomes feasible to devise a program to
reduce or eliminate fouling.

1122 Pressure Drop

Pressure drop from uniform fouling is usually small and not likely to cause a
hydraulic limit. If pressure drop increases rapidly, it is most likely due to a local-
ized restriction created by deposition of solids in the feed or possibly by salts
coming out of solution as process conditions change. This kind of pressure drop can
occur with little or no loss of heat transfer efficiency. However, if the pressure drop
forces reduced flowrates, heat transfer efficiency may begin to deteriorate as a
secondary effect.
Very few process units have the instruments needed to monitor individual exchanger
pressure drops continuously. Therefore, periodic field pressure surveys should be
conducted. Consider installing “dP” meters across exchangers that are repeat
offenders. See Section 521 of the Instrumentation and Control Manual for a discus-
sion of differential pressure (dP) meters.

1130 Monitoring Frequency

The ideal monitoring program is to have continuous data gathering and perfor-
mance calculation, as outlined in Figure 1100-5. However, this program is only
practical if data collection and calculations are handled by a computerized moni-
toring system.
Collecting data manually and performing hand calculations are a possible alterna-
tive but depend on the availability and commitment of manpower. To date,
Company experience with “manual” monitoring programs has not shown them to be
effective.
Fig. 1100-5 Ideal Monitoring Program
Monitor Data Required Calculate
Heat transfer Stream flowrates Basics: Q, Ucl, Us, rf
Temperatures in/out Design Parameters:
Stream properties • • • Velocities
Mechanical data • • • Flow regimes
• • • Skin temperatures
Hydraulics Pressure in/out for each Clean dP
stream

1100-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 1100 Performance Monitoring

1140 Computer Programs

The best type of monitoring system is a continuous one, made possible with moni-
toring packages, such as CHEVMON. However, to date no one has incorporated a
complete heat exchanger monitoring package into CHEVMON. There is no ques-
tion that this can be done. Anyone interested in pursuing this problem should
contact the Monitoring and Controls Division of ETD for assistance.
The HTRI programs, and HEXTRAN, are available on the mainframe computer if it
is necessary to calculate performance manually.
The HTRI programs (see Section 280) are best for evaluating the performance of
individual exchangers. HEXTRAN is the choice for studying a complete network of
exchangers, such as a crude unit preheat train. Contact the Engineering Analysis
Division for assistance in using these programs.
Heat exchanger monitoring may be available as part of the service provided by
outside organizations. Some chemical suppliers provide this today for crude units.
They collect and process the data, using their own computer tools, and compile the
results.
Monitoring is only as good as its input. Therefore, to maintain quality there should
be some degree of data checking built into continuous monitoring systems. For
instance, a difference in hot and cold streams duties should prompt an investigation
into flowmeters and/or TIs.
In “manual” monitoring systems, spreadsheets (Lotus) may be available or easily
created to assist in checking the quality of data. If you need assistance, contact the
appropriate process division within ETD Technical Services.

December 1989  1989 Chevron USA Inc. All rights reserved. 1100-7
1100 Performance Monitoring Heat Exchanger and Cooling Tower Manual

1150 Example: Evaluating a Heat Exchanger

Figure 1100-6 shows the mechanical and operating data needed to evaluate a typical
heat exchanger. The specific case chosen is a crude unit crude/vacuum residuum
exchanger.

Fig. 1100-6 Example: Evaluating a Heat Exchanger

Mechanical data should not change much with time. It is good procedure, however,
to periodically recheck the data (especially after a shutdown) in case field modifica-
tions were made.
Operating data—flows, temperatures and pressures—need to be collected for each
analysis. The fluid properties and composition may only need periodic checking.
The frequency will have to be determined for each service. If process conditions
change frequently, then new analyses for stream composition and properties must be
obtained regularly.

1100-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 1100 Performance Monitoring

Figure 1100-7 provides sample calculations using the operating data from
Figure 1100-6. The comments give some interpretation of the results; however, the
historical monitoring data are required for more definitive interpretations.

Fig. 1100-7 Sample Calculations for Evaluating a Heat Exchanger

HEAT BALANCE
BPOD  A  DT  Cp = MMBtu/hr
Tube side: 91,500  12.77  (508-488)  0.68 = 15.89
Shell side: 22,500  14.78  (650-580)  0.68 = 15.83
Comment: The quality of the data looks good because the two sides balance.
SERVICE HEAT TRANSFER COEFFICIENT, Us:
LMTD = [(650-488)-(580-508)] / Ln [(650-488)/(580-508)]
= 111F
F = 0.98 [1 shell / 2 tube pass]
Us = Q / (A  LMTD  F)
= 15,900,000 / (4000  111  0.98)
= 36.5 Btu/hrft2F
FOULING RESISTANCE, rf:
rf = (1 / Us) - (1 / Ucl)
= (1 / Us ) - [ (1 / hi) + (1 / ho) + rw]
hio = 375 Btu/hrft2F From Hand Calculations or Computer Simulation
ho = 142 Btu/hrft2F
rw = 0.0003 hrft2°F/Btu
rf = (1/36.5) - [(1/375) + (1/142) + 0.0003]
= 0.0274 - 0.0100
= 0.0174 hrft2F/Btu
Comments: Fouling represents over 60% of the total heat transfer resistance. In a well designed unit rf
should not exceed 20 to 25% of the total resistance. Check design parameters such as
velocities, flow regimes, and skin temperatures against good design practice. Look for
possible feed contaminants, misoperation, etc.
PRESSURE DROP:
Calculated Observed Observed/Calculated
Tube side: 8 10 1.2
Shell side: 5 50 10.0
Comments: The shell side pressure drop indicates that something unusual is occurring. Review the
fouling history, rf, and pressure drop history to see if the buildup has been gradual or
sudden.

December 1989  1989 Chevron USA Inc. All rights reserved. 1100-9
1100 Performance Monitoring Heat Exchanger and Cooling Tower Manual

1160 Monitoring Followup

If a monitoring program is successful, the user will be guided toward some followup
activity. Likely actions include:
1. Schedule heat exchangers that warrant cleaning. (Also identify those that do not
warrant cleaning.)
2. Redesign exchangers to better fit current operation and thereby improve perfor-
mance.
3. Modify process conditions to alleviate fouling or to take better advantage of
available hardware.
4. Modify chemical treatment program to alleviate fouling.
5. Improve instrumentation if monitoring results are inconsistent.
6. Modify process control if performance is erratic.
7. Modify the monitoring program itself to reflect process and/or equipment
changes.

1100-10  1989 Chevron USA Inc. All rights reserved. December 1989
2100 General Information

Abstract
This section discusses types of cooling water systems, types of cooling towers and
cooling tower components and provides a list of contacts and references for more
information on cooling water systems.
The Cooling Tower Institute (CTI) has developed the industry standards for cooling
towers. The commonly used standards are referenced in Specification
EXH-MS-1317, included in this manual. Copies of the CTI manual or the CTI spec-
ifications can be obtained from:
Cooling Tower Institute
P.O. Box 73383
Houston, TX 77273
(713) 583-4087

Contents Page
2110 Cooling Water vs. Air Coolers 2100-3
2120 Types of Cooling Water Systems 2100-3
2121 Open Recirculating Cooling Water System
2122 Closed Loop Cooling Water System
2123 Tempered Cooling Water System
2130 Types of Cooling Towers 2100-5
2131 Natural Draft Towers
2132 Forced Draft Towers
2133 Induced, Mechanical Draft Cooling Towers
2140 Cooling Tower Components 2100-7
2141 Fills
2142 Drift Eliminators
2143 Cooling Tower Basin

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-1
2100 General Information Heat Exchanger and Cooling Tower Manual

2150 Contacts and References for Cooling Towers 2100-12

2151 Contacts: Cooling Tower Manufacturers and Specialists
2152 References
Note Figures 2100-6 through 2100-10 are 11  17 foldouts at the end of the
section.

2100-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

2110 Cooling Water vs. Air Coolers

Typically, Chevron uses an open, recirculating cooling water system in order to cool
fluids in the plant to temperatures lower than can be attained with air coolers. When
environmental concerns and water resource limitations do not control, cooling water
can also be more attractive than air cooling for many other reasons.
Section 100 of this manual contains more detailed information comparing water to
air coolers.

2120 Types of Cooling Water Systems

Like steam, power, and instrument air, cooling water is a critical “utility.” It is
required for all operating conditions, including startups, shutdowns, and upset
conditions. Cooling towers are typically large wooden structures that use circu-
lating ambient air to cool warm water. In a cooling tower, normally referred to as
just the “tower,” the hot water is cooled by exchange of its sensible and latent heat
with relatively cool air.
About 80% of the cooling occurs from evaporation of a small portion of the water as
the air flowing through the tower contacts water cascading from the top to the
bottom of the tower. Most of the Company’s cooling towers use a recirculating
water system to conserve water resources and minimize operating costs.
There are three types of cooling water systems.
1. Open, recirculating cooling water system
2. Closed loop cooling water system
3. Tempered cooling water system

2121 Open Recirculating Cooling Water System

Figure 2100-6 shows a typical open recirculating cooling water system using a
cooling tower.
During normal operation of a cooling tower, all available cells should be used. Each
fan in the cell normally contains two-speed motor drivers. The cooling tower basin
and pump forebay serve as system storage and are sized for a minimum of
15 minutes of water storage at the normal circulating rate. Makeup water is based on
the concentration in the circulating water. The cooling tower blowdown can ideally
be flow controlled by a conductivity analyzer.
The system circulating pumps are usually horizontal, centrifugal, double-suction,
and single-stage. Pump discharge pressure is in the range of 55 to 65 psig. There are
normally three or four pumps, with at least two of them being turbine-driven to
assure 100% water circulation rate during a power failure. Under normal circum-
stances, the electric motor-driven pump and one turbine-driven pump operate while
the other remaining turbine-driven pumps are on automatic standby. The pumps take
suction from the forebay and discharge to a common header. A common minimum

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-3
2100 General Information Heat Exchanger and Cooling Tower Manual

flow recirculation line, directly back to the tower, is provided at the pump dis-
charges for pump protection and initial prestartup conditioning.
Chemicals are used to control scaling, corrosion, and fouling. The most important
chemical is the corrosion inhibitor. Chemicals and components of the water treating
package are discussed in Section 2400.

Typical Process Cooling Water System

Cooling water circulation is controlled by starting and stopping the standby cooling
water circulation pumps based on pressure in the discharge header. There is no flow
or temperature control on the circulating stream. A bypass recycles water to the
cooling tower basin for low flow pump protection. Makeup to the cooling tower is
on level control. Temperature, flow, and pressure of the cooling water to the plants
and other users is monitored, frequently, in the control room. Return cooling water
at the cooling tower inlet is sampled and analyzed for conductivity, chlorine, pH,
and corrosion inhibitor. Output from the conductivity analyzer resets a flow
controller on the cooling tower blowdown line. Blowdown flow is also used to regu-
late the addition of dispersant and corrosion inhibitor to the system. Output from the
other analyzers regulates the addition of chlorine, acid or caustic, and corrosion
inhibitor.

2122 Closed Loop Cooling Water System

Figure 2100-7 is a process flow diagram for a typical closed loop cooling water
system.
“Open” systems, as described in the previous section, typically cool 100°F to 125°F
hot returning water to about 5°F to 10°F above the ambient wet bulb temperature at
the site. A “closed” system supplies a flow of high purity, inhibited cooling water at
around 100°F. Its function is to cool bearings, jackets, etc., of rotating equipment in
the plant, where the quality of the “open” cooling water is unacceptable for the ser-
vice. The heat from a closed system is removed by heat exchangers using the open
cooling water system. This system typically uses demineralized water for initial fill
and subsequent makeup. It has its own pumping, minimum flow recirculation line,
surge tanks, and chemical feed systems. Pump discharge pressure is normally about
the same as in an open cooling water circulating system, 55 to 65 psig. Corrosion
inhibitor is added to the system upstream of the circulating water pumps.
Closed loop systems should include two surge tanks, each sized for 5 minutes of
water storage at total system design flow. They should be nitrogen blanketed to
minimize the oxygen content in the water. Changes in fluid volume, temperature,
and chemical balance should be monitored at the tanks. Under normal conditions,
both tanks should be online, but either tank can be isolated for maintenance.

2100-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

2123 Tempered Cooling Water System

Figure 2100-8 is a process flow diagram for a typical tempered cooling water
system.
A tempered cooling water system is another closed loop system circulating high
purity inhibited cooling water, normally at a minimum of 140°F to process heat
exchangers that cool streams with high pour points. Pump discharge pressures are
about 100 to 115 psig for these systems. This system also includes two surge tanks,
circulating pumps, and a common minimum flow recirculating line. The system
should use demineralized or high quality water for initial fill and subsequent
makeup. Corrosion inhibitor is added intermittently to the system upstream of the
circulating pumps. Since the water is only cooled to 140°F, induced draft air cooled
heat exchangers remove process heat from the system. The temperature is main-
tained by bypassing hot return water around the air coolers.

2130 Types of Cooling Towers

The oldest means of cooling and storing circulating water was to discharge the hot
water into a pond of sufficient area so that the water was cooled by the air passing
over the surface of the pond. Spraying the heated water into the air over the pond
increased the rate of cooling and reduced the area to around 5% of that required
without sprays. Because of this large space requirement and the windage losses with
spray ponds, the cooling tower was developed. A modern mechanical draft cooling
tower requires about 2% of the area of a spray pond, and less than 0.01% of the area
for a cooling pond or reservoir. Cooling towers come in all sizes. Some are small
enough to be portable. Others take up more plot space than one of the plants they
serve, with capacities of 100,000 GPM and duties of up to 1 billion Btu/hr.
For industry, cooling towers are normally rectangular, wood, or metal structures.
The hot water is delivered to the top and falls through baffling-type fill into the
basin below. Air enters at the bottom or side of the tower, cooling the water by
convection and partial evaporation. The circulation of the air through the tower may
be either by natural draft, or by a forced or induced draft fan.

2131 Natural Draft Towers

Chevron still has a few natural draft towers in our producing and pipeline centers.
The air circulates through a natural draft cooling tower because of the temperature
difference between the air inside and outside of the tower. Natural draft towers have
been very popular with utilities, who can allow much longer payouts and frequently
use the natural draft principle in their large hyperbolic towers. Hyperbolic towers
are not used by the oil industry and therefore are not discussed in this manual.

2132 Forced Draft Towers

Chevron also has a few forced-draft cooling towers. Forced-draft towers are usually
smaller than induced-draft towers and have either centrifugal fans located at the
base of the tower (which is constructed as a plenum to provide positive pressure
airflow through the fill material) or axial fans on the side of the tower.

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-5
2100 General Information Heat Exchanger and Cooling Tower Manual

For the past 30 years, induced-draft cooling towers have been purchased instead of
forced draft cooling towers (for capacities greater than 2,000 GPM/cell). Forced-
draft cooling towers have the following problems:
1. High air velocities through the fill and difficulty with promoting good air distri-
bution across the tower.
2. Greater difficulty with recirculation.
3. Lower efficiency and more plot plan space.
As a result of these problems, the Company normally purchases induced-draft
cooling towers.

2133 Induced, Mechanical Draft Cooling Towers

Induced-draft towers pull air into the tower. The induced draft is provided by a
propeller-type axial fan located in the stack at the top of the tower.

Counterflow Design
See Figure 2100-1. The induced-draft fan pulls the air into the inlet louvers at the
bottom of the tower and up through the tower. In the counterflow tower, the
returning hot water is piped to a distribution system of headers and lateral piping
connected to pressure flow nozzles, which are located below the top of the tower
just below the mist eliminators. The nozzles spray the water as droplets that then fall
across the fill which acts as baffles to allow surface contact between the droplets
and the rising air. The primary advantage of the counterflow tower is its efficiency.
The coldest water contacts the driest air, and the warmest water contacts the moist
humid air.

Fig. 2100-1 Counterflow Induced-Draft Tower

2100-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

Crossflow Design
See Figure 2100-2. The induced-draft fan pulls the air into the inlet louvers placed
along the tower’s sides, spanning its entire length and height. Air is introduced
perpendicular to the falling water. The crossflow tower uses risers to pipe the
returning hot water to the top of the tower where it is discharged to open gravity
distribution decks adjacent to the shroud protecting the fan. The floor of this deck
contains gravity flow nozzles, and the water level in the deck controls the rate of
water flow onto the fill.
Fig. 2100-2 Crossflow Induced-Draft Cooling Tower

The water falls through orifices in the nozzles, and as it flows it is distributed across
the fill. The fill acts as baffles to allow surface contact between the droplets and the
crossflowing air. Mist eliminators, forming a “V” at the center of the tower, mini-
mize drift and windage losses as the air is pulled up to the fans in the center of the
tower.
Since the early 1970s, counterflow towers have used a lower pressure spray system,
coupled with large plenum areas. In most cases, this feature has made counterflow
towers preferable to crossflow towers because they are less expensive to install and
operate than crossflow towers. Crossflow towers are often specified when a low
flow rate or potentially contaminated air atmosphere is involved.

2140 Cooling Tower Components

To become familiar with the components of cooling towers, refer to the following
figures for details of the standard components for the two main types of cooling
towers that most petroleum plants use:
1. Figure 2100-9, Counterflow Cooling Tower: Perspective—Typical Parts and
Framing
2. Figure 2100-10, Crossflow Cooling Tower: Transverse Elevation—Typical
Parts and Framing
See Appendix I for a glossary of commonly used terms (CTI Bulletin 109).

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-7
2100 General Information Heat Exchanger and Cooling Tower Manual

2141 Fills
The two most significant decisions when purchasing a new cooling tower are:
1. Should we buy counterflow or crossflow?
2. Should we buy “splash” fill or “film” fill?
Section 2130 above compares the relative merits of counterflow and crossflow
towers. This section discusses splash fill and film fill.
If there is no plugging problem (i.e., good water treating) and no serious hydro-
carbon attack, the counterflow tower with polyvinyl chloride (PVC) film fill usually
will have the lowest overall cost (installed cost + fan power cost + pumping cost.)
Most large crossflow towers use splash fill (normally a PVC material). However,
some companies have small crossflow towers which stack thin sheets of plastic fill
on very close centers and can be considered film fill.
Figures 2100-3 and 2100-4 illustrate the two basic types of cooling tower fill. Fill
material is used to maintain an even distribution of water across the horizontal plane
of the tower and to create as much water surface area as practical to enhance evapo-
ration and sensible heat exchange. The water/air interface ratio can be improved by
creating either a large number of droplets or many thin vertical sheets of water. Fill
materials are commercially manufactured from wood, PVC, polystyrene, cellulose,
and before environmental constraints, from asbestos cement board.
Fig. 2100-3 Splash Fill

Splash Fill
For years splash fill was the standard fill for cooling towers. Splash fill is
constructed of successive layers of staggered impact surfaces. Small droplets are
formed as the warm water falls through the fill and splashes off each layer.
Typical splash fill consists of redwood battens in thicknesses of 3/8 inch to 1 inch
and 1.5-inch to 2-inches deep, installed vertically on the narrow edge. These battens
are on 0.25-inch to 2.5-inch centerline and form a deck that is usually installed hori-

2100-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

zontally in the cooling tower on from 12- to 24-inch vertical centers. Each row in
elevation is rotated 90 degrees to the row above it. The battens are 6 feet long. The
decks formed are 3 feet by 6 feet and two of them fill the 6-foot cross-sectional area
between the columns in the tower. Typically, there are about 20 decks in a tower.
Where conditions permit, these splash decks are also made in egg-crate type injec-
tion molded polypropylene and PVC. They have the following advantages over film
fill discussed below:
1. There is no plugging as the spacing is far more open than in a film fill.
2. Inspection of every component of the tower is greatly simplified by moving
sections of the deck out a piece at a time, and by being able to move easily
down the inside of the tower.

Film Fill
Film fill came into use in cooling towers about 15 years ago (see Figure 2100-4). It
has a honeycomb configuration and is usually a PVC material. This fill spreads the
water droplets into thin sections throughout the cells of the fill, thereby cooling a
large surface area for the same energy. If the water is always clean, it behaves as
predicted. As film fill is more efficient than splash fill, it takes up much less volume
in the tower. Two to 3 feet of film will provide the equivalent surface of 20 to
30 feet of wood splash fill, making the return header and the elevation of the tower
much lower. This results in lower head required and lower pumping costs.
Fig. 2100-4 Film Fill

Cooling tower support posts are normally on 6-foot centers. Typical film fill is
manufactured in 3-foot by 6-foot cross-sectional areas with layer thicknesses of 6,
12, and 18 inches. Total depth in the tower typically ranges from 1 foot to 6 feet. As
an example, where icing could be a problem, a 6-foot thickness could have a 1-foot
clear elevation at each 2-foot interval to allow for extra heavy supports under each
2 feet of fill. Film fill is manufactured with various size flutes (openings) and sheet
thicknesses.

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-9
2100 General Information Heat Exchanger and Cooling Tower Manual

Film fill is most commonly used in counterflow towers and is normally installed
just above the intake louver opening. In a retrofit it is usually installed just below
the existing water distribution system. Good water distribution is very important
with film fill. The use of film fill is not recommended where there is potential for
hydrocarbon leakage into the fill, particularly waxy or heavy hydrocarbon. A tower
has collapsed due to hydrocarbon contaminating the middle 12 inches of a 30-inch
fill. The “mushy” condition in the middle could not be observed from the top or
bottom. Biofouling is another serious consideration for film fill. Bacterial growth is
most rapid at 98.6°F. For many of our towers, this is about half way through the film
fill. Bacteria thrive on hydrocarbon, ammonia, nitrates, sulfates, etc. See
Section 2440 for more information on biological fouling and control.

2142 Drift Eliminators

Drift is water droplets which are entrained in the air stream as it passes through the
tower. These water droplets contain dissolved and suspended solids in proportion to
their concentration in the circulating water. Drift eliminators are baffles that cause
the hot air with entrained water droplets to change direction a number of times. This
process causes the droplets to hit the eliminator surface at every change of direction
and fall back into the tower. The efficiency of a drift eliminator is a function of its
design. Figure 2100-5 shows a sketch of the three major drift eliminator designs:
1. Herringbone, the least efficient.
2. Waveform, of intermediate efficiency.
3. Cellular, the most efficient.
Drift eliminators are constructed of wood, PVC, polystyrene, or cellulose. PVC now
dominates this group for the following reasons:
1. The returning water does not pass through them.
2. There is no sunlight to develop bacteria.
The drift eliminators fit as a series inside 3-foot wide sections with end flanges. For
a counterflow cooling tower, the drift eliminators are located 2 to 4 feet below the
fan deck across the entire cross-section of the tower. For a crossflow tower, the drift
eliminators are located in the middle one-third section of the tower. See
Figures 2100-1 and 2100-2.

2100-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

Fig. 2100-5 Different Types of Drift Eliminators

2143 Cooling Tower Basin

The cooling tower basin should, as a minimum, be designed to provide:
1. Sufficient volume to allow 3 to 5 hours of operation if makeup water is lost.
2. A specific volume of fire water in case the regular water system fails.
3. A volume of cold water to allow a short period of operation below a specified
temperature in case of power failure. For example, the basin may allow
10 minutes operation before tower outlet temperature rises above 95°F,
assuming no convection effect.
4. The necessary head for proper operation of the cooling water circulating
pumps.
Cooling tower forebay design is covered in more detail in Section 2240.

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-11
2100 General Information Heat Exchanger and Cooling Tower Manual

2150 Contacts and References for Cooling Towers

2151 Contacts: Cooling Tower Manufacturers and Specialists

Included in this section is a list of cooling tower manufacturers, specialists, and
resources for maintenance and repair.
Cooling Tower Manufacturers
Marley
P. O. Box 2912
Mission, Kansas, 66201
913/362-1818
Custodis-Ecodyne
P. O. Box 1267
Santa Rosa, Ca., 94502
707/544-5833
Hamon
245 US Highway 22 West
Bridgewater, N. J., 08807
201/725-3311
BAC Pritchard
P. O. Box 7322
Baltimore, Maryland, 21227
301/799-6312
Contact the Utility Systems and Energy Management Section, ETD, for informa-
tion and recommended contacts and specialists in these areas: (1) cooling towers,
(2) water treating, (3) upgrades and repairs

2152 References
Industry Specifications:
Cooling Tower Institute, Inc.
P. O. Box 73383
Houston, Texas 77273
(713) 583-4087
Company Documents:
Johnsen, C. W., “FCC Cooling Tower Electronic Vibration Switches,” 1/31/89. IMI,
Richmond Refinery.

2100-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2100 General Information

Outside Documents:
1. Hydraulic Institute Standards for Centrifugal, Rotary & Reciprocating Pumps,
14th Edition, Hydraulic Institute, 1983.
2. Nystrom, James B., et al., “Modeling Flow Characteristics of Reactor Sumps,”
Journal of the Energy Division, ASCE, Vol. 108, No. EY3, November 1982.
3. Padmanabhan, M., and G. E. Hecker, “Scale Effects on Pump Sump Models,”
Journal of Hydraulic Engineering, ASCE, Vol. 110, No. 11, November 1984.
4. Prosser, M. J., The Hydraulic Design of Pump Sumps and Intakes, British
Hydromechanics Research Association/Construction Industry Research and
Information Association, 1980.
5. Sweeney, Charles E., et al., “Pump Sump Design Experience: Summary,”
Journal of the Hydraulics Division, ASCE, Vol. 108, No. HY3, March 1982.

December 1989  1989 Chevron USA Inc. All rights reserved. 2100-13
2100 General Information Heat Exchanger and Cooling Tower Manual

2100-14  1989 Chevron USA Inc. All rights reserved. December 1989
2200 Cooling Tower Design Guidelines

Abstract
This section discusses key cooling tower design parameters, electrical facility instal-
lation, environment/safety/fire protection considerations, and forebay design.

Contents Page
2210 Key Parameters 2200-2
2211 Heat Load (Duty)
2212 Circulating Water Rate (GPM)
2213 Wet Bulb Temperatures
2214 Optimizing Cooling Tower Costs
2215 Makeup Water
2216 Blowdown and Cycles of Concentration
2220 Electrical Installations 2200-9
2221 Area Classification
2222 Materials
2223 Installation
2230 Environmental/Safety/Fire Protection Considerations 2200-10
2231 Effluent Quality
2232 Air Quality
2233 Safety
2234 Fire Protection
2240 Cooling Tower Forebay Design 2200-15
2241 General Information
2242 Forebay Design
2243 Hydraulic Model Testing
2244 Standard Drawings
2245 References
Revision History 2200-27

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-1
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

2210 Key Parameters

This section discusses the key design parameters that must be considered when
purchasing or rating a cooling tower. The actual rating procedure is in Section 2300.

2211 Heat Load (Duty)

The tower duty is calculated using the following equation:

Duty Q MMBH = mCp  (Th - Tc)

(Eq. 2200-1)
where:
m = Circulation water flow in pounds per hour.
Cp = Specific heat in Btu/lbF
Th = Hot water to the tower, F
Tc = Cold water from the cooling tower basin, F
Converting pounds per hour to gallons per minute and using a Cp of 1,
Q (MMBH) = 500  GPM  (Th - Tc)
The 500 comes from converting Item 1 from GPM to lb/hr: 
(8.33 lb/gal  60 min/hr) = 500.
The calculated heat load is usually increased by a factor of 10 to 20% to obtain the
design heat load.

2212 Circulating Water Rate (GPM)

Conversely, if we have the duty and we want to find the circulating water rate
assuming a temperature range:

Q
GPM = -------------------------------
-
500  T h – T c 
(Eq. 2200-2)
The circulation rate and temperatures are developed by looking at:
1. All the heat exchanger duties in the cooling tower network.
2. The cooling water flow rates and temperatures to satisfy the design conditions
for the heat exchangers.
By summing all the duties of the heat exchangers in the network and taking the
weighted averages of all the inlet and outlet temperatures of the circulating water in
GPM, Th and Tc can be determined. For each circulating water rate there is a unique
hot and cold water temperature combination.

2200-2  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

2213 Wet Bulb Temperatures

Determining the design wet bulb temperature is an important decision, as invest-
ment costs are involved. Figure 2200-1 lists the ambient design wet bulb tempera-
tures at a number of our operating centers.

Fig. 2200-1 Design Wet Bulb Temperatures at Several Company Locations

Location Design Wet Bulb F
Anchorage, Alaska 59
Bahamas, Freeport 79
Cedar Bayou (Bayport, Texas) 82
El Paso, Texas 70
El Segundo, California 70
Hawaii 73
Kaybob 61
Marietta, Ohio 77
Mt. Belvieu (Bayport, Texas) 82
Orange, Texas 80
Pascagoula, Mississippi 79
Philadelphia, Pennsylvania 76
Port Arthur, Texas 82
Richmond, California 65
Salt Lake, Utah 65
St. James, Louisiana 80
St. John, N. B. 65
Vancouver (Burnaby) 68

Considerations for Design Wet Bulb

1. Cooling towers should be oriented so that the longitudinal axis is aligned with
(parallel to) the prevailing wind. If the plot plan will not accommodate this
orientation, the wet bulb temperature shown in Figure 2200-1 may need to be
increased by 1F.
2. Cooling tower performance can be measurably affected by external influences
on the wet bulb temperature of the air entering the tower. Examples of this are
localized heat sources situated upwind, drift from adjacent cooling towers,
recirculation of exit air caused by large structures adjacent to the tower, etc. For
more information on recirculation, request a copy of CTI Bulletins PFM-110
and PFM-116. The external influences discussed here should be evaluated, and
if appropriate, shown wet bulb design temperatures may need to be raised an
additional 2F.

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-3
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

If a cooling tower is being located where the Company has no experience, the
design wet bulb temperature should be obtained from the local weather bureau or
local airports. Industry’s normal practice is to use the wet bulb temperature at the
5% level. This is the temperature that the wet bulb will be below over 95% of the
time during the summer months.

2214 Optimizing Cooling Tower Costs

For a given heat duty and design wet bulb temperature, you can use the following
three parameters to optimize the cooling tower cost.
1. The temperature of the water returning to the tower.
2. The range—the difference in temperature between the hot water returning to
the tower and the cold water from the cooling tower basin. (Cooling ranges
normally fall between the limits shown in Figure 2200-2.)
3. The approach—the difference in temperature between the cold water from the
cooling tower basin and the ambient wet bulb temperature.

Fig. 2200-2 Acceptable Cooling Tower Temperature Range for Different Types of Plants
Type of Plant Range, F
Refineries 25-45
Power Plant Steam Condensing 10-25
Chemical Processes 15-25
Air Conditioning/Refrigeration 5-10

Tower Size Factor

The tower size factor is an empirical way of comparing various combinations of the
parameters discussed above. Figure 2200-3 plots the “Tower Size Factor” for
assumed returned water temperatures, known wet bulb temperatures, and resultant
ranges and approaches. The return temperature, range and approach that satisfy the
process and project limitations and result in the lowest “Tower Size Factor” will
also result in the lowest cooling tower costs.
Example:
Assume this is Hawaii, with a temperature of water back to the tower of 118F and a
wet bulb temperature of 73F (118  73 = 45). Move vertically up the chart at 45 to
the range of 35F, or an approach of 10F, which is consistent mathematically. Move
horizontally to the left to the design wet bulb temperature; then move down to the
left, following the curves to the “Tower Size Factor.” For our example, the Tower
Size Factor is about 0.93.

2200-4  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines
Fig. 2200-3 Tower Size Factor
December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-5
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Determine If the Tower Meets Design Requirements

It is easy to determine if the tower meets design requirements because of the effort
the CTI has put into resolving past problems that cooling tower manufacturers have
had with their completed towers meeting design criteria.
Our Specification EXH-EG-1317 itemizes the following as the sole responsibility of
the vendor:
1. Meet the operating conditions of the Data Sheet (EXH-DS-1317—CTI Bid
Form).
2. Be certain that the tower is a CTI code tower.
3. Meet all applicable codes and ordinances.
In addition to these requirements, the purchase order should require the manufac-
turer to supply the appropriate data so that a CTI Acceptance Test under ATC-105
can be performed (with appropriate equations to financially penalize the manufac-
turer if the tower does not meet “design.”)

2215 Makeup Water

Water losses (and consequently makeup water rate) from a cooling tower are the
sum of:
1. Evaporation. The cooling tower “cools,” mainly by evaporation. To approxi-
mate this loss, use 1% of the circulation rate for each 10 degrees Fahrenheit of
cooling.
2. Drift. This is the water that leaves the tower with the air. In the past the
maximum drift was specified at 0.2% of the water circulated. With modern
advances in drift elimination, this has been significantly reduced. For towers
purchased in early 1989 we have been receiving guarantees of 0.008% of the
circulation rate for drift loss. This loss carries the impurities that are in the
water and the chemicals added in the water treatment program. See
Section 2230 for the environmental concerns for drift.
The rate of water through the fill material (“Water Loading”) for most of our
towers is about 4 GPM/ft2. Drift is not dependent on water loading. Increasing
air velocity does result in greater drift. Typical air flows in cooling towers are
300 to 700 ft/min. Velocities in the stack are in the range of 1500 to
2000 ft/min.
3. Blowdown. This is the one water loss of the three that is adjustable, once the
tower is running. It controls the “cycles of concentration.”

2200-6  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

2216 Blowdown and Cycles of Concentration

Blowdown from a circulating water system is necessary to prevent scale-forming
compounds from exceeding their respective solubilities. If water is not removed
from the system, the dissolved solids present in the make-up will concentrate and
deposition will take place. A high total dissolved solids (TDS) level also increases
the system corrosiveness. On the other hand, from an economic viewpoint, it is
desirable to minimize blowdown in order to minimize water usage. Cycles of
concentration is the term employed to indicate the degree of concentration of the
circulating water as compared to the makeup. For example, two cycles of concentra-
tion indicate the circulation water has twice the solids concentration of the makeup
water. Cycles are usually based on concentration of chloride (where water is not
chlorinated) or magnesium and sodium ions (because they almost never precipitate
under operating conditions). The chemical suppliers can also run soluble calcium
concentration to determine cycles.

Blowdown Equations
Blowdown rates from a circulating water system can be calculated using the
following equations:
Mu = E + Bd + W = E  C/(C-1)
(Eq. 2200-3)
C = E + Bd + W/Bd = Stw/Smu
(Eq. 2200-4)
Bd = E / (C - 1)
(Eq. 2200-5)
where:
Mu = Makeup, GPM
E = Evaporation loss, GPM
Bd = Blowdown, GPM
W = Drift loss, GPM
C = Cycles of concentration (defined below)
Stw = Solids concentration in tower water
Smu = Solids concentration in makeup water
For each unit of total dissolved solids (TDS) added with the makeup, one unit of
TDS must be removed as blowdown. We have:
Smu  Mu = Stw  Bd
(Eq. 2200-6)
or
Stw/Smu = Mu/Bd = C
(Eq. 2200-7)

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-7
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Example of Blowdown and Cycles of Concentration Calculations

Given: Circulation rate = 13,000 GPM
Delta T = 120F - 85F = 35F
Cycles of concentration = 5

Calculations:
E = 13,000  (0.1/10)  35 = 455 GPM
(Eq. 2200-8)
Note Evaporation is usually 1% of circulation rate for each 10F change across
the tower.

CE 5  455
M u = ------------- = ------------------ = 569 GPM
C–1 4
(Eq. 2200-9)

Bd = Mu - E = 569 - 455 = 114 GPM.

Assuming W = 0.
(Eq. 2200-10)
Figure 2200-4 shows the reduction of blowdown for the example above with
increased cycles of concentration. The law of diminishing returns starts to apply at
the higher cycles. However, minimizing the blowdown is very desirable in a zero
effluent discharge location. Blowdown can also be expressed as a percent of the
makeup flow rate. In this example,
% Bd = (114/569)  100 = 20%
(Eq. 2200-11)

Sizing Acid and Inhibitor Systems

The above equations can also be used when sizing inhibitor and sulfuric acid
pumps. In both cases, it is necessary to know the makeup water rate to the system.
This rate, together with cycles of concentration, is used to calculate the inhibitor and
acid consumption. For calculation purposes, the amount of corrosion inhibitor
required to be added to the makeup water is the total inhibitor level desired in the
system divided by the cycles of concentration. For example, if 50 ppm are recom-
mended for the circulating water, then 10 ppm are added to the makeup water if the
system is cycled five times.
Multiplying this makeup dosage in ppm by the millions of pounds of makeup per
day will result in the pounds of inhibitor requirements. In the above example, the
daily makeup rate is 569 gallons per minute or 6.8 million pounds/day. Multiplying
this by 10 ppm, the daily inhibitor requirement amounts to 68 pounds.

2200-8  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

Fig. 2200-4 Example: Blowdown vs. Cycles of Concentration

2220 Electrical Installations

Cooling towers present special problems for the installation of electrical facilities.
Moist, corrosive conditions normally exist; hence, moisture-and corrosion-resistant
materials are required. In addition, because flammable gases or vapors may be
present under some conditions, equipment suitable for the appropriate hazardous
area classification is required.
Standard Drawing GD-P1011 shows the typical area classification requirements and
installation details and lists recommended materials.

2221 Area Classification

Leaks in water-cooled heat exchangers will normally result in leakage of process
fluid into the cooling water. If the process fluid is a gas or a hydrocarbon liquid with
a flash point lower than the cooling water temperature, gas or vapor will be released
from the cooling water at the tower. In case of a tube rupture in a high-pressure gas
heat exchanger, large quantities of gas will be entrained in the water. This gas may
cause pressure surges in the cooling water return line that may rupture the cooling
water piping on the tower. Thus, it is possible for flammable gases or vapors to be
released at the cooling tower, sometimes in large quantities.
However, an abnormal condition involving equipment failure must exist—i.e., a
leak in a heat exchanger—in order for flammable gases or vapors to be present at a
cooling tower. Thus, the appropriate classification is Class I, Division 2.

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-9
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

2222 Materials
Because of the corrosion problem, aluminum conduits and fittings should be used.
Electrical equipment enclosures should be aluminum or corrosion-resistant mate-
rials. For corrosion resistance, all aluminum materials should have a copper content
of less than 0.4%.
Typical Class I, Division 2, wiring methods should be used. Conduits should be of
rigid metal with threaded connections. Fittings should have threaded hubs and cast
gasketed covers. Push buttons should be explosion proof, and vibration switches
should be hermetically sealed (mercury type) in cast enclosures, or explosion proof.
Receptacles should be explosion proof, of the arc-tight type designed so that arcs
will be confined within the case of the receptacle. Lights should be enclosed and
gasketed. Conduit seals should be provided as normally required in classified areas.

2223 Installation
Installation details shown on Standard Drawing GD-P1011 should be used. Wher-
ever practical, conduits should be routed on the exterior of the tower. However, the
conduit may be run below the upper deck if required. Conduit runs across the upper
surface of the deck can be ramped over. In all cases, the conduits should be
routed away from any cooling water piping that might move during upset
conditions and cause damage to conduits and fillings.

2230 Environmental/Safety/Fire Protection Considerations

2231 Effluent Quality

Chromate vs. Nonchromate Corrosion Inhibitors
Environmental regulations are forcing drastic limitations on or elimination of the
chromium in waste water. The National Pollution Discharge Elimination System
(NPDES) and the Environmental Protection Agency (EPA) limit the discharge of
total and hexavalent chromium from our process plants.
Cooling tower blowdown constitutes a large portion of a typical plant’s waste water.
The alternatives are either chromium removal from cooling tower blowdown or the
use of an alternative ultra-low or nonchromate treatment. Chromate
removal/recovery equipment on cooling tower blowdown streams is usually more
expensive than nonchromate inhibition. However, automatic control of chemical
concentrations and an excellent microbiological program are a must for a nonchro-
mate program to perform successfully.
Nonchromate treatments can be expected to reduce corrosion on mild steel only
down into the range of 3 to 5 mils per year. Even with higher corrosion rates, the
cost of nonchromate treatments run from 1.5 to 2.0 times the cost of a chromate-
based treatment program.

2200-10  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

The selection of the proper corrosion inhibitor should be made by the process plant
on an individual basis based on economics and operational reliability. Section 2400
and Appendix J give guidelines on the various corrosion inhibitor systems.

Minimizing Blowdown
Minimizing blowdown makes sense from both an economic and environmental
standpoint. Depending on the location, makeup water costs can range from 40 cents
to $4.00 per 1000 gallons.
Normally, the plant effluent systems are capable of handling cooling tower blow-
down streams. However, if large volumes of cooling tower blowdown must be
disposed of and the blowdown contains high levels of total dissolved solids (TDS)
and metal-based water treating chemicals, this practice may be unsatisfactory.
Possible future Best Available Technology (BAT) Effluent Regulations may also
require a reduction in effluent flow rate. For these reasons, methods of minimizing
cooling tower blowdown are being investigated.
Typically, cooling tower blowdown is composed of less than 0.5% by weight of
dissolved solids. The cost of disposal by such means as solar ponds, evaporation
plants, and deep well injections depends on the volume discharged. Other blow-
down treating methods, such as chrome removal processes (which the Company has
not used to date) are also dependent on the volume. Therefore, every effort should
be made to minimize the amount of water going to ultimate disposal. Other special
processes are side stream softening or side stream softening combined with an elec-
trodialysis or reverse osmosis unit. The clean effluent from these processes can be
recycled to the tower to reduce the amount of cooling tower blowdown.
Blowdown is discussed in detail in Section 2216 and Section 2422.

Use of Biocides
In some areas, effluent must meet fish toxicity requirements. Biocides can be toxic
to fish and must be used with care. They should be chosen so that a minimum
amount is used with a maximum potential for degradation in the effluent system.
Biocides may also have an adverse effect on the water treatment systems. A rough
indication of this can be obtained by comparing the biological oxygen demand
(BOD) for a sample of normal effluent water and a sample of effluent containing
biocide at the concentration expected in the effluent. A low BOD result in the pres-
ence of biocide indicates a potential toxicity problem. These tests should be
conducted before a new biocide is used.

Impounds Around Chemical Areas

As discussed in Section 2530, all chemical injection facilities should be contained
by berms. The impoundage should be large enough to hold the contents of the
largest container in case of a rupture.

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-11
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

2232 Air Quality

Drift
The drift off the cooling tower contains solids and other additives in proportion to
the level of solids and additives in the recirculating water. The most significant
contaminant is hexavalent chromium (Cr+6) if it is being used as a corrosion inhib-
itor. Hexavalent chromium emissions can be controlled by:
1. Limiting the average chromate concentration in the recirculating water (pres-
ently 13 ppm maximum in the petroleum and chemical industries).
2. Eliminating chromate-based chemical completely from the water treating
programs.
3. Retrofitting towers with higher efficiency drift eliminators.
4. A combination of 1 and 3 above.

Minimizing Drift
Manufacturers claim they can guarantee drift rates from 0.02% down to 0.001% of
the recirculation rate. To achieve the lower drift numbers requires some additional
investment and 3% to 5% added fan horsepower. These low numbers are difficult to
measure. The measuring techniques vary and several different sampling train
configurations have been developed. The drift rates have not given consistent
results.

2233 Safety
Chemical Handling
The safety considerations for handling water treatment chemicals and chlorine are
discussed in Section 2500.

Wood Deterioration
Wood deterioration in platforms and stairs has been a problem. Decay organisms
also affect the nonwater wetted areas of the cooling tower. All cooling towers
should be inspected regularly for any signs of cracking or deterioration. This is
particularly critical for towers where pressure-treated Douglas fir and non-heart-
wood redwood are the principal materials of construction. These two types of wood
have a history of deterioration and therefore higher maintenance costs.

Fan Vibration
Excessive fan or gearbox vibration has caused many fan failures. Obviously, this
can be a significant personnel hazard. The primary purpose of cooling tower vibra-
tion switches is to detect high fan/gearbox vibration and shut down the fan motor
before a failure occurs. A secondary purpose of the switch is to allow surveillance
of machine condition in operation so that failures can be predicted ahead of time
and preventive maintenance performed. While mechanical switches have proven

2200-12  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

inadequate in meeting the primary purpose and incapable of providing the second
purpose, electronic monitor/switches can meet both requirements.
Mechanical vs. Electronic Switches. After tests in 1987 comparing the commonly
used mechanical switch (Metrex 5175-01) and an electronic switch (PMC Beta
Model 440), Richmond Refinery is now recommending the use of electronic
switches for cooling tower fans. For more information on this testing, please contact
the Richmond Refinery IMI group and request the 1/31/89 report entitled “FCC
Cooling Tower Electronic Vibration Switches.”
Previously, cooling tower fans at Richmond Refinery have been equipped with
mechanical vibration switches (Metrix Model 5175-01 or Robertshaw Model 365
Vibraswitch). Recent experience has shown these mechanical switches provide
inadequate protection against catastrophic failures of cooling tower fans. Alterna-
tively, electronic switches provide all of the following essentials for protective shut-
downs:
• Good sensitivity and repeatability at generated vibration frequencies (espe-
cially low frequencies, 3 to 30 Hz)
• Transducer mounted on gearbox housing for good signal detection (not on
auxiliary piping or cooling tower structure where the vibration signal is attenu-
ated)
• Testing capability with fan running
• Time delay or shutdown bypass for startups
• Remote reset capability
Mechanical switches cannot be mounted on the gearbox and are not testable on-the-
run because mechanical switches do not have a remote test function. Furthermore,
bench tests have shown that, even with new mechanical switches, sensitivity and
repeatability are inadequate to detect destructive vibrations.
In addition to the above vibration switch essentials, electronic switches provide the
following features to meet the secondary purpose of applying predictive mainte-
nance techniques:
• AC output for monthly surveillance
• 4 to 20 mA output for remote vibration monitor/recorder
Mechanical switches are self-contained and are not designed to have these capabili-
ties.
Installation. Richmond Refinery now uses the PMC Beta Model 450 (see
Figure 2200-5 for the specifications and settings Richmond uses for these switches.)
Other manufacturers offer similar switches.
Switch electronics are mounted on the cooling tower in explosion proof housings.
Four of 14 switches mounted at Richmond had corrosion problems on PC boards
attributed to moisture intrusion during installation. Long term reliability of elec-
tronics in this environment has yet to be proven. Currently, the PMC Beta switches

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-13
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

are fully operational and are providing continuous protection and gearbox vibration
data via the DC Plus data collector.

Fig. 2200-5 FCC Cooling Tower Vibration Switches, Specifications and Settings
Manufacturer:
PMC Beta Corporation
4 Tech Circle
Natick, MA 01760
(617) 237-6920
Model: PMC Beta, Model 450 D-R supplied with:
• 480 VAC input transformer (L1 & L2 of 480 V System)
• 480 VAC 3 Amp Relay for Shutdown Circuit
• 0.1 to 1.5 in/sec range
• AC output on BNC Connector on Switch Panel
• AC output sensitivity = 278 MV/in/sec
Starting Lockout Terminals
3/4 FNPT connections drilled at right, left, bottom
Model 160 E transducer
Field-Configurable Settings for Cooling Tower Gearboxes:
Shutdown setpoint = 0.4 to 0.5 in/sec
Alarm Setpoint = 60% to 80%
Shutdown Relay = Normally closed (NC)
Alarm Relay = Not used
Shutdown Relay Time Delay = 3 seconds
Alarm Relay Time Delay = 3 seconds
Remote Reset = Not used

Maintenance. Perform periodic maintenance (every 3 months) in conjunction with

monthly vibration monitoring functions.
Change Corrosion Inhibitor Packet. Due to the moist environment, corrosion inhib-
itors are installed in the housings of the transformer/power supply, vibration switch,
and transducer. Corrosion inhibitor: Hoffman Corrosion Inhibitor,
Part No. A-HCI-1DV, size 0.25"  1.25"  3".
Relubricate Housing Threads with Grease. Housing cover threads corrode and must
be coated with Crouse-Hinds Anti-Seize Screw Thread Lubricant Sealer,
Part No. STL-2.
Reference—Johnson, C. W., “FCC Cooling Tower Electronic Vibration Switches,”
1/31/89, IMI, Richmond Refinery.

2200-14  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

Safety Considerations
1. When working on mechanical equipment (like the fan), utilize the electrical
lock-out feature.
2. Cooling tower fill and drift eliminators are not safe working surfaces. They
should be evaluated from existing access walkways, from air inlet openings, or
from temporary planking that spans column lines.
3. A “buddy” system should be used whenever entering any part or hatch on a
cooling tower. Only qualified people familiar with the mechanical components
and understanding the safety hazards should inspect the tower.
4. Always replace coupling guards before putting any cooling tower cell back into
service.
5. In cold climate locations, ice formation can damage tower components and be a
safety hazard. Icing procedures should be available and in good working condi-
tion, anytime the temperature drops to around 40F.

2234 Fire Protection

Nearly all of our cooling towers are made of wood and, because we are cooling
hydrocarbons in most of the exchangers, have wooden splash fill. Cooling towers
are fire hazards, particularly when idle. Recommendations for fire protection are as
follows:
1. Prohibit smoking, open lights, and warm-up fires anywhere near the tower.
2. Supervise closely any welding or cutting operations.
3. Locate new cooling towers remote from any equipment that produces sparks.
4. Provide hydrants with adequate pressure and hoses to reach all sections of the
cooling tower.
5. Install fire sprinkler systems that automatically deluge any fire source.

2240 Cooling Tower Forebay Design

This section provides basic concepts and guidelines for cooling tower forebay
design. Past experience has shown that a poorly designed cooling tower forebay will
severely impact cooling tower operation and pump life because of the following
associated problems:
• Pump cavitation
• Pump vibration
• Pump equipment damage
• Reduced pump efficiency
• Excessive noise

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-15
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Good design is especially important when large pumps (over 300,000 GPM) are
used; large pumps are more susceptible to rough running and vibration, and thus
require “better” forebay conditions for satisfactory performance.
Accordingly, the following standards are applicable to forebays equipped with either
horizontal or vertical pumps of the following capacities:
• 3000 to 300,000 GPM
• 300,000 GPM and greater
These standards do not apply to facilities with pump capacities less than 3000 GPM,
because small pumps are not usually used in cooling tower forebay applications.
These design guidelines may also apply to facilities with the same function as a
forebay, e.g., pumping station sumps. For facilities with pump capacity less than
3000 GPM, facility design should follow the pump manufacturer’s recommenda-
tions.
The information in this section is based on research conducted by the Hydraulic
Institute and British Hydromechanics Research Association. Forebay designs should
be analyzed using a hydraulic model; most models are efficient, relatively inexpen-
sive, and reliable.

2241 General Information

The forebay is an intake structure that collects and supplies a flow of water to the
suction point of the circulation pumps. The flow conditions that govern pump
performance are a function of the hydraulic design and upstream approach flow. A
“good” forebay design results in a uniform, steady, single phase flow and satisfac-
tory pump performance.
Conversely, inadequate design may cause adverse flow conditions and hydraulic
problems such as uneven flow distribution and large scale turbulence. The most
damaging conditions, however, are vortices near the pump column (vertical pumps)
and in the corners and along the walls and floor of the forebay. Even a small amount
of air entrained in the vortices will cause pump cavitation and vibration and may
lead to severe pump damage.
To avoid these above problems, the forebay design should achieve and maintain the
following conditions:
• Uniform distribution of flow entering the forebay
• Minimal circulating flows in the forebay
• Filled zones of separation
• Minimal significant fluid rotation
The following design standards provide an initial design basis. Note that these stan-
dards are subject to variation with individual applications. Hydraulic model testing
will physically analyze the preliminary design and may suggest structural modifica-
tions toward the development of the final design.

2200-16  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

2242 Forebay Design

General
Forebay design is based on the Hydraulic Institute Standards for sump design.
Continuing research on rectangular, free surface wet pit sumps with 3000 to
300,000 GPM capacities has yielded guidelines in pump position and approach
distance. All recommended distances are functions of the rated pump capacity at
design head.
As pump capacities exceed 300,000 GPM, however, the casing wall thickness (and
rigidity of support) increases disproportionally with the hydrodynamic loading on
the pump. Consequently, large capacity pumps are more prone to vibration and
demand better forebay design than smaller pumps. Using more stringent acceptance
criteria to measure “satisfactory” performance, the British Hydromechanics
Research Association has developed recommended dimensions based on the bell
diameter of large pumps.
The following sections contain general forebay design guidelines according to pump
type (i.e., horizontal or vertical pumps) and suggested forebay dimensions according
to pump capacity (i.e., 3000 to 300,000 GPM or pumps larger than 300,000 GPM).

Guidelines for Horizontal and Vertical Pumps

The following general guidelines are applicable to forebays with capacities
exceeding 3000 GPM and either horizontal or vertical pumps.
1. Ideally, a straight channel approaching the pump suction point(s) will deliver
uniform flow to the pump(s). Avoid any obstructions and/or turns that will
cause abrupt changes in flow direction; e.g., sharp corners and rapidly
diverging passages may induce eddy currents and vortices.
2. Unavoidable obstructions such as columns and cross braces should be stream-
lined to reduce the trail of alternating vortices; these vortices form in the wake
of the obstructions as water flows past.
3. Maximum velocity of the flow approaching the pump(s) should be 1.0 foot per
second. Straightening vanes and/or a longer forebay length may reduce
velocities; if properly located near the mouth of the forebay inlet, trash screens
may also function as straightening vanes.
4. A longer forebay length may also be necessary to dissipate the kinetic energy
associated with steeply sloped floors, weirs, and steps, and therefore prevent
aeration.
5. “Dead pockets” of the forebay which contain stagnant water (e.g., corners
behind the suction point) may be eliminated via simple fillets or complex form-
work.
6. The inlet to the forebay should be below the normal operating water level to
avoid aeration.

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-17
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

7. In multiple-pump installations, water should not flow past one pump suction
point to reach another; i.e., pumps should not be placed in line with the flow of
water. To maintain even flow distribution, the water stream entering the forebay
should be normal to the line of pumps and along the line of symmetry.
8. For suction bells that must be placed in line of flow, an open front cell around
each intake may induce a more uniform flow into the pumps. Cells may be
unnecessary if both the longitudinal distance between intakes and the ratio of
forebay to pump size are quite large.
9. In multiple pump installations, rounded or “ogived” separating walls may be
beneficial if pumps operate simultaneously. Otherwise, separating walls should
be avoided.
10. To avoid uneven flow distribution in multiple-pump installations, pumps should
not be placed around the edge of the forebay.
11. To avoid upstream flooding, forebay volume should be sized to accommodate
the maximum design flow during pump operation. When constant-speed pumps
are used, volume must also be adequate to prevent short cycling (rapid “on-off”
operation) of the pumps.
12. Double screens should be placed ahead of the suction of the cooling water
pumps, particularly in new installations to screen out foreign materials. Screens
should be removable, while in service, for cleaning.

Guidelines for Horizontal Pumps Only

The following general guidelines are applicable to forebays with horizontal pumps
at capacities in excess of 3000 GPM; these standards should be used in conjunction
with the information of the preceding section. Refer to Standard Drawing
GD-Q99594 for layout and piping details for horizontal pump suction lines.
1. Submergence for net positive suction head and minimal vortexing should be
according to pump manufacturer’s recommendations. On average, minimum
submergence of the suction intake is as follows:
a. Two line diameters when the intake is located in the forebay floor
b. One line diameter when the intake passes through the forebay wall
2. Vortex prevention plates just below the water surface may also be necessary to
prevent vortexing.
3. To mitigate any upstream flow disturbances, the minimum length of the suction
line should be ten line diameters.
4. An expansion joint and pipe anchor may be installed between the forebay wall
and pump to prevent overloading of the pump case.
5. Under suction lift conditions, suction piping should maintain an upward slope
to the pump; this slope helps prevent air entrainment and cavitation.

2200-18  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

6. Under flooded suction conditions, the following conditions should be main-

tained:
a. Suction piping should be level or maintain a gradual downward slope to
the pump; the piping should not extend below the pump suction flange.
b. Diameter of the intake mouth should not be smaller than the diameter of
the suction piping.
c. A gate valve should be installed in the suction piping between the forebay
wall and expansion joints. The pump may then be “disconnected” from the
forebay during inspection and maintenance.

Guidelines for Vertical Pumps Only

The following general guidelines are applicable to forebays with vertical pumps at
capacities exceeding 3000 GPM; these standards should be used in conjunction with
the guidelines above for both horizontal and vertical pumps.
1. Submergence for net positive suction head and minimal vortexing should be
according to pump manufacturer’s recommendations. Typically, minimum
submergence is two times the suction bell diameter.
2. Necessary changes in floor elevation should occur at least three suction bell
diameters upstream of the pump column(s).
3. In multiple pump installations where pumps must be placed in line of flow,
turning vanes under each suction bell may deflect the flow upward and directly
into the pump. Vanes may be unnecessary if both the longitudinal distance
between intakes and the ratio of forebay to pump size are quite large.
4. In multiple pump installations where flow distribution is skewed and pumps do
not operate simultaneously, flow splitters may redirect the flow to the suction
bells. Flow splitter lengths should be greater than four bell diameters.

Recommended Dimensions: 3000 to 300,000 GPM Capacity

The recommended forebay dimensions and layouts as shown in Figures 2200-6
through 2200-8 are applicable to facilities with either horizontal or vertical pumps
in the 3000 to 300,000 GPM capacity range (see also Standard Drawing
GD-Q99594). All dimensions are based on the rated capacity of each pump at
design head.
Dimension C is the distance between the bottom lip of suction bell and the forebay
floor. It is an average value subject to changes suggested by the pump manufacturer.
Dimension B is the recommended maximum distance between the centerline of the
suction bell and the forebay back wall. If actual Dimension B exceeds the suggested
length for structural or mechanical reasons, a “false” back wall may be installed.
Dimension S is the recommended minimum center-to-center distance between
suction bells. In single pump installations, it is the minimum forebay width.

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-19
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Dimension H is the suggested “normal low water level.” It is not the minimum
submergence required to prevent vortexing; submergence is normally defined as the
quantity H minus C.
Dimension Y is the minimum distance between the bell centerline and the first
upstream obstruction inside the forebay. For most bell designs, Dimension Y is
approximately three bell diameters.
Dimension A is the minimum overall forebay length when the average flow
velocity in the forebay is less than 2.0 feet per second.

Fig. 2200-6 Sump Dimensions vs. Flow, 3000 to 300,000 GPM Capacity (Courtesy of the Hydraulic Institute)

2200-20  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

Fig. 2200-7 Elevation of Basic Forebay Design, 3000 to 300,000 GPM Capacity (Courtesy of the Hydraulic Institute)

Fig. 2200-8 Plan of Basic Forebay Design, 3000 to 300,000 GPM Capacity (Courtesy of the Hydraulic Institute)

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-21
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Recommended Dimensions: Pumps Larger Than 300,000 GPM

The recommended forebay dimensions and layouts as shown in Figures 2200-9
through 2200-11 are applicable to facilities with either horizontal or vertical pumps
in the 300,000 GPM-plus capacity range. Dimensions are based on the intake or
suction bell diameter; unless noted otherwise, dimension symbols are identical to
those previously noted.
Dimension D is the diameter of the pump intake or suction bell.
Dimension X is the recommended distance between the edge of the bell and the
back forebay wall.
Dimension d is the diameter of the suction line or pump column.

2243 Hydraulic Model Testing

Because the hydraulic problems associated with forebay design are functions of
many variables, analysis of expected flow conditions is difficult. Unfortunately,
outside circumstances often force the designer to deviate from the design stan-
dards—and expected resulting flow conditions—described herein. On these occa-
sions, scaled hydraulic model testing may be the best method to analyze the
preliminary design.
In-situ simulation, while another possible alternative, is usually impractical. A
scaled model is more efficient because the system geometry can be quickly and
easily modified. The forebay size may be adjusted, various screen blockages
modeled, and instrumentation located in all areas of interest to measure momentum,
velocity distribution, and velocity changes at obstructions. Model forebay walls are
usually constructed of Plexiglas so that modelers and engineers may observe flow
patterns throughout the model.
The model should encompass all forebay components likely to influence the flow
entering the pump(s). Model boundaries should be located in areas where flow
pattern control has minimal boundary effects on the system. Models normally use
either equal Froude numbers or velocities; no significant scale effects occur in 1:2
and 1:4 models.
When conducted by an independent laboratory or the pump manufacturer, hydraulic
models are relatively inexpensive, reliable tools to analyze the hydraulic perfor-
mance of a preliminary design. Modifications suggested by models may also result
in substantial savings in later forebay construction, operation, and maintenance.
Since 1986, hydraulic models have been used to analyze the Richmond Refinery’s
flow splitter box of the 1A and 2A Separators, pump station of the Deep Water
Outfall, and No. 13 Separator.

2200-22  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

Fig. 2200-9 Elevation of Basic Forebay Designs, Pumps Larger than 300,000 GPM (From Hydraulic Design of Pump
Sumps and Intakes by Prosser.  1980 by the Construction Industry Research & Information Assn.,
London. Used with permission.)

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-23
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Fig. 2200-10 Plan of Basic Forebay Design, in Plane of Uniform Flow Approaching the Pumps, 300,000 GPM - Plus
Capacity (From Hydraulic Design of Pump Sumps and Intakes by Prosser.  1980 by the Construction
Industry Research & Information Assn., London. Used with permission.)

Fig. 2200-11 Plan of Basic Forebay Design, 300,000 GPM - Plus Capacity (From Hydraulic Design of Pump Sumps and
Intakes by Prosser.  1980 by the Construction Industry Research & Information Assn., London. Used with
permission.)

2200-24  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

In addition to developing possible structural modifications to improve flow condi-

tions in preliminary forebay design, models may also be used to correct conditions
in existing forebays. These improvements, the usual basic recommendations of a
model, are:
Increase the “normal low water level”. Usually, to simultaneously increase the
“normal low water level” and accommodate the desired operating forebay volume,
the forebay must be deepened. This change may increase excavation and engi-
neering costs.
Install antivortex devices. Devices such as cones, splitters, grids, and extension
plates may prevent or reduce vortexing in the forebay. The devices shown in Figures
2200-12 and 2200-13 should also be selected in consultation with the pump manu-
facturer.
Reshape the approach flow. Modifications may occur in the existing piping that
supplies the forebay and/or the inlet to the forebay.
Fig. 2200-12 Modifications to Intake Design to Reduce Vortices (From Hydraulic Design of Pump Sumps and Intakes by
Prosser.  1980 by the Construction Industry Research & Information Assn., London. Used with permis-
sion.)

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-25
2200 Cooling Tower Design Guidelines Heat Exchanger and Cooling Tower Manual

Fig. 2200-13 Other Modifications to Intake Design to Reduce Vortices (From Hydraulic Design of Pump Sumps and
Intakes by Prosser.  1980 by the Construction Industry Research & Information Assn., London. Used with
permission.)

2200-26  1989–2010 Chevron U.S.A. Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2200 Cooling Tower Design Guidelines

2244 Standard Drawings

The following standard drawing is included in the Standard Drawings and Forms
section of this manual.
• GD-Q99594 Piping and Screen Details, Suction Pit for Cooling Tower Basin.

2245 References
1. Hydraulic Institute Standards for Centrifugal, Rotary & Reciprocating Pumps,
14th Edition, Hydraulic Institute, 1983.
2. Nystrom, James B., et al., “Modeling Flow Characteristics of Reactor Sumps,”
Journal of the Energy Division, ASCE, Vol. 108, No. EY3, November 1982.
3. Padmanabhan, M., and G. E. Hecker, “Scale Effects on Pump Sump Models,”
Journal of Hydraulic Engineering, ASCE, Vol. 110, No. 11, November 1984.
4. Prosser, M. J., The Hydraulic Design of Pump Sumps and Intakes, British
Hydromechanics Research Association/Construction Industry Research and
Information Association, 1980.
5. Sweeney, Charles E., et al., “Pump Sump Design Experience: Summary,”
Journal of the Hydraulics Division, ASCE, Vol. 108, No. HY3, March 1982.

Revision History
Date Description Author Sponsor
December 1989 Initial release
Februray 2010 Errata: — Replaced references to GB-Q99594 with GD-Q99594 BJBL RPHO

December 1989  1989–2010 Chevron U.S.A. Inc. All rights reserved. 2200-27
2300 Ratings/Retrofits

Abstract
This section discusses rating and retrofitting a cooling tower, and includes a simpli-
fied method for evaluating a cooling tower, with sample calculations, and ways to
retrofit existing towers to improve performance and efficiency.

Contents Page
2310 Background on Rating a Cooling Tower 2300-2
2311 Purpose of Rating
2312 Qualified Cooling Tower Raters
2320 Simplified Method of Evaluating a Cooling Tower 2300-3
2330 Possible Retrofits to Existing Cooling Towers 2300-15
2331 Ways to Change L/G, Liquid Water as a Ratio to Air
2332 Ways to Increase KaV/L
2333 Ways for More Efficient Mixing and Distribution of Air-Water

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-1
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

2310 Background on Rating a Cooling Tower

2311 Purpose of Rating

To rate a newly constructed tower or upgrade or retrofit an existing one, the perfor-
mance of cooling towers should be evaluated to determine:
1. If a newly purchased tower meets its guaranteed operating condition.
2. The conditions before and after a guaranteed “Upgrading” of the tower.
3. For an operating cooling tower, the effects of several significant parameters,
such as:
a. Increased heat load requirements.
b. Lower cold water temperature requirements from the tower.
c. Increased GPM requirements.
4. The effects of changes in mechanical features, such as the type and amount of
fill in the tower, the horsepower of the fan, the flow rate, etc.

2312 Qualified Cooling Tower Raters

Outside Resources
For either new construction, or for upgrades, the following independent sources for
cooling tower performance testing are currently considered acceptable:
1. Environmental Systems Corporation, Knoxville, Tennessee, (615) 688-7900.
The current authorized contract agency for testing for the Cooling Tower Insti-
tute (CTI).
2. Midwest Research Institute, Kansas City, Missouri, (816) 753-7600. The
previous testing agency for CTI and still competent and capable.
3. Willa, Inc., St. Louis, Missouri, (314) 965-5051. Willa wrote the Cooling Tower
Performance Testing Manual for CTI. For over 10 years, after formation of the
CTI, Willa performed all CTI testing under ATC-105, the testing procedure
Willa wrote.

2300-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Company Participation
Below is a summary of the costs, equipment, and Company personnel required to
assist in a qualified cooling tower performance test:
1. As a nonmember of CTI, Chevron normally pays $3500 to $5000 per test with
the first two agencies listed above. Willa’s rate is currently $600/day for time,
$300/day for travel, $200/day for rental use of Performance Instrumentation,
plus travel (coach), and living expenses. In addition to being a qualified tester,
Willa can also make recommendations for tower performance improvements.
2. Thermal performance tests in accordance with CTI ATC-105 require up to
$35,000 of testing instrumentation equipment to be brought onsite by the
testing agency. Three or four personnel must be onsite at the plant dedicated to
the testing for most of their time for up to 2 full days.
For any “guaranteed” improvement to an existing cooling tower to have any
meaning, a tower performance test should be conducted both before and after the
work being performed. Penalty clauses for not meeting guarantees are included in
Model Specification EXH-MS-1317, Induced Draft Cooling Towers, included in
this manual.

2320 Simplified Method of Evaluating a Cooling Tower

The following approach is a simple way for an operating center to find out what a
cooling tower will do at any operating range and any wet bulb temperature. This
procedure is not precise, but it is within the accuracy of the temperature measure-
ment instrumentation. It can, therefore, be very useful.
Because the heat discharged by the water in the tower equals the heat absorbed by
the air going up the tower, a number of parameters cancel out. All that is needed to
rate a tower are the four following temperatures and the CTI performance curve:
1. Hot water temperature (Thw), F
2. Cold water temperature (Tcw), F
3. Inlet air wet bulb temperature (twb), F
4. Outlet air temperature (tao), F
Instrumentation requirements are limited to a thermometer and a sling psychrom-
eter. A psychrometer is an inexpensive instrument used to measure wet bulb temper-
atures by passing air over the bulb of a thermometer that is covered by a cloth or
wick saturated with water. Heat is removed by adiabatic evaporation. Either a sling
or a mechanically aspirated psychrometer is acceptable, provided the instrument is
properly shielded from radiation and the air across the wick is limited to
1000 feet/minute.
As shown in the following discussion, water flow (L) and air flow (G) are not
required to evaluate the tower.

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-3
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

From the heat balance for a cooling tower:

L  Cp  (Thw - Tcw) = G (eao - ewb)

where:
L = Liquid water, lb/hr
G = Air flow, lb/hr
eao = Enthalpy, Btu/lb of the exit air from the tower.
(See Fig. 2300-1.)
ewb = Enthalpy, Btu/lb of the air entering the tower.
(See Fig. 2300-1.)
Figure 2300-1 lists the enthalpies of all cooling tower air-water vapor mixtures. In
the figure, eao and ewb are found at their corresponding temperatures.
Cp, the specific heat for liquid water (Btu/lbf) is so close to 1, that we can call it 1.
Therefore:
e ao – e wb
L- = -------------------------
--- enthalpy-
- = -----------------------
G T hw – T cw range
(Eq. 2300-1)
This L/G is the basis of the cooling tower rating curves in the CTI Performance
Curves Manual. For every practical wet bulb temperature (50F through 90F in
steps of 2 degrees), the manual provides a “curve” for each range of 4, 6, 8, 10, 12,
14, 15, 16, 18, 20, 22, 24, 26, 28, 30, 35, 40, 50, 60, 80, and 100 degrees Fahren-
heit. An example of one of these curves is shown in Figure 2300-2, Example of a
CTI Performance Curve.
The only other two parameters on these performance curves are:
1. “KaV/L” is the Cooling Tower Characteristic Equation. It is developed by
substituting the heat balance equation for the tower into the mass transfer equa-
tion.
Increasing this number with a constant liquid-to-air ratio on the tower results in
better performance from the tower. That is, for two towers, each with the same
L/G and the same return inlet water temperature, the tower with a higher design
KaV/L will produce the colder water back to the operating plant. KaV/L is
discussed in more detail in Section 2332 below.
The KaV/L is an integration of the temperature with respect to the enthalpy.
This value is difficult to calculate for a given tower. For the simplified method
discussed here, we can avoid making this calculation by reading it from the
charts in the CTI Performance Curves Manual.
2. The “approach” is defined as the difference between the cold water tempera-
ture and the inlet wet bulb air temperature. These two temperatures are two of

2300-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Fig. 2300-1 Enthalpy of Saturated Air-Water Vapor Mixture (at 29.921 in. Hg) (1 of 2)(Courtesy of the American Society
of Heating, Refrigerating and Air Conditioning Engineers)

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-5
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

Fig. 2300-1 Enthalpy of Saturated Air-Water Vapor Mixture (at 29.921 in. Hg) (2 of 2)(Courtesy of the American Society
of Heating, Refrigerating and Air Conditioning Engineers)

2300-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Fig. 2300-2 Example of a CTI Performance Curve (Courtesy of the Cooling Tower Institute)

the four recorded temperatures listed above. Approach can therefore be calcu-
lated.
“Approach” = Tcw - twb
With the proper performance curve in the manual for the range and the wet bulb, we
can plot the L/G and the “approach.” At their intersection, KaV/L can be found on
the ordinate.
We now have an excellent tool for answering two performance questions about the
tower.
Performance Question 1.

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-7
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

How will the tower perform for any other range and wet bulb temperature?
Knowing L/G and KaV/L for the tower, we can use L/G as the abscissa and KaV/L
as the ordinate on any other curve in the CTI Performance Curves Manual. By
locating the intersection of these two parameters on any curve, we find the
“approach” that the tower is capable of at that range and wet bulb temperature for
that curve. Knowing the approach, we can easily calculate the cold water tempera-
ture from the tower for that range and wet bulb temperature.
Example: Four temperatures taken at the tower are:
Hot water temperature (Thw) = 120F
Cold water temperature (Tcw) = 85F
Inlet air wet bulb temperature (twb) = 78F
Outlet air temperature (tao) = 102.6F

L e ao – e wb 76.58 – 41.58
---- = -------------------------
- = --------------------------------- = 1.0
G T hw – T cw 120 – 85
(Eq. 2300-2)
“Approach” = Tcw — twb = 85F - 78F = 7F
As shown on Figure 2300-2, at the intersection of an L/G of 1 and an approach of
7F, KaV/L for the tower equals 2.35.
As discussed above, an L/G of 1 and a KaV/L of 2.35 now define the design of the
tower. They can now be used to find the approach and the temperature of the water
from the tower for any range and wet bulb temperature. (See Figure 2300-3, for an
example.)
The curve was plotted reading the approach from a series of CTI performance
curves which only varied the wet bulb temperature. In all cases the range was held
at 35F, and the various wet bulb performance curves were entered (with L/G equal
to 1 and KaV/L equal to 2.35), to find the “approach.”
The wet bulb temperature when added to the “approach” gives the temperature of
the water from the cooling tower. This is also shown on the plot. This graph is a
useful tool in estimating the temperature of the water from a given cooling tower for
various wet bulb temperatures.

2300-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Fig. 2300-3 Example: Approach Temperature and Temperature of Water from Tower Basin versus Ambient Wet Bulb
Temperature

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-9
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

Performance Question 2.
What will happen if we change something on the tower?
Examples of changes are:
1. What if we change the liquid to gas ratio (L/G) for a given tower:
a. By increasing the water flow over the tower.
b. By installing high horsepower fans.
2. What if we increase the fill, in particular if we replace some of the existing
splash fill in the tower with film pack cellular fill.
These are questions most frequently asked when better performance is required of a
tower. From the discussion above, we already know the range, the wet bulb temper-
ature, and L/G and KaV/L for the present condition of the tower.
To answer these questions, we must also take into consideration the type of fill used
in the cooling tower because the type of fill affects the CTI performance curves.
There are basically two types of cooling tower fill: splash fill and film pack fill. See
Section 2141 for more detail about these fills.
1. Splash fill. For years splash fill has been the standard fill for cooling towers.
Typically, splash fill is formed by 1/2-inch boards on 6-inch checkerboard alter-
nating positions on 1-foot elevation differential. This fill normally occupies the
bottom two-thirds of a counterflow tower. It has been used for many years and
continues in use in over 60% of new and upgraded cooling towers in the petro-
leum industry. Splash fill is not adversely affected by hydrocarbon leakage into
the tower.
On the CTI performance curve, splash fill will have a performance curve with a
negative slope of 0.6. See Figure 2300-4.
2. Film pack fill. Film pack fill is a PVC cellular fill that has much more wetted
area per unit volume than splash fill. It has a honeycomb configuration.
Munters and Brentwood are well-known manufacturers of film pack fill. If the
water is clear, film pack fill performs as predicted but should not be used with
waxy or heavy oils.
On the CTI performance curve, film pack fill will have a performance curve
with a negative slope of 0.8. Figure 2300-5 is a CTI blank form on which the
performance curve for any film pack fill can be plotted.

By superimposing the straight line plots of the fill performance curves on the CTI
performance curve of Figure 2300-2, we can answer the questions posed above, 1.a
and 1.b (See Figure 2300-6).

2300-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Fig. 2300-4 CTI Tower Characteristic Curves at 0.6 Slope (Courtesy of the Cooling Tower Institute)

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-11
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

Fig. 2300-5 Blank CTI Graph Form for Plotting CTI Performance Curve (Courtesy of the Cooling Tower Institute)

2300-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

Fig. 2300-6 Example of CTI Performance Curves for Fill at 0.6 Slope Superimposed on CTI Performance Curves for
Cooling Tower (7F Approach). (See text for explanation.) (Courtesy of the Cooling Tower Institute)

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-13
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

To answer Performance Question 2.1.a (increasing the water flow):

From our example in Figure 2300-2, assume that the fill is splash fill and that we
increase the water flow by 20%. Figure 2300-6 shows the increase of the
“approach” temperature for an increase in flow of 20%. For this case, follow the 0.6
negative slope of the tower characteristic curve from the design L/G of 1.0 to 1.2
because we have increased “L” (liquid) by 20%. At L/G of 1.2, the approach has
increased from 7F to 9F. Therefore, the water temperature from the tower basin
will be 2F warmer. (Tcw = “Approach” + twb.) The impact of this increase in water
temperature and increase in flow rate can be evaluated by a process or heat
exchanger engineer.
To answer Performance Question 2.1.b (installing high horsepower fans):
Horsepower varies to the third power of air flow. Therefore, an increase in motor
size by one-third, say installing 200 horsepower for 150 horsepower, the increase in
air flow (200/150) to the one-third power equals 110%, or a 10% increase in air
flow. If L/G for the tower is 1.0, increasing the air flow by installing 200 horse-
power drivers would reduce L/G to 1.0/1.1 or 0.91.
Figure 2300-6 also shows the reduction of the approach temperature for an increase
in fan horsepower of one-third; for this case, follow the 0.6 negative slope of the
tower characteristic curve from the design L/G of 1.0 to 0.91. The approach has
decreased from 7F to slightly less than 6F. Therefore, the water temperature from
the tower basin will be slightly more than 1F cooler (Tcw = “Approach” + twb). The
benefits of this decrease in water temperature, with no change in the flow rate
“over” the tower, can be evaluated by a process or heat exchanger engineer.
To answer Performance Question 2.2. (increasing the fill):
In this case we add to the cooling tower a fill with more wetted surface. Manufac-
turers of this type of fill (Munters, Brentwood, etc.), furnish fill performance curves
that can be superimposed on the CTI performance curves to evaluate the effect of
changing fill. The increased KaV/L and the new approach can be located directly
above and on the same L/G, assuming negligible changes in liquid and air flow as a
result of adding a foot or so of film pack for splash fill.
Maximum improvement by changing from splash fill to film pack:
• Crossflow towers can usually only be improved by 5 to 10%. If all factors are
favorable, maybe by 12%.
• Counterflow towers can usually be improved by as much as 20 to 25%.
Before deciding to replace splash fill with more efficient film pack, you must,
however, evaluate the following key potential problems.
• The potential for plugging the film pack with hydrocarbon and any process
chemical contamination.
• The need for increased air moving energy due to increased air static pressure
drop.
Because of these problems, only about 40% of new and upgrade towers for refinery
and oil oriented cooling towers are switching to film pack.

2300-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2300 Ratings/Retrofits

2330 Possible Retrofits to Existing Cooling Towers

2331 Ways to Change L/G, Liquid Water as a Ratio to Air

Listed below are several ways to change the ratio of liquid water to air (L/G).
1. Increase horsepower. Air flow through the tower varies with horsepower to the
one-third power.
2. Larger fan. Air flow is directly proportional to the fan diameter.
3. More efficient fan. Fan efficiencies are often not as good as manufacturers’
claims. The best fans when tested under field conditions have fan efficiencies
of 70 to 73%, not of 80% or better, as they claim.
4. Velocity recovery stack. Eighty percent of the velocity reduction through the
fan cylinder (stack) can be recovered. A velocity recovery stack is in the form
of a flared diffuser. It provides a gradual increase in cross-sectional area with a
resultant decrease in existing air velocity. This converts velocity pressure to
static pressure and reduces the work on the fan.
5. Decrease in static pressure:
a. Arrangement of fill. Turn fill parallel to air flow.
b. Selection of drift eliminators. Comparison of velocity against pressure drop
must be evaluated.
c. Louvers. Except for protection against icing leave louvers off. For counter-
flow towers, make sure there is a middle wall.
6. Increasing active plan area (cross-section of tower where water and air can
flow). Make sure all areas of the tower are open to water and air flow.
7. Convert old trough design to PVC spray distribution. The PVC distribution
gives better distribution and reduces pressure drop.

2332 Ways to Increase KaV/L

Although it is not necessary to understand the parameters in this dimensionless heat
transfer characteristic equation to be able to use the simplified cooling tower rating
system presented in this section, they are as follows:
K = Overall average unit energy transfer conductance per square foot
water surface area within the tower
a = Area per unit volume of tower
V = Active volume of cooling tower
L = Water circulation rate, volume or weight per unit time

December 1989  1989 Chevron USA Inc. All rights reserved. 2300-15
2300 Ratings/Retrofits Heat Exchanger and Cooling Tower Manual

Increase KaV/L by the following:

1. Increase the amount of fill in the tower.
2. Use a higher efficiency splash fill.
3. Use a more efficient film fill.
4. Raise distribution piping and add fill.

2333 Ways for More Efficient Mixing and Distribution of Air-Water

1. Full cone nozzle. To assure uniform distribution of the water across the fill.
2. Finer spray. But not too fine, so that drops can escape through the drift elimina-
tors.
3. Complete coverage. Make sure all areas of the tower have air and water flow in
contact.
4. Increase plenum height. In a counterflow tower, for minimum static pressure
drop and maximum contact, the plenum height should be a minimum of 8 feet
above the drift eliminators and the base of the stack.

2300-16  1989 Chevron USA Inc. All rights reserved. December 1989
2400 Water Treatment System Design

Abstract
Treating recirculating cooling water is necessary to protect the plant and minimize
operating costs. This section, in conjunction with Appendix J, discusses the mecha-
nisms of fouling and corrosion in a cooling water system, the water treatment
methods, and the sources of makeup water.
Each system must be individually evaluated to establish specific water treatment
requirements, type of controls and supporting equipment, and operating costs.
System effluent quality resulting from chemical treatment must be in accordance
with the local governmental regulations.
For any cooling tower evaluation, the costs should include the cooling tower itself,
any incremental makeup water treating equipment, and cooling tower blowdown
costs. Operating cost, including maintenance for cooling towers, can be $10/GPM-
year, especially if nonchromate inhibitors must be used instead of the more effec-
tive, less expensive chromate inhibitors.

Contents Page
2410 Components of Water Treating Package 2400-3
2420 Scale Control 2400-3
2421 pH Control
2422 Blowdown Control
2423 Antiscalant
2430 Fouling Control 2400-6
2431 Causes of Fouling
2432 Cooling Water Velocity Design
2433 Use of Antifoulants and Dispersants to Control Fouling
2440 Microbiological Growth Control 2400-7
2441 Cause of Microbiological Growth
2442 Chlorine Injection
2443 Nonoxidizing Biocide

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-1
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

2450 Corrosion Control 2400-8

2451 Mechanisms of Corrosion
2452 Types of Corrosion Inhibitors
2453 Evaluation of Corrosion Inhibitor Alternatives
2460 Monitoring of Corrosion and Fouling Corrosion 2400-14
2461 Corrosion
2462 Fouling
2463 Microbiological Growth
2470 Automatic Control Systems 2400-15
2480 Sources of Reused Water for Makeup to Cooling Towers 2400-15
2481 Boiler Blowdown or Concentrated Effluent from Reverse Osmosis Water
Treater
2482 Treated Plant Waste Water
2483 Municipal Secondary Treated Effluent
2484 Process Water
2490 Choosing a Water Treatment System Supplier 2400-17

2400-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

2410 Components of Water Treating Package

Components of an open recirculating cooling water treatment package may include:
1. Acid required to control the pH to prevent scaling. See Section 2420.
2. An antiscalant to protect against calcium carbonate, magnesium silicate, and
calcium sulfate. See Section 2420.
3. Antifoulant or dispersents (if necessary). See Section 2430.
4. A biocide to kill or inhibit bacterial or algae fouling. See Section 2440.
5. A surfactant to supplement the biocide and help it effectively penetrate inor-
ganic deposits that may shelter bacterial growth.
6. A corrosion inhibitor for steel and an additional corrosion inhibitor for copper
alloys, if needed. See Section 2450.
It should be noted that antifoulant, dispersant, antiscalant, or scale inhibitors are
frequently used interchangeably. When discussing with individual suppliers, it is
best to ask for their particular definition of terms. Some chemical compounds, for
example, polyphosphates and phosphonates, serve several purposes, such as corro-
sion inhibition and scale inhibition.
Sections 2420 through 2450 discuss the above mechanisms along with recom-
mended treatments.

2420 Scale Control

Scaling in cooling water systems is normally the result of exceeding the solubility of
scale-forming compounds, such as calcium sulfate, magnesium silicate, etc. In
almost all cases, scaling occurs first in the hottest portions of the system. The result
of scaling is a hard, crystalline deposit that reduces heat transfer efficiency and
possibly exchanger capacity. The primary means of preventing scale are pH control
and blowdown. Antiscalant compounds are also frequently used.

2421 pH Control
Deposition of calcium scale on tubes is normally avoided by addition of sulfuric
acid to control pH between 6.5 to 7.0. At 180F, the solubility of calcium sulfate
(reaction product of sulfuric acid and calcium carbonate) at 1200 ppm is higher than
the solubility of calcium carbonate at 200 ppm.
The following are recommendations for pH control:
1. Provide automatic pH control using on-off control of the acid. Design the
forebay acid injection system to achieve optimum distribution and dilution of
acid to the pump suction lines.
2. Concentrated sulfuric acid has a specific gravity of 1.86. It will tend to sink to
the bottom of the tower unless it is agitated into the water to dilute it.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-3
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

Figure 2400-1 shows a polyvinyl-chloride mixing trough for adding the acid to
a flowing water stream, which continuously recirculates from the cooling water
pumps into the cooling tower forebay.

Fig. 2400-1 Cooling Tower Acid Injection

2400-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

3. Provide a means for the operator to adjust the capacity of the acid valve or
pump. Based on the pH recorder, the operator should adjust the pH controller
(pHRC) so that the acid is “on” half of the time and “off” half of the time. The
acid valve or pump should accordingly be sized for 200% of the steady-state
acid rate.
4. Choosing the most representative sample location is critical. The cooling water
pump discharge header is the preferred location to measure pH, because there is
little residence time between the acid injection point and the pump. However,
piping arrangements can preclude this spot. Cooling towers frequently have
multiple pumps and underground headers that can go off in several directions.
It is difficult to locate a sample point at which the sample is always representa-
tive.
The return line at the cooling tower is frequently the best place to measure pH. It is
always flowing, so the sample is always representative. On large towers the resi-
dence time in the tower is usually large compared to the residence time in the
cooling water loop, so that sampling the return line will result in acceptable pH
control.

2422 Blowdown Control

As discussed in Section 2300, chemical treatment and wastewater discharge treat-
ment costs are directly related to the blowdown rate and the cycles of concentra-
tion. The most economical operation is the one that will concentrate to the
maximum without formation of deposits on heat transfer surfaces or without
dramatically increasing corrosion rates.
Figure 2200-4 (Section 2200) illustrates that above 8 to 10 cycles of concentration
increasing the cycles of concentration has minimal effect on the blowdown rate.

Automatic Blowdown Control

The following are recommended procedures to provide automatic blowdown control
using total dissolved solids (TDS) as an indicator:
1. Provide continuous manual blowdown of 80% of the total blowdown rate.
2. Use a conductivity analyzer and monitor to monitor total dissolved solids
(TDS).
3. Install an on-off, conductivity controller to operate a blowdown trim valve. The
trim valve should be in parallel with the manual blowdown valve. Also, the
trim valve should be sized for 40% of the total rate, as it is “on” only half the
time.
4. Install an event recorder to mark the times when the trim valve is open. Based
on the event recorder, the operator should adjust the manual blowdown valve to
keep the trim valve “on” half the time and “off” half the time.
ETD’s Monitoring and Control Systems Division is available for assistance in
designing this system.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-5
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

2423 Antiscalant
Antiscalant compounds are also used to avoid calcium scales. The common scale
inhibitors are polyphosphates (sodium hexametaphosphate or sodium tripolyphos-
phates), phosphonates, and low molecular weight polymers. Wood by-products
lignins and tanins also have limited scale inhibiting properties. In many instances,
the scale-inhibiting properties of the polyphosphonates are enhanced when they are
combined with polymers and/or surface active agents.
For more detailed information on antiscalants, see “Scale Deposition,” in Power
Magazine Special Report, “Cooling-Water Treatment for Control of Scaling,
Fouling, Corrosion” in Appendix J. The following Tables in this report are suitable
references:
Table 1: Scale Occurrence in Cooling Water Systems
Table 2: Calcium Carbonate Indicators, Indexes
Table 3: Scaling Severity Keyed to Index
Table 4: Leading Scale-control Agents
Table 5: Scale Control with Phosphonates

2430 Fouling Control

2431 Causes of Fouling

Fouling in cooling systems is caused by the deposition of suspended solids on heat
exchanger surfaces. Typical fouling materials are microbiological deposits, corro-
sion products, suspended solids in the makeup water (such as silt and dirt), airborne
products that are scrubbed out by the cooling tower, and process contamination.
System design and treatment control will minimize fouling due to corrosion prod-
ucts and microbiological activity, but the control of fouling caused by suspended
materials introduced either in the makeup water or from the air may require the use
of a suitable dispersant.

2432 Cooling Water Velocity Design

In general, most fouling control problems are connected with low circulating water
velocities. Some common low water velocity places are engines or compressor
jackets and shell side coolers. Low velocities in any type of equipment permits
suspended solids in the circulating water to settle out. For this reason, when
designing a cooling water system, water velocities should be maintained above
8 feet per second.

2400-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

2433 Use of Antifoulants and Dispersants to Control Fouling

For detailed information on fouling, antifoulants, and dispersements, see the Power
Magazine Special Report, “Cooling-Water Treatment for Control of Scaling,
Fouling, Corrosion,” in Appendix J. There is, however, mixed opinion as to the
effectiveness of these chemicals. Heat Transfer Research Institute has tested many
of these chemicals and has found none it can recommend.
Fouling deposits caused by periodic process contamination can be minimized or
removed with the application of special types of dispersants. These dispersants are
usually fed at relatively high rates during the contamination period and for several
weeks thereafter. The treatment levels employed are usually on the order of 100 to
150 ppm. Because process contamination is, hopefully, infrequent, no special
feeding equipment is provided for the oil solubilizing dispersants.

2440 Microbiological Growth Control

2441 Cause of Microbiological Growth

Environmental conditions in most cooling systems are ideal for the growth of unde-
sirable microbiological organisms. These organisms are usually in the form of algae
and various types of bacteria that create fouling and corrosion problems. For these
reasons, it is important that microbiological activity be minimized. This is usually
accomplished through the application of gaseous chlorine and/or various types of
proprietary nonoxidizing biocides.

2442 Chlorine Injection

From a cost standpoint, the least expensive oxidizing biocide is gaseous chlorine fed
to the circulating water system by means of a chlorinator. Chlorine should be
considered for all systems except where the chlorine demand may be abnormally
high due to the presence of various types of organics or reducing agents. Some
compounds that will raise the chlorine demand are hydrogen, sulfide, ammonia,
various sulfites, certain types of amines and some kind of hydrocarbons. Chlorine
demand must be satisfied before chlorine is available to function as a biocide.
Although high chlorine demand is usually not encountered using fresh makeup
water, it can be a problem with various types of reclaimed waters. High chlorine
demand can also be caused by process contamination in the circulating water
system. If there is any question when designing a new system, a sample of makeup
water should be tested for determining the chlorine demand.
The disadvantages of using chlorine are:
1. Chlorine renders both pine and redwood susceptible to attack by fungi. It solu-
bilizes the natural toxic extractives from the wood.
2. Chlorine is often not compatible for use with most of the nonmetal corrosion
inhibitors which operate at pH of 7.5 or above. Chlorine at this pH accelerates
wood deterioration.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-7
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

3. Bacteria tend to become immune to chlorine after prolonged exposure. This

may be overcome by “slug” feeding an effective nonoxidizing biocide at
specific intervals.
Chlorine can be added continuously or intermittently. Free residual of 0.5 to 1 ppm
chlorine is usually desired. It can be controlled by an oxidation reduction potential
control system. Chlorine injection is covered in detail in Section 2540.

2443 Nonoxidizing Biocide

Nonoxidizing biocides are mainly of two types:
1. Enzyme poisons, such as methylene bisthiocyanate, acrolein, and metal salts
(copper sulfate).
2. Toxicants that disrupt the cell wall and cytoplasm, such as chlorophenols, dode-
cylguanidine hydrochloride and quaternary ammonium salts.
These compounds can be highly toxic and can be expensive to use. Their use must,
therefore, be carefully controlled. Very often these compounds are used in conjunc-
tion with chlorine. The feeding program usually involves continuous use of chlo-
rine followed by shock treatment with an effective nonoxidizing biocide at specific
intervals. The need for biocide addition can be determined by standard bacterial
count methods. Once a pattern of bacterial growth versus biocide requirement is
established, the need for continued bacterial monitoring should be minimal.
Proprietary nonoxidizing biocides provided by the water treatment suppliers are
usually liquid products available in 30-gallon or 55-gallon drums. These materials
are usually “batch” fed periodically to the cooling tower basin. Although the appli-
cation can be automated, the usual practice is to make the addition manually,
utilizing a suitable bucket or container.
Refer to Section 2230 for information on the environmental impacts of using
biocides. For more detailed information on biocides, see “Biological Fouling” in
Power Magazine Special Report, “Cooling-Water Treatment for Control of Scaling,
Fouling, Corrosion,” in Appendix J.

2450 Corrosion Control

2451 Mechanisms of Corrosion

Refer to Appendix J and the Corrosion Prevention Manual for detailed information
on the corrosion mechanisms related to cooling water.

2400-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

2452 Types of Corrosion Inhibitors

The corrosion inhibitors most frequently used in cooling water systems are classi-
fied as passivators. These inhibitors passivate the metal by encouraging a tenacious
metal oxide or other film to form on the metal surfaces. Combinations of chromate,
polyphosphates, and zinc have been the most used inhibitors. Because of environ-
mental regulations, chromates and zinc are being restricted to very low treatment
levels or eliminated. Other corrosion inhibitors include nitrites, silicates, amines,
and various other organic agents.
Various corrosion inhibitors are listed in Figure 2400-2. Normal concentrations,
advantages, and disadvantages are given in this table.
For more detailed information on corrosion inhibitors see the section “Corrosion
Control” in Power Magazine Special Report, “Cooling-Water Treatment for Control
of Scaling, Fouling, Corrosion” in Appendix J.

Copper Alloy Corrosion Inhibitors

Corrosion inhibitors for copper alloys must supplement the organic-type steel corro-
sion inhibitors. The inhibitors for copper are generally organic sulfur compounds or
an azole (five-membered ring containing at least one atom of nitrogen). It is impor-
tant to recognize that inhibitors other than chromates may not adequately protect
copper by themselves.

2453 Evaluation of Corrosion Inhibitor Alternatives

The evaluation of alternative corrosion inhibitors should include:
1. Metals to be protected
2. If any chromium is used, the desired discharge levels in the effluent
3. Relative performance of the inhibitor
4. Quality of makeup water
5. Comparative costs of chemicals
6. Ease of pH control
7. Process system leakages and airborne contaminants near the cooling tower
8. Toxicity level of the inhibitor/biocides
9. Number of cycles of concentration
10. Effect of inhibitor on biogrowth
11. Biodegradability of the inhibitor
Because these parameters are dependent on specific conditions at the operating site,
no general recommendations can be made.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-9
 Fig. 2400-2 Corrosion Inhibitor Summary (1 of 4)

2400-10

2400 Water Treatment System Design

Concentration in Disadvantages or Reasons for Caution in
Inhibitor Type Circulating Water, ppm Advantages Use

Chromate 200 and above 1. Easy to control concentration 1. Toxic (affects physiological processes in
(High Concentration Chromate 2. Gives reasonable protection even with mammals)
Seldom Used Except in Small process leakage into cooling water 2. Expensive
Systems)

Chromate-Phosphate About 30-70 Chromate 1. Very good protection [<1 mil per year (mpy)] 1. Toxic
About 5-10 Phosphate due to dianodic inhibition 2. Expensive
2. Excellent inhibition of copper and its alloys 3. pH variations may cause formation of
3. Keeps surfaces very clean orthophosphate deposits
4. Polyphosphate enhances bacteria growth
 1989 Chevron USA Inc. All rights reserved.

Chromate-Zinc About 20 Chromate 1. Superior corrosion protection at low cost (<1 1. Toxic
About 3 Zinc mpy). (This is the most commonly used 2. Susceptible to CrO4= reduction and
inhibitor in open circulating systems. Chro- resulting high corrosion when process leaks
mate is the anodic inhibitor. Zinc is the develop
cathodic inhibitor.)
2. Inhibits copper and aluminum corrosion
Low Chromate-Zinc About 5 Chromate 1. Least toxic of all chromate inhibition at low 1. Contributes less chrome but more zinc to
About 8-10 Zinc cost refinery effluent
2. Gives adequate corrosion inhibition at low 2. Very susceptible to failure if process leaks
cost develop

Polyphosphate 3-5 controls scale 1. Relatively inexpensive 1. Hydrolyzes to orthophosphate. (Orthophos-

100-1000 complexes with Ca++ and 2. Nontoxic phate is less effective as a corrosion inhib-

Heat Exchanger and Cooling Tower Manual

Mg++ to prevent precipitation itor.)
3. Acts as scale preventive as well as a corro-
20-25 for corrosion protection sion inhibitor 2. Reverted orthophosphate can react with
calcium to form deposits if concentrations
4. Can be used with chromate, zinc, or are high
organics to improve corrosion and scale
control 3. Contributes nutrient to bacteria or algae
4. Not a superior steel corrosion inhibitor by
itself
5. Unsatisfactory inhibitor of copper and
aluminum corrosion

Chromate-Phosphate Zinc About 15-20 Chromate 1. Extremely good protection (<1 mpy) due to 1. Toxic
December 1989

2-5 Zinc anodic/cathodic inhibition 2. Chromate and zinc in discharge

2-5 Phosphate 2. Protects copper also
3. Zinc provides extra corrosion control
 Fig. 2400-2 Corrosion Inhibitor Summary (2 of 4)

December 1989

Heat Exchanger and Cooling Tower Manual

Concentration in Disadvantages or Reasons for Caution in
Inhibitor Type Circulating Water, ppm Advantages Use

Polyphosphate-Zinc 25-40 Phosphate 1. Corrosion protection performance 1. Has the disadvantages of polyphosphate
5-15 Zinc approaches chromate. (Polyphosphate is inhibitor alone. (See items 1–3 under poly-
the anodic inhibitor; zinc is a cathodic inhib- phosphate.)
itor.) 2. Zinc present in blowdown water
2. Cost about the same as chromate-zinc 3. Requires a supplemental inhibitor, such as
treatment mercaptobenzothiazole, to provide protec-
tion against copper corrosion

Polyphosphate-Ferrocyanide 20 pp, Phosphate 1. Adequate corrosion protection 1. Has the disadvantages of polyphosphate
7 ppm Ferrocyanide 2. Low toxicity. (Ferrocyanide is stable and will inhibitor alone. (See items 1–3 under poly-
 1989 Chevron USA Inc. All rights reserved.

not release HCN at low pH.) phosphate.)

2. Must have at least 100 ppm CaCO3 present
in circulating water
3. Must have close pH control to avoid trical-
cium phosphate precipitation

Polyphosphate-Organic 3-6 ppm Phosphate 1. Claimed to have very low toxicity and easily 1. Limited refinery experience with this product
(Organic portion of these inhibitors is biodegradable 2. Part of the corrosion protection at higher-
proprietary information) 2. Phosphate content low enough to prevent than-normal pH (range variable). (This
any significant contribution to eutrophication means Ca++ or Mg++ solubility may be
3. No heavy metals exceeded.)

4. Low cost—Equal to chromate in cost. 3. Reversion of polyphosphate may be

problem

Phosphonate 5-10 as scale inhibitor 1. Greater temperature stability than polyphos- 1. Not a satisfactory inhibitor by itself at
phates; does not hydrolyze to give large normal pH levels (6.5-7.0), especially at
amounts of orthophosphate higher temperatures

2400 Water Treatment System Design

2. Polymerlike structure helps inhibit forma- 2. Presence of organic material will tend to
tion of scale (as does polyphosphate) increase biogrowth. (This leads to increased
biocide consumption and costs.)

Phosphonate-Zinc 30-40 ppm (3-4 ppm Zinc) 1. Corrosion protection almost equal to chro- 1. Expensive
(This is used with a moderate pH mate 2. Zinc is present in blowdown water
increase to about 7.8. Higher pH inter-
feres with the use of zinc) 3. Biocide costs go up due to increased
biogrowth potential resulting from higher
organic content of the cooling water
4. Requires a supplemental copper corrosion
inhibitor
2400-11
 Fig. 2400-2 Corrosion Inhibitor Summary (3 of 4)

2400-12

2400 Water Treatment System Design

Concentration in Disadvantages or Reasons for Caution in
Inhibitor Type Circulating Water, ppm Advantages Use

Nitrites About 200 to 500 ppm as Nitrate 1. Satisfactory iron protection by formation of 1. Susceptible to attack by oxidizing agents
gamma iron oxide film and certain bacteria
2. Does not protect copper
3. Requires very high concentration of nitrite in
the beginning

Silicates About 200 to 500 ppm as Silicate 1. Fairly good protection—used when chro- 1. Requires formation of iron-oxide to be effec-
mate, phosphates, or nitrites cannot be tive
used 2. Polymeric species of silicates are required
 1989 Chevron USA Inc. All rights reserved.

3. Not as good protection

4. Does not protect copper

Organic-High pH Depends on solids concentration in 1. Gives corrosion protection in order of 2-5 1. High pH means that for reasonable cycles
circulating water but generally 20-100 mpy (Chromate is <1 mpy) of concentration of the cooling water, solu-
ppm 2. Lower acid consumption bility of calcium and magnesium salts are
exceeded; acrylate or phosphonate acts as
3. No heavy metals
a scale inhibitor; if phosphonate injection
4. Fairly low toxicity, biodegradable (slowly) ceases for any reason, fouling will occur
almost immediately
2. High pH may mean increased cooling tower
wood deterioration; more susceptible to
fungus attack
3. High pH limits use of low cost biocides
4. These treatments can be expensive at high

Heat Exchanger and Cooling Tower Manual

solids concentration
5. Chlorine biocide cannot be used at high pH
due to accelerated cooling tower deteriora-
tion

Lignins and Tannins Highly variable (depending on the 1. Easily biodegradable and nontoxic 1. Does not give adequate protection at higher
quality of the wood products from 2. Low cost water temperatures encountered in most
which lignins and tannins are derived) refinery cooling systems
December 1989
 Fig. 2400-2 Corrosion Inhibitor Summary (4 of 4)

December 1989

Heat Exchanger and Cooling Tower Manual

Concentration in Disadvantages or Reasons for Caution in
Inhibitor Type Circulating Water, ppm Advantages Use

Lignins (or Tannins) and Zinc Variable 1. Improved corrosion protection compared to 1. Organic lignins and tannins still do not give
lignins alone good high temperature protection - even
with zinc present
2. Zinc in blowdown may be objectionable
from standpoint of meeting future regula-
tions
3. Lignins and tannins are highly variable in
composition; one product may work;
another may not
 1989 Chevron USA Inc. All rights reserved.

4. Biogrowth is a real problem in towers using

these products because lignin and tannin
are food for bacteria
5. Chlorine biocide cannot be used with these
inhibitors
6. Overfeeding results in iron-lignin fouling
Organo Sulfur-Zinc 50-100 ppm 1. Biodegradable (slowly); moderate toxicity 1. Expensive at the recommended concentra-
2. Good protection for copper and steel tion; lower concentrations may be sufficient
for adequate protection; depends on each
3. Corrosion protection relatively unaffected by application
process leakage into the cooling water
2. Contains zinc which will appear in the blow-
down water
3. Incompatible with chlorine biocide—forms a
residue which will foul exchangers

Note mpy = Mils Per Year Corrosion Rate

2400 Water Treatment System Design

2400-13
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

2460 Monitoring of Corrosion and Fouling Corrosion

2461 Corrosion
Test Exchangers
Various water treatment manufacturers offer test heat exchangers for evaluation of
corrosion control. The exchangers are offered free if corrosion inhibitors are
purchased. Although exchangers do show relative corrosion rates as water treating
conditions change, they are not recommended because they usually operate under
boiling conditions and at flow velocities considerably less than anything used in the
operating center’s exchangers. Thus, the conditions are not representative and may
lead to unnecessary purchase of chemicals.

Coupons
A coupon monitoring program is used in many towers to monitor corrosiveness of
the water. The big advantage of such a program is that it gives visual evidence of
corrosion. If corrosion rates start to drift upward (or move up rapidly), the cause can
be identified and corrected before serious trouble develops. The coupons also show
pitting tendencies and crevice corrosion (under Teflon washers) and show sludge
and silt deposition.

Continuous Measurement
Continuous measurement generally falls into two categories: the electric resistance
technique and the applied potential technique. With either technique, corrosion
measurements are made quickly without removing the sensing device from the
installation. Measurements are continuous, sometimes revealing sudden variations
in corrosion rates. The applied potential technique is the simplest and more reliable.
Contact ETD’s Materials Division or the Monitoring and Control Systems Division
for assistance in this area.

2462 Fouling
Antifouling programs can be monitored by the following methods:
• Test heat exchangers. (See cautions discussed above.)
• Specialized equipment giving a relatively quick readout of fouling tendencies
such as the monotube deposition tester. This tester has a window that permits
visual check of fouling. It also records temperature differences to reveal fouling
rates.

2400-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

2463 Microbiological Growth

Monitoring methods for cooling tower slime control programs have remained the
same for years. The principal method is macroscopic and is strictly observation of
the surfaces in the system.
The plate count is a method for measuring the density of living bacteria in a sample
of water. When carried out according to the procedure in the standard methods for
examination of water and waste water, the result is called a standard plate count.
The plate count correlates well with the microbiological growth.

2470 Automatic Control Systems

The use of automatic controls can increase the reliability of a cooling tower chem-
ical treatment program and is very important for success of nonchromate treatment.
The factors affecting the relative importance are listed in Figure 2400-3. For more
details on the actual control packages, see Section 2500 of this manual.

Fig. 2400-3 Factors Affecting the Decision to Use Automatic Controls

Low Chromate/Nonchromate 
Chromate Treatment Treatment
1. Continuous Corrosion Corrosion rates are usually below 1 Corrosion monitoring highly desirable
Measurement mil/year; wide latitude in CrO4 to regulate proper inhibitor concentra-
concentration available to attain tion to achieve <5 mpy corrosion
corrosion rates <5 mpy; corrosion rates
monitoring not required
2. Continuous Deposition Chromate treatment provides good Risk of fouling the system—moni-
Measurement antideposition control—monitoring not toring almost essential
required
3. Biological Monitoring Not essential Higher pH enhances biogrowth—
monitoring desired to avoid exces-
sive slime growth
4. Continuous Control of Desirable to reduce inhibitor cost but Constant inhibitor concentration
Inhibitor Feed Rate not necessary for success of program essential for success of treatments
and to avoid system fouling
5. pH Control Necessary to prevent scaling; fairly Necessary to avoid fouling of system
wide range and scale deposition
6. Continuous Blowdown Desirable to save chemicals Necessary to hold chemical cost to
acceptable level

2480 Sources of Reused Water for Makeup to Cooling Towers

Waste water reuse is an important approach to water conservation and pollution
control. Cooling towers using reused makeup water require more careful attention
than those having good quality fresh water. Common sources of reused water are
discussed below.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-15
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

2481 Boiler Blowdown or Concentrated Effluent from Reverse Osmosis Water

Treater
Cascading of these waters as cooling tower makeup should be considered. Boiler
blowdown is low in hardness and relatively low in dissolved solids. However, boiler
blowdown will be significantly hotter than return cooling water, so the tower may
need additional duty or the blowdown will need to be cooled.
The concentrated effluent (reject) from a reverse osmosis plant is normally only at a
maximum of four cycles of concentration. It can ideally be used in a cooling tower
operating at 12 cycles of concentrations.

2482 Treated Plant Waste Water

The use of treated plant waste water for cooling tower makeup conserves water and
reduces the net effluent discharge. The characteristics of the treated water, e.g., total
dissolved solids, biological oxygen demand (BOD), metal, etc., will determine its
potential for reuse. Additional controls may be required for increased corrosion,
microbiological fouling, and foaming. Several of our refineries are reusing treated
effluent as cooling tower makeup.

2483 Municipal Secondary Treated Effluent

The use of municipal secondary-treated waste water may increase biological
fouling, calcium phosphate scale, and foam. This could result in substantial increase
in treatment chemical costs and would require efficient control of operating parame-
ters. Because of high orthophosphate levels in this water, external phosphate
removal may be necessary. There is also the risk of additional load on the refinery
treatment system due to the presence of other contaminants in the municipal waste
water. For example, heavy metals in the municipal waste water may make it diffi-
cult for the operating facility to meet discharge limitations on heavy metals.

2484 Process Water

Generally, process water is not recycled directly back to cooling towers without
considerable pretreatment. Process waters include desalter water, stripped sour
water and process wash water. Desalter water and process wash waters are not
considered suitable for reuse as cooling tower water because they contain oil, high
total dissolved solids, and other contaminants. Sour water stripper bottoms (SWSB)
may contain phenol, total dissolved solids, and traces of ammonia and hydrogen
sulfide. SWSB may in some cases be suitable for reuse as cooling tower makeup.
However, the stripping efficiency for sulfide and ammonia removal is critical to its
reuse. Sulfides reduce chromate inhibitors to ineffective chromium hydroxide
sludge. Hydrogen sulfide is also corrosive to cooling water equipment. Ammonia
can cause poor pH control within the cooling system, is corrosive to copper alloys,
deactivates chromate based corrosion inhibitors, and, as a nutrient, contributes to
biological fouling. Dissolved organics such as phenols in the SWSB may also

2400-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2400 Water Treatment System Design

increase biological fouling. SWSB may also contain polysulfite if this is used for
cyanide corrosion control. This may pose possible sulfur plugging problems.
Before using any of the above waste water streams as cooling water, a complete
analysis and a study of the possible adverse effects of the contaminants should be
done.
Another possible route to consider would be side stream softening with lime and
soda ash. This treatment removes silica and calcium and magnesium hardness,
permitting higher cycles of concentration. To date, Chevron has not installed any of
these systems, but they should be evaluated as water sources and effluent require-
ments become more limiting.
For some added information on sources of reused water, see the section “Water
Reuse” in Power Magazine Special Report, “Cooling-Water Treatment for Control
of Scaling, Fouling, Corrosion” in Appendix J.

2490 Choosing a Water Treatment System Supplier

The following information may be helpful in purchasing chemicals and services for
treating cooling water:
1. Chemicals can be purchased as individual components for a relatively simple
treatment. However, for complex formulations, it is advisable to purchase the
chemicals from suppliers directly in bulk quantities. The cost of these treat-
ments usually includes the cost of services to be provided by a technical service
representative of the supplier. This is an important aspect and must be weighed
while comparing proposals. The representative may be able to seek exceptional
technical guidance from his “home office” for his clients. Representatives
should be required to check the system at regular intervals.
2. There is a wide choice of treatments available, and costs of chemicals among
suppliers are competitive. There are two approaches:
a. Invite several companies to submit their proposed recommendations based
on a survey of the cooling system, including its design, metallurgy, oper-
ating temperatures, makeup water quality, treatment being used, and the
condition of the system with respect to fouling and corrosion. This proce-
dure will result in a number of different approaches being submitted by the
suppliers.
b. Work with the chemical vendor considered the best one in the area, in a
“partnering alliance.” The target is quality improvement (using statistical
evaluation) for all the chemical requirements for the entire process plant.
3. The supplier must provide chemical formulation of the treatment recom-
mended, including the toxicity, BOD level, etc.
4. All suppliers must provide references of successful similar applications. Differ-
ences in operating conditions should be highlighted.

December 1989  1989 Chevron USA Inc. All rights reserved. 2400-17
2400 Water Treatment System Design Heat Exchanger and Cooling Tower Manual

5. Caution must be exercised before accepting the claims made by the suppliers
regarding the efficiency of any treatment. Suppliers should not be allowed to
test their products at our expense.
6. Complete details should be included to compare the total treatment costs on a
common basis (say, chemical costs/1,000 gallons of blowdown water).

2400-18  1989 Chevron USA Inc. All rights reserved. December 1989
2500 Chemical Injection System Design

Abstract
This section discusses the process control and design of a chemical injection system
for a cooling tower. Included are discussions of automatic controls; design consider-
ations; types of containers for feeders and storage; location of facilities; chlorine
injection facilities, with sizing, piping, and location; system commissioning; and
safety requirements.

Contents Page
2510 Process Control 2500-2
2520 Automatic Controls 2500-2
2521 Instrumentation
2522 Packaged Controls
2530 Chemical Injection System Design 2500-3
2531 Introduction
2532 Chemical Entry
2533 Design Considerations
2540 Chlorine Injection Facilities 2500-6
2541 Introduction
2542 Nature of Chlorine
2543 Design and Process Considerations
2544 System Commissioning

December 1989  1989 Chevron USA Inc. All rights reserved. 2500-1
2500 Chemical Injection System Design Heat Exchanger and Cooling Tower Manual

2510 Process Control

Figure 2500-1 shows the common components of a typical cooling tower process
control scheme. The mechanisms of corrosion and fouling and how they are
controlled are discussed in Section 2400. Consult the ETD Monitoring and Control
Systems Division for help in this area.
Fig. 2500-1 Typical Cooling Tower Process Control

2520 Automatic Controls

2521 Instrumentation
Automatic units ensure control of critical variables at the optimum values, resulting
in an efficient system at a minimum operating cost. Close regulation of parameters
is also important for minimizing waste disposal. Many practical features can be built
into the instrumentation to achieve maximum dependability. Such features include:
1. Sensing units located at the cooling tower to minimize response time.
2. Preamplification of the signal at the sensing point so it can be transmitted any
distance without interference.
3. Instruments located in the control room or other locations where personnel are
available.
4. Dependable, solid-state electronic recorders to observe trends.

2500-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2500 Chemical Injection System Design

Generally, an automatic system provides a package to measure and control the

inhibitor level (chromate or nonchromate), conductivity for blowdown control, pH,
and, in many cases, corrosivity. The corrosivity device serves as a backup or as an
override on the system to provide an alarm or perform certain functions that are
programmed in for rapid restoration of normal conditions.
Chlorination programs can be automated by using ORP (oxidation reduction poten-
tial) instruments. Liquid nonoxidizing biocides are frequently injected into the
cooling system automatically by timing devices which operate proportioning feed
pumps.

2522 Packaged Controls

Both Uni-Loc and Magna Corporation offer packaged control systems for pH, blow-
down, and inhibitor addition. Controls should be designed so that they may be
expanded to use the various control options available. As environmental consider-
ations restrict, and probably eliminate, the discharge of chromates, more sophisti-
cated cooling tower controls will be needed.

2530 Chemical Injection System Design

2531 Introduction
This section discusses the design, equipment selection, and installation of chemical
injection facilities for additives injected from:
• Drums
• Semibulk portable containers
• Bulk storage tanks
The discussion applies to facilities used for purchased-outside products typically
intended as additives to process plant streams, such as antifoulants, antifoams, anti-
oxidants, corrosion inhibitors, boiler feed water additives, acids, caustics and
cooling tower additives.

2532 Chemical Entry

We are legally required in many locations to document and provide specific mate-
rial hazard information for all personnel who might be exposed to the hazardous
materials. These requirements include:
• Reviewing and approving all chemicals which are used in a plant
• Posting the identity of the material and the appropriate hazard warning on the
container
• Providing a Material Safety Data Sheet before the material is used

December 1989  1989 Chevron USA Inc. All rights reserved. 2500-3
2500 Chemical Injection System Design Heat Exchanger and Cooling Tower Manual

The engineer is responsible for verifying that all of the local rules and regulations
have been complied with. Consult the local safety and environmental specialists for
the most up-to-date rules and regulations.

2533 Design Considerations

Richmond Refinery Drawings D-253080, B-603003, and B-603005 are Standard
P&IDs for bulk, semibulk, and barrel chemical injection facilities. These drawings
are attached at the end of this section as Figures 2500-2, 2500-3, and 2500-4. Full
size prints and CAD files are available from the Richmond Refinery Drafting
Department.
Note that the attached P&IDs were generated at the Richmond Refinery for their
Process Hazard Safety Committee. Consequently, some of the information, such as
references to refinery instructions, is specific to Richmond.

Types of Containers
The choice of containers for feeding chemicals, such as acid, caustic, inhibitor,
dispersant, biocide, etc., depends on the type of controls, the desired inventory, ease
of replenishing stock, handling problems, cost of chemicals in bulk and drum lots,
and, most importantly, the safety and environmental considerations involved.

Bulk Storage Tanks (see Figure 2500-2 drawing D-253080)

For large quantities and where tank truck deliveries are required, bulk storage tanks
are the best option. Often vendors will supply the tank; however, the tank must
sometimes be upgraded to meet local standards.
Steel storage tanks should be designed to UL-142 or equal (unless a pressure vessel
is required, then use ASME Code, Section VIII, Division 1). All connections to the
vessel must be 3/4-inch minimum size and must be seal-welded or flanged. Vessels
containing flammable material require a bottom fill nozzle and a vent system sized
for emergency venting (per NFPA 30 and API STD-2000).
High-density polyethylene (HDPE) storage tanks may be used if they are compat-
ible with the material stored (HDPE should not be used to contain combustibles).
HDPE tanks should be designed to Poly Cal Plastics’ specification PCPP48614, or
equal, for cross-linked polyolefin tanks.

Semibulk, Portable Returnable Container (see Figure 2500-3 drawing B-603003)

Semibulk containers offer the advantages of returnable containers supplied by the
vendor and are easily handled by a forklift or crane. Skid-mounted standard units
with attached pumps offer great convenience at low first cost and eliminate or
reduce capital expenditures. Material unit cost is usually cheaper than for drums.
However, not all vendors offer this option. Note that some upgrade of vendor-
proposed piping and equipment is usually required to meet local standards.

2500-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2500 Chemical Injection System Design

Drums (see Figure 2500-4 drawing B-603005-0)

Use of drums should be avoided unless the quantity to be used is very small and
returnable, or bulk facilities cannot be justified. Drums may be an acceptable alter-
nate, however, for temporary uses (test runs). Drums are costly to dispose of, waste
chemicals because they do not empty completely, and may result in higher
employee exposure. Drum handling is difficult and can cause injuries.

Location of Facilities
The facilities should be located in an easily accessible area, preferably around the
perimeter of the plant. Give consideration to the access of the truck delivering the
bulk shipment, ease of operator access on normal rounds, and to the nature and
potential hazard of the material (i.e., decomposition of the material if near a fire).

Injection Pumps
The injection pump(s) should be located as close as possible to the vessel while still
maintaining the proper clearances. For bulk storage facilities, if possible, elevate the
pump(s) approximately 3 feet for ease of operation and maintenance. Milton Roy,
Pulsafeeder, or Williams pumps may be used, depending on the application.

Instrumentation
A combination level gage/rate meter is allowable; however, it should be heavy duty
to avoid spills resulting from mechanical damage. The suggested model is a
KENCO calibration gage (or equal). Since each facility’s pump flow rate is
different, the KENCO model number must be obtained from the vendor.

Piping
Consider seal welded or socket welded piping 3/4-inch minimum. The type of
chemical may require a material other than steel. The ETD Materials Division or the
material vendor should be consulted. Some injection quill designs are shown on the
referenced P&ID. To facilitate pump hookup or to decrease costs in long discharge
runs, stainless steel tubing (or another alloy) is acceptable (1/2-inch O.D.,
0.065-inch minimum wall).

Lighting
Provide adequate lighting for the operating and maintenance activities required.
Consult the local safety engineer for the level of lighting required.

Berms
The decision to berm the facilities should be based on the volume, and on the
impact on the effluent treatment system in the event of a spill. The intent is to
contain any spills to a confined area and to keep the material out of the drainage
system. Berms should be large enough to hold all contents of the largest tank in case
of rupture and should contain any slight spillage from the loading spot and from
taking the facilities out of service for maintenance. A 2-inch minimum gate valve
(normally closed) should be provided in the berm, through which rainwater can be

December 1989  1989 Chevron USA Inc. All rights reserved. 2500-5
2500 Chemical Injection System Design Heat Exchanger and Cooling Tower Manual

drained from the bermed area. Adequate washdown facilities should be provided
near the facilities.

Safety Requirements
Each facility must be evaluated to determine if the following safety equipment is
needed:
• Safety showers and eyewash stations a minimum of 10 feet from any potential
leak source and a maximum of 50 feet (unobstructed) from the facility for
corrosive chemicals
• Safety signs
• Barriers or guard posts around the perimeter per Standard Drawing GA-S99975
(located in the Civil and Structural Manual)
• Scott Air-Pak
• Posting of instructions for the driver of the bulk delivery truck

2540 Chlorine Injection Facilities

2541 Introduction
This section gives guidance on the design of cooling tower chlorine injection
systems. In order to complete a detailed design, you must understand the hazards,
regulations, and safe handling guidelines for chlorine. These subjects are discussed
in detail in the Piping Manual, Section 1500, and in general terms in this section.

2542 Nature of Chlorine

Chlorine is a nonflammable gas which is liquified under pressure. It will react
chemically (often vigorously) with almost all elements and with many inorganic and
organic substances, usually with the evolution of heat. The gas has a characteristic
odor and a greenish-yellow color, and is about two and one-half times as heavy as
air. Thus, if it escapes from a container, it will seek the lowest level in the building
or area where the leak occurs. Although dry chlorine will not corrode most metals, it
is very corrosive (forms HCl) when moisture is present. Therefore, never use water
on a chlorine leak because resulting corrosive conditions always make the leak
worse.
There are potential health hazards associated with the use of chlorine. Chlorine gas
is primarily a respiratory irritant. In sufficient concentration the gas irritates mucous
membranes, respiratory system, and skin. In extreme cases, the difficulty of
breathing may increase to the point where death can occur from suffocation. Never
attempt a rescue without adequate respiratory protection. Liquid chlorine, in contact
with skin or eyes, will cause burns. Consult the Material Safety Data Sheet for addi-
tional information.

2500-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2500 Chemical Injection System Design

2543 Design and Process Considerations

The primary use for chlorine is for cooling water treatment. Chlorination reduces
the growth of algae and fungi which, left untreated, can cause severe deposition
problems and cell attack on metals.
Chlorine normally comes in 150-pound or 1-ton containers. Drawings D-603001
and D-603002 are standardized P&IDs that can be used as a basis for designing
chlorine facilities for cooling water treatment. These drawings are attached at the
end of this section as Figures 2500-5 and 2500-6. Full-size prints and CAD files can
be obtained from the Richmond Refinery Drafting Department.

Sizing of the Facilities

The dependable continuous discharge rate of chlorine gas from a single 100-pound
or 150-pound cylinder without frosting under normal conditions (70°F) is about
1.75 pound/hr against a 35 psi back pressure. The rate for a 1-ton container is about
15 pound/hr under similar conditions.
If the gas discharge rate from a single container will not meet requirements, two or
more can be connected to a manifold and discharged simultaneously, but all
containers should be at the same temperature to prevent transfer of gas from a warm
container to a cool container. If a container becomes completely filled with liquid
and the container valve is then closed, the container may burst from hydrostatic
pressure.
Minimize the number of chlorine cylinders that are hooked up to the system. Addi-
tional cylinders and associated piping increase the risk of having a chlorine release.
Normally a 4-week supply should be adequate.

Where to Inject Chlorine

Chlorine should be injected into the tower basin on the opposite side of the tower
from the forebay to maintain a sterile condition in the tower basin. If chlorine is
added to the forebay, it will not reach the basin because it evaporates through aera-
tion as the water cascades down over the tower fill. In order to efficiently sweep the
basin, a chlorine distributor that extends the full length or width of the basin is
normally employed. This distributor should be located approximately 2 feet below
the basin water level.

Piping
Piping arrangements should be as simple as possible. Joints should be flanged or
welded with the number of flanged joints held to a minimum. Piping systems should
be well-supported and adequately sloped to allow drainage; low spots should be
avoided.
Construction materials and ratings for Monel pipe are shown on the attached P&IDs
for cooling water chlorination facilities. The cylinders are connected to the piping
system by a Chlorine Institute transfer hose. A pressure-reducing valve should be
installed between the cylinder and the regulator.

December 1989  1989 Chevron USA Inc. All rights reserved. 2500-7
2500 Chemical Injection System Design Heat Exchanger and Cooling Tower Manual

Piping Expansion and Protection

Suitable allowance should be provided for pipe expansion due to changes in temper-
ature. Liquid chlorine has a high coefficient of thermal expansion. If liquid chlorine
(containing no gas bubbles) is trapped between two valves, high pressure will
develop with an increase in the temperature of the chlorine and a rupture could
occur. For this reason the standard designs shown on the P&IDs have only one pair
of isolation valves (one main plus backup).

Preventing Liquid to the Chlorinator

No liquid chlorine can enter the chlorinator, as this usually damages the chlorinator
components. The following methods can be used to reduce the chances of liquid
chlorine entering the chlorinator:
• Install a pressure reducer (as shown on the P&IDs) to flash any residual liquid
• Have a 120 volt heater installed in the chlorinator (most can come factory-
equipped)
• Install a liquid trap in the piping just upstream of the chlorinator

Chlorinator Design
Chlorinators are usually sized to provide 1 ppm of chlorine continuously to the
circulating water. For example, for a system with a circulation rate of 13,000 gallons
per minute (156 million pounds per day) the chlorinator should be sized to deliver
approximately 156 pounds per 24 hours. A margin of 20 to 25% should be added to
the chlorinator design capacity.
Chlorinators are typically automated by equipping them with timers that permit
either a daily or several times weekly application. These timers should also permit
complete flexibility for the length of each application. However, the typical dura-
tion is somewhere between 2 and 8 hours. The timer is usually connected to a sole-
noid valve on the water supply to the chlorinator. Since a flow of water is necessary
to operate a chlorinator, interrupting this flow with a suitable automatic, full-
opening valve provides a simple and reliable method of turning the equipment on
and off.

Vacuum Eductor
The chlorinator is actuated by an inducted vacuum from an eductor. The eductor
uses the flow of water to pull a slight vacuum on the chlorine system. The chlorine
flow can be controlled by a variable rate rotameter.

Where to Locate Chlorine Containers

Exposure of containers to flame, intense radiation heat, or steam lines must be
avoided. If the fusible plug reaches 158°F the plug will melt and chlorine will
escape to the atmosphere. For cooling water treatment, the cylinders can be stored
outdoors but they should be located at least 50 feet from the tower or from any other
flammable materials. If they must be located closer than this, a building or fire wall
should be designed and constructed to protect all elements of the chlorine system

2500-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2500 Chemical Injection System Design

from fire hazards. Positive hold-down clamps should be provided for all chlorine
containers.

Miscellaneous Safety Equipment

The chlorine facilities should be well lighted and identified by “Cl2 Hazard” and
“No Parking” signs around the perimeter. Barriers of guard posts should also be
installed around the perimeter. An eyewash, Scott Air-Pak, protective clothing, and
shower should be located within 50 feet of the chlorine.

2544 System Commissioning

Cleanup After Fabrication
It is especially important to clean all portions of a chlorine system before use
because chlorine may react violently with cutting oil, grease, or other foreign mate-
rials. Cleaning may be accomplished by pulling a cloth saturated with trichloreth-
ylene (or other suitable chlorinated solvent) through each length of pipe.
Chlorinated solvents can produce serious physiological effects if not used in strictest
compliance with the solvent manufacturer’s safety recommendations. Never use
hydrocarbons or alcohols because residual solvent may react with chlorine.

Pressure Testing
After hydrostatic testing, the piping must be dried. This can be done by passing
steam through the lines from the high end until the lines are thoroughly heated.
While steaming, allow condensate and foreign matter to drain out. While the line is
still warm, nitrogen (or dry air) should be blown through the line until it is dry; this
may require several hours.
After drying, the system should be pressurized to 150 psi with dry air or nitrogen
and tested for leaks by application of soapy water to the outside of joints. Chlorine
gas may then be introduced gradually and the system tested for leaks. Never attempt
to repair leaks by welding until all chlorine has been purged from the system. When
all detectable leaks have been repaired, the system should be retested.

December 1989  1989 Chevron USA Inc. All rights reserved. 2500-9
2500 Chemical Injection System Design Heat Exchanger and Cooling Tower Manual

2500-10  1989 Chevron USA Inc. All rights reserved. December 1989
2600 Cooling Tower Operations/
Troubleshooting

Abstract
This section briefly discusses cooling tower commissioning and some suggestions
for cooling tower prestartup plans. Figure 2600-1 is a troubleshooting table for
mechanical/electrical components.

Contents Page
2610 Cooling Tower Commissioning 2600-2
2620 Troubleshooting 2600-2

December 1989  1989 Chevron USA Inc. All rights reserved. 2600-1
2600 Cooling Tower Operations/ Troubleshooting Heat Exchanger and Cooling Tower Manual

2610 Cooling Tower Commissioning

In order to minimize subsequent corrosion problems, it is important to properly
commission a cooling tower. This involves properly cleaning the tower, the circula-
tion piping, and heat exchangers prior to operation.
The following prestartup plans should be considered and, if appropriate, included in
the operating instructions:
1. Circulate sufficient cooling water through the tower for a period long enough to
leech out readily soluble materials in the fungicide used to impregnate the
wood. A piping bypass should be used during this operation in order to avoid
passing the circulating water through heat exchangers. Many fungicides contain
heavy metals that can create serious corrosion problems in the cooling system.
The length of time required for prestartup circulation will vary from 2 to 4
weeks depending on the circulation rate. The circulation rate and time should
be reviewed with the cooling tower manufacturer. The pumping rate must also
be chosen based on the selected pump(s) characteristics.
2. Blow down the cooling water system at the full design blowdown rate to elimi-
nate the last of the remaining scale and debris.
3. A detergent solution should be circulated through the system piping and heat
exchanger equipment. This cleaning step is designed to remove grease, dirt, and
loose mill scale. Usually, 24 to 48 hours are required for proper cleaning.
Following this step, the system should be thoroughly flushed and then treated
with a highly concentrated solution of corrosion inhibitor designed to rapidly
form a protective film on all metal surfaces. The tubes and shell should be
drained in so far as possible just before startup.
4. Special handling and/or treatment of blowdown and drainage may be required
during commissioning to avoid an effluent quality problem.

2620 Troubleshooting
Figure 2600-1 outlines common problem areas and possible solutions in the opera-
tion of a cooling tower.

2600-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2600 Cooling Tower Operations/ Troubleshooting

Fig. 2600-1 Troubleshooting Mechanical/Electrical Components (Courtesy of Power Magazine Special Report,
“Cooling Towers,” 3/73)
Trouble Possible Cause Remedy
Uneven water  Broken or plugged nozzles and/or distribu- Replace or repair defective parts; clean
distribution tion piping; broken fill; distributing pan out of distribution system and pump suction
level; excessive or uneven water flow screen; adjust water flow to design condi-
tions
Cold water too warm Overpumping; fill improperly installed; not Adjust water flow to design conditions;
enough air make sure all fill sections are intact and
properly installed; make certain motor hp is
correct; see if fill or eliminators need
cleaning; see if anything is obstructing inlets
or discharge
Excessive water drift Broken or plugged distribution system; Replace or clean nozzles; see that all fill
broken or missing drift eliminators; fan and eliminator sections are in place and
pitched above design; overpumping intact; pitch fan to design conditions; reduce
water flow to tower design conditions
Noisy gears and bear- Worn bearings or gear set; warped gearing; Check oil for level and contamination; adjust
ings in speed reducer low oil level; contaminated oil; protective oil shield; replace worn bearings
shield rubbing gear case; bearing fatigue
Excessive movement in Worn high-speed and low-speed bearings Replace worn bearings and oil seals; check
speed-reducer pinion and tooth contact of gears after replacing gears
low-speed shafts and/or bearings
Vibration in couplings and Misalignment of couplings; foreign matter Realign coupling and recheck alignment
drive shaft adhering to coupling; shaft out of balance after 30 days; tighten motor and speed-
bent or off-center; worn bearings or bent reducer hold-down bolts
shaft in motor or gear unit
Unusual motor noise Motor running single-phase; electrical Stop motor, try to restart (unit won’t start if
unbalance; worn bearings single-phased); check wiring, controls,
motor and all three lines, correct if required;
check lubrication; replace bad bearings
Motor won’t start Line trouble; single-phasing at start; Check source of power supply and correct;
improper connections; load too heavy check connections with circuit diagram;
disconnect motor to see if it starts without
load; reduce load or, if necessary, replace
motor with unit of greater capacity
Motor, motor-bearing Overload (measure load, compare with Check for excessive friction in motor drive
overheating nameplate rating); misalignment; excessive or unit; check for overvoltage, improper
end thrust; too much grease (ball or roller connections; realign set; reduce thrust from
bearing); insufficient lubricant drive or machine; relieve grease supply and
boost oil to points set by manufacturer

December 1989  1989 Chevron USA Inc. All rights reserved. 2600-3
2600 Cooling Tower Operations/ Troubleshooting Heat Exchanger and Cooling Tower Manual

2600-4  1989 Chevron USA Inc. All rights reserved. December 1989
2700 Inspection and Maintenance

Abstract
This section briefly discusses cooling tower inspections and maintenance and
includes a Cooling Tower Institute (CTI) checklist for inspection and a list of main-
tenance activities for key tower components.

Contents Page
2710 Inspection 2700-2
2720 Maintenance 2700-2

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-1
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2710 Inspection
Cooling tower inspections are usually carried out by experienced cooling tower
manufacturers or specialists. See Section 2150 for a list of possible contacts.
Included in this Section is a copy of Cooling Tower Institute’s Cooling Tower
Manual, Chapter 13, “Inspection of Cooling Towers.” It includes detailed check-
lists for inspection of all of the cooling tower components.

2720 Maintenance
Cooling tower maintenance is normally carried out by experienced cooling tower
contractors. See Section 2150 for a list of possible contacts.
Figure 2700-1 lists the common maintenance activities which should be scheduled
for cooling towers.

Fig. 2700-1 Preventive Maintenance Schedule for Key Tower Components

Seasonal  Seasonal 
Component Weekly Monthly Shutdown Startup
Fan motor — Check mounting bolts Clean and cover Lubricate, if required;
tighten mounting bolts
Speed reducer Check oil level, visual Check oil level, check Overfill with oil, clean Refill with new oil and
inspection mounting bolts and cover tighten mounting bolts
Fan wheel Visual inspection, Check bolts, check Clean and cover Clean and paint hub,
clean foreign matter “set screws” check blade pitch
Drive shaft Visual inspection Check shaft alignment Clean and cover Realign
Distribution Remove foreign Remove foreign Drain and clean Inspect and clean
system matter matter
Cold-water basin Remove foreign Remove foreign Drain and clean Inspect and clean
matter matter
Sump and screen Remove foreign Remove foreign Drain line and shut off Inspect and clean
matter matter
Overflow Visual check Visual check — Tighten bolts
Connections Visual check Tighten bolts — —

2700-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-3
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-5
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-7
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-9
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-11
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-13
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-15
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-17
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-18  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-19
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-20  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-21
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-22  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-23
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-24  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-25
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-26  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-27
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-28  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 2700 Inspection and Maintenance

December 1989  1989 Chevron USA Inc. All rights reserved. 2700-29
2700 Inspection and Maintenance Heat Exchanger and Cooling Tower Manual

2700-30  1989 Chevron USA Inc. All rights reserved. December 1989
3100 Specifying and Purchasing
Procedure

Abstract
This section outlines a suggested procedure for ordering heat exchanger equipment.
This is a general procedure that does not address local practices or administrative
and approval requirements. It may be modified for local preferences or used as is.

Contents Page
3110 Steps for Purchasing Heat Exchangers 3100-2
3120 Quotation Request 3100-2
3130 Bid Evaluation 3100-3
3131 Technical Analysis
3132 Cost and Delivery Analysis
3133 Commercial Terms
3140 Purchasing the Exchanger 3100-5
3150 Model Quotation Request and Requisition 3100-5

December 1989  1989 Chevron USA Inc. All rights reserved. 3100-1
3100 Specifying and Purchasing Procedure Heat Exchanger and Cooling Tower Manual

3110 Steps for Purchasing Heat Exchangers

Listed below are the major steps involved in purchasing heat exchanger equipment.
1. Obtain process requirements and physical and chemical properties of the fluids.
2. Determine operation and maintenance and construction requirements or
concerns.
3. Investigate site and utility conditions and restraints (e.g., space, power, cooling
water, underground interferences—pipe, old foundations, etc.).
4. Select the type of heat transfer equipment (shell and tube, air cooler, double
pipe, plate exchanger, etc.).
5. Prepare quotation request—see Model Quotation Request or Purchase Order,
Section 3150.
6. Complete the exchanger data sheet(s)—see Data Sheet Guide for details. (The
Data Sheet Guides are located after the appropriate specification.)
7. Prepare special drawings (not normally required).
8. Attach applicable standard drawings.
9. Forward bid package (items 5 through 8) to Purchasing.
10. Obtain quotations (by Purchasing).
11. Evaluate quotations, recommend vendor, discuss with client, and obtain appro-
priate expenditure approvals.
12. Prepare materials requisition—see Model Quotation Request or Purchase
Order.
13. Forward the approved bid analysis and revised (if needed) list sheets, specifica-
tions(s), data sheet(s) and drawings, through channels to Purchasing for place-
ment of the order.

3120 Quotation Request

Work with your local purchasing organization to obtain written quotations. Usually
4 to 6 weeks are needed. Purchasing is responsible for the selection of bidders.
However, you should recommend bidders for inclusion (or exclusion) when project,
technical, quality, or other considerations make it advisable.
Specify that the vendors complete the Company data sheet(s). Vendors often use
their own data sheets which can be confusing when making bid comparisons, or
they omit valuable information.
On equipment with which you are unfamiliar, you should go over the technical
aspects of your bid request with local technical specialists, or with technical special-
ists of the Engineering Technology Department.

3100-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 3100 Specifying and Purchasing Procedure

3130 Bid Evaluation

Four key criteria must be satisfied to select a vendor: technical features, cost,
delivery, and commercial terms.
For “Engineered” equipment, such as complex shell and tube and air cooled heat
exchangers, the engineer performs the technical, cost, and delivery analysis. The
result of his work will be to recommend the vendor whose bid: (1) meets all the
requirements of the bid request, and (2) results in the lowest evaluated overall cost
to the Company. The major considerations in reaching these results are covered in
the following section.

3131 Technical Analysis

Standard forms are available to facilitate the technical analysis of the bids. Use
EXH-EF-20 for shell and tube exchangers and EXH-EF-20-A for air coolers. These
forms are located in the Standard Drawings and Forms section of this manual.
Use the appropriate sections of this manual for guidance in making the technical
analysis. The technical items which should be considered in the bid analysis along
with the appropriate section to refer to for advice are listed below.
• Exchanged Duty
• Heat Transfer Rate—Service, Clean—Section 210, Section 300 (for specific
services)
• Mean Temperature Difference (MTD)—Section 210
• Surface Area (bare and finned)—Section 210, Section 480
• Fouling Resistance—Section 240, Section 300
• Pressure Drop—Section 220
• Design Pressure and Temperature—Section 510
• Corrosion Allowance—Section 800
• Materials of Construction (shell, tubes, fans, etc.)—Section 800 and 
specifications
• Codes—Section 120
• TEMA Size and Type (for shell and tube exchanger)—Section 410
• Extended Surface Specifications (if applicable)—Section 270
• Tube and Shell Side Velocities (where applicable)—Section 220
• Number of Shell and Tube Side Passes (where applicable)—Section 440
• Number of Bays (for air cooled exchangers)—Section 600
• Platforms and Ladders (for air cooled exchangers)—EXH-MS-1826
• Motor and Fan Specifications (for air cooled exchangers)—EXH-MS-1826
• Noise requirement (for air cooled exchangers)—EXH-MS-1826
If any exceptions are stated or alternatives proposed, they must be clearly under-
stood and they must meet the technical, maintenance, and performance require-
ments of the quotation request.

December 1989  1989 Chevron USA Inc. All rights reserved. 3100-3
3100 Specifying and Purchasing Procedure Heat Exchanger and Cooling Tower Manual

3132 Cost and Delivery Analysis

Once the bidder’s proposal(s) is understood, you can proceed with the cost and
delivery analysis. To save effort, first run a rough screening of the bids to determine
if some may be eliminated because of unacceptable technical, commercial, and/or
cost differences. Then evaluate (cost) the acceptable differences (if any) among the
remaining bids that are within the scope of the specifications and that are of value to
the Company (lower energy consumption, higher efficiency, mechanical features,
etc.). Finally, to make the bids comparable, apply these estimates to the appropriate
bid(s) to determine which offering provides the lowest overall evaluated cost to the
Company.
The cost and delivery analysis can be made using locally available forms, such as
PD 137/PD 135, Summary of Bids, available from Purchasing.
When evaluating vendors’ quotations, the following cost and delivery items should
be considered:
• Equipment Cost
• Freight
• Spare Parts (requirements and costs) Tools, Test Rings, etc.
• Export Preparation
• Penalty for Progress Payment
• Escalation (clause)
• Differential Piping Cost
• Erection Cost
• Operating Cost—This is especially important for water cooled exchangers.
• Estimated Cost to Meet Specifications
Once you have evaluated the quotations, decided on the lowest evaluated cost
offering, and recommended a vendor, talk again with operating or project represen-
tatives. Make sure they understand and agree with the reasons for your selections,
especially if your decision conflicts with their preferences.

3133 Commercial Terms

Commercial terms are Purchasing’s responsibility. They often add special require-
ments to the bid package besides the standard preprinted terms and conditions that
are always included. After negotiations with the lowest bidder(s) (during the course
of the bid analysis) these may be modified and/or others added.
The “low bidder” is then that bidder whose proposal: (1) meets the minimum
requirement of the specification(s) (including quality and reliability); (2) results in
lowest evaluated overall cost to Company; and (3) contains acceptable commercial
terms and delivery.

3100-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual 3100 Specifying and Purchasing Procedure

3140 Purchasing the Exchanger

Prepare the proper requisition form and forward it (through local channels) to
Purchasing complete with an approved bid analysis and all the documents in the
original bid package modified and/or updated with information received from the
recommended vendor.
The engineer’s job is not done when the exchanger is placed on order. Vendor
followup is important. You must make sure the exchanger is being built to 
specifications—through review of vendor’s drawings. The engineer must follow 
the fabrication progress closely so that any schedule slippage is noted and acted
upon early.
Inplant expediting through Purchasing is another way to follow the progress of an
order, but it is costly. If inspection is required, read your inspector’s reports and
keep in touch to make sure he is doing his job correctly and that you are getting the
required tests and inspection. Section 900 discusses inplant inspection in more
detail.

3150 Model Quotation Request and Requisition

This model has been developed to aid in the preparation of quotation requests and
requisitions for shell and tube heat exchangers, and for air cooled heat exchangers.
The entire text will not necessarily apply to every quotation request or requisition,
and all requirements are not included in this model. The user must select the appli-
cable portions for each request or requisition. Where blanks appear, the user must
choose wording applicable to the particular equipment involved. In many cases
additional items will need to be added to some or all of the sections.

December 1989  1989 Chevron USA Inc. All rights reserved. 3100-5
3100 Specifying and Purchasing Procedure Heat Exchanger and Cooling Tower Manual

3100-6  1989 Chevron USA Inc. All rights reserved. December 1989
References

1. American Petroleum Institute (API), API Recommended Practice 632, 

“Winterization of Air Cooled Heat Exchangers,” First Edition, April 1988.
2. American Society of Mechanical Engineers (ASME), ASME Boiler and 
Pressure Vessel Code, Section I and Section VIII.
3. COPI General Specifications, San Ramon, California.
4. Heat Transfer Research Inc. (HTRI), HTRI Design Manual.
5. HTRI Heat Exchangers Corporate Programs.
6. Tubular Exchanger Manufacturers Association (TEMA), Standards of the
Tubular Exchanger Manufacturers Association, Seventh Edition, 1988.

December 1989  1989 Chevron USA Inc. All rights reserved. Reference-1
 References Heat Exchanger and Cooling Tower Manual

Reference-2  1989 Chevron USA Inc. All rights reserved. December 1989
Appendix B Single-Phase Heat Transfer Coefficients

This appendix presents correlations for single-phase convective heat transfer inside
tubes and outside bare tube bundles.

Contents Page
B1.0 In-Tube Turbulent Heat Transfer B-2
B2.0 Turbulent Entry Region Effects B-3
B3.0 In-Tube Laminar Heat Transfer B-3
B4.0 In-Tube Transition Between Laminar and Turbulent Flow B-4
B5.0 Bundle Cross-Flow Heat Transfer B-4

March 1994  1994 Chevron USA Inc. All rights reserved. B-1
Appendix B Single-Phase Heat Transfer Coefficients Heat Exchanger and Cooling Tower Manual

B1.0 In-Tube Turbulent Heat Transfer

The HTRI correlation for turbulent in-tube heat transfer is:

0.79
Nu = .025  Re  Pr 0.42   for Re  10,000
(Eq. B-1)
where:
Nu = hi(di/12)/k = Nusselt Number (dimensionless)
Re = V(di/12)/= Reynolds Number (dimensionless)
Pr = 3600Cp/k = Prandtl Number (dimensionless)
hi = Inside heat transfer coefficient, Btu/hrFft2
di = Tube inside diameter, inches
Cp = Fluid specific heat, Btu/lb°F
k = Fluid thermal conductivity, Btu/hr°Fft
 = Fluid density, lb/ft3

V = Fluid velocity, ft/sec

 = Fluid bulk viscosity, lb/ftsec

w = Fluid wall viscosity, lb/ftsec

(Note:  in lb/ft sec = 0.000672   in centipoise)

For liquids,  = (/w)m
m = 0.11 for heating the tube side liquid
m = 0.167 for cooling the tube side liquid
For gases,  = (Tbabs/Twabs)n
Tbabs = Absolute temperature of the bulk gas, R
Twabs = Absolute temperature of the tube wall, R
n = 0.5 for heating the tube side gas
n = 0.0 for cooling the tube side gas

B-2  1994 Chevron USA Inc. All rights reserved. March 1994
Heat Exchanger and Cooling Tower Manual Appendix B Single-Phase Heat Transfer Coefficients

B2.0 Turbulent Entry Region Effects

In turbulent flow, the heat transfer in the first few diameters from the inlet is higher
than for fully developed flow given by Equation B-1. This is of no consequence in
overall thermal performance. However, locally enhanced heat transfer near the tube
entrance governs the overall geometry and mechanical integrity of shell side boilers
with hot gas in the tubes in Hydrogen Plants, Fluid Catalytic Cracking Units, Sulfur
Plants, some Sulfuric Acid Plants and Ethylene Plants. Heat transfer is enhanced
about a factor of two in smooth tubes without ferrules. With ferrules, the enhance-
ment is more and is correlated by:

Numax / Nu = 22.5  (di/df)0.68 Re-0.21  1

(Eq. B-2)
where:
df = Ferrule I.D., inches
The effect of locally enhanced heat transfer in boilers cooling very high-tempera-
ture gases is discussed in Sections 350 and 390.

B3.0 In-Tube Laminar Heat Transfer

In-tube laminar flow in heat exchangers is always uneconomical, and sometimes
leads to “viscosity plugging” (Sections 213 and 610). Laminar heat transfer in tubes
may be characterized by:

Nu  3.65 for Re  2000

(Eq. B-3)
This is the asymptotic limit for constant property fluids.
Laminar heat transfer is enhanced in the entry region where thermal boundary layers
are developing, where the wall viscosity is much less than the bulk viscosity, and
where extreme transverse temperature and density gradients cause transverse natural
circulation eddies. HTRI Design Manual, Volume I, Section C3, presents correla-
tions that include these effects. They are not included here because laminar flow
should be avoided. An approximate magnitude for laminar flow heat transfer is
needed to estimate heat transfer coefficients in the transition region between laminar
flow and fully turbulent flow.

March 1994  1994 Chevron USA Inc. All rights reserved. B-3
Appendix B Single-Phase Heat Transfer Coefficients Heat Exchanger and Cooling Tower Manual

B4.0 In-Tube Transition Between Laminar and Turbulent Flow

Turbulent flow Equation B-1 applies for Re  10,000. Laminar flow Equation B-3
applies for Re  2000. For 2000  Re  10,000, the heat transfer coefficient may be
estimated by linear proration with respect to Re as follows:

Nu =   Nu2000 +(1 - )Nu10000

(Eq. B-4)
where:
Nu2000 = Nu at Re=2000 and Nu10000 = Nu at Re=10,000
and,  = 1.25 - Re/8000

B5.0 Bundle Cross-Flow Heat Transfer

The HTRI correlation for flow across staggered tube bundles is:

Nux = 0.38 Rex0.60  Pr0.33   for Rex  4000

(Eq. B-5)
where:
Nux = ho(do/12)/k (dimensionless)
Rex = Vx(do/12)/ (dimensionless)
ho = Outside heat transfer coefficient, Btu/hr°Fft2
do = Tube outside diameter, inches
Vx = the cross flow velocity between tubes at an average section in the
bundle, ft/sec
The cross flow area and the fraction of the total flow that crosses the bundle must be
known to estimate cross flow velocity.
The cross flow area may be calculated as follows:
Ax =    4  0.5  D b  1 – d o  P t BS  144 for 30 degree and 90 degree layouts
Ax =  2  4  0.5  D b  1 – d o  P t BS  144 for 45 degree layouts
Ax =    3  0.5  D b  1 – d o  P t BS  144 for 60 degree layouts

where:
Ax = crossflow area, ft2
Db = bundle diameter, inches
Pt = tube pitch, inches
BS = baffle spacing, inches
The flow fraction that crosses the bundle is usually 70 to 75% of the total flow. A
precise estimate requires computer analysis.

B-4  1994 Chevron USA Inc. All rights reserved. March 1994
Appendix C Two-Phase Heat Exchangers

This appendix presents the HTRI flow regime map for two-phase flow and a simple
pseudo single-phase model for two-phase heat exchangers that is valid for annular
or churn flow.
For annular or churn flow regimes, liquid coats the heat transfer surfaces and a
continuous gas phase flows in the core. Liquid droplets are entrained in the core
flow.
The phrase annular flow is descriptive of the shape of the liquid film at the tube
surface for in-tube flow. Churn flow describes the corresponding flow regime for
flow across tube bundles. HTRI uses the phrase annular flow to describe a contin-
uous gas phase core flow with liquid wet tubes for both shell and tube side flows.
The HTRI flow regime map for shell and tube side flow is shown in Figure C-1.
Fig. C-1 HTRI Flow Regime Map

The abscissa Rlh in Figure C-1 is given by Equation C-1.

Rlh = [ 1 + (l/v)Y/(1-Y) ]-1

(Eq. C-1)

December 1989  1989 Chevron USA Inc. All rights reserved. C-1
Appendix C Two-Phase Heat Exchangers Heat Exchanger and Cooling Tower Manual

where:
l = Liquid density
v = Vapor density
Y = Weight fraction vapor
The ordinate in Figure C-1 is given by Equation C-2 for tube side flow and
Equation C-3 for shell side flow.
Tube side:

Cgt = (1/Gt)[(di/12)gv(l-v)(1-Y)/Y]1/2
(Eq. C-2)
Shell side:

Cgs =(1/Gx)[((Pt-do)/12)gv(l-v)(1-Y)/Y]1/2
(Eq. C-3)
where:
Gt = Mt/At = tube side mass flux, lb/hrft2
Gx = (MtFB/Ax)= shell side mass flux, lb/hrft2
Mt = Total flow rate, lb/hr
At = Total tube side flow area, ft2
Ax = Total shell side flow area, ft2 (See Appendix B.)
FB = Fraction of flow that crosses bundle (HTRI “B” stream) 
(Typically 0.6 for two-phase shell side flow)
di = Inside tube diameter, inches
do = Outside tube diameter, inches
Pt = Tube pitch, inches
g = Gravitational constant = 4.17108 ft/hr2
Two-phase heat exchangers that primarily exchange sensible heat (e.g., feed/effluent
exchangers in hydroprocessing plants) may be modeled as a single phase exchanger
as described below, provided the flow regimes on the shell and tube sides are
annular (Cgt and Cgs 0.3). Economic velocities for two-phase mixtures given in
Section 220 of this manual are usually in the annular flow regime.

C-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix C Two-Phase Heat Exchangers

C1.0 Pseudo Single Phase Model

In the annular flow regime, heat transfer resistance between the gas and the liquid is
negligible and can be ignored. The governing resistance is convection and conduc-
tion in the liquid film covering the heat transfer surface. This type of flow can be
modeled as a pseudo single-phase fluid using “no-slip” mixture density and
velocity, mixture heat capacity (or enthalpy), liquid viscosity, and liquid thermal
conductivity. “No-slip” means that the liquid and gas are assumed to flow at the
same velocity. Mixture properties are calculated as follows:
Mixture density:

m = v / [Y + (v/l)(1-Y)]
(Eq. C-4)
Mixture specific heat:

Cpm = (Y)Cpv + (1-Y)Cpl

(Eq. C-5)
where Cpv and Cpl are specific heats of the vapor and liquid respectively.
Mixture velocity:

Vm = Mt/(Afm)
(Eq. C-6)
where Af is the flow area (At or Ax).
These pseudo single-phase fluid properties can be input to the HTRI single phase
simulation program (ST) or used with the simpler methods described in this manual.
This approach has been validated against Chevron in-plant test data for well mixed
flow only.

December 1989  1989 Chevron USA Inc. All rights reserved. C-3
Appendix C Two-Phase Heat Exchangers Heat Exchanger and Cooling Tower Manual

C-4  1989 Chevron USA Inc. All rights reserved. December 1989
Appendix D Shell Side Hydrocarbon Boiling Correlations

This appendix gives boiling heat transfer correlations for horizontal shell side
hydrocarbon boiling with internal circulation. These correlations were developed by
Heat Transfer Research Inc.

Boiling Heat Flux

Single tube maximum (dryout) heat flux:
q1max = 664Pc Pr0.28 (1-Pr)0.73
and:
Fqbq1max = Maximum bundle flux, qbmax
FqdoFqbq1max = Incipient dryout heat flux, qbdo (steam heated)
FqshFqdoFqbq1max = Incipient dryout heat flux, qbdo (sensible heat medium)
Fof = Operating flexibility factor (usually 0.8)
Fqu = Uncertainly factor (0.7)
Recommended maximum design heat flux for bundles:
qdmax = FofFquFqshFqdoFqbq1max
where:
Fqb = qbmax/q1max = e-x
= Factor for maximum bundle heat flux
and x = (0.22) [(1-Gb)/Gb]2/3
and Gb = (Db)L/(144A) = Db/(ntdto)
Fqdo = 33.3 Gb - 0.632, Limit: 0.5  Fqdo  0.7
= Factor for bundle dryout with steam heating
Fqsh = (1-e-N)/N
= Heat flux correction for sensible heat medium
where:
N = UcA/(MCp)
and
Uc = Overall heat transfer coefficient
A = Heat transfer area
M = Mass flow rate of heat medium
Cp = Specific heat of heat medium

December 1989  1989 Chevron USA Inc. All rights reserved. D-1
Appendix D Shell Side Hydrocarbon Boiling Correlations Heat Exchanger and Cooling Tower Manual

Boiling Heat Transfer Coefficients

Single tube boiling heat transfer coefficient:
h1 = 0.046 Pc0.48 q2/3 FhpFhc
where:
Fhp = 2.733(0.12 + Pr0.4)/[(1-Pr)0.3] (= 1 for Pr=0.029)
= pressure correction
Fhc = 1/[1 + 2.75(q/q1 -1)0.2(BR/DR)0.75]
= Boiling range (BR) correction
and q1 = 1000 Btu/hrft2 for BR  56F
q1 =25000/(BR)0.8 Btu/hrft2 for BR > 56F
and DR = Tc - 1.12 Tsat
Bundle boiling heat transfer coefficient:
hb = Fhsc FhshFhbh1
where:
Fhb = 1 + 0.175[(1-Gb)/Gb]2/3 Limit: Fhb  2.5
= Convective enhancement from multiple tubes
Fhsh = Fqsh2/3
= Coefficient correction for sensible heat medium
Fhsc = (1 - Qsc/Qt)2/3
= Correction for subcooled feed

Boiling Side Temperature Difference

Average tube wall temperature:
Twall = Tsat + qdmax/hb
Nomenclature
A = Heat transfer area, ft2
BR = Boiling Range, F (dew point - bubble point)
Cp = Heat medium specific heat, Btu/lbF
Db = Bundle diameter, inches
DR = Tc - 1.12 Tsat, F
dto = Outside tube diameter, inches
Fhb = Factor for convective enhancement of boiling coefficient
Fhc = Composition correction to heat transfer coefficient
(boiling range correction)

D-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix D Shell Side Hydrocarbon Boiling Correlations

Nomenclature (Cont’d.)
Fhp = Pressure correction to heat transfer coefficient
Fhsc = Factor for heating subcooled feed to boiling temperature
Fhsh = Factor for sensible heat medium
Fof = Operating flexibility factor (usually 0.8)
Fqb = Bundle to single tube maximum heat flux ratio
Fqdo = Fraction of maximum bundle heat flux when dryout occurs
Fqsh = Ratio of maximum to average heat flux (for sensible heat medium)
Fu = Correlation uncertainty factor (0.7 for design)
Gb = Bundle geometry factor (ratio of peripheral area to heat transfer area)
h1 = Pool boiling heat transfer coefficient, Btu/hrft2 (outside of single
tube)
hb = Bundle heat transfer coefficient, Btu/hrft2 (shell side boiling in
bundles)
L = Tube length, inches
M = Mass flow rate of heat medium, lb/hr
N = Number of transfer units (based on heat medium)
nt = Number of tube sheet holes
P = Pressure, psia
Pc = Critical pressure, psia (mole fraction average for mixtures)
Pr = Reduced pressure = P/Pc
Qsc = Duty to heat subcooled feed, Btu/hr
Qt = Total duty, Btu/hr
q = Average bundle heat flux, Btu/hrft2 (heat flux is variable with
sensible heat media)
q1 = Reference heat flux in boiling range correction factor
q1max = Maximum heat flux for single tube, Btu/hrft2 (same as dryout heat
flux for single tube)
qbmax = Maximum heat flux for bundle, Btu/hrft2 (extensive dryout in center
of bundle)
qbdo = Incipient dryout heat flux for bundle, Btu/hrft2 (steam heated)
qdmax = Recommended maximum design heat flux, Btu/hrft2 
(average value with sensible heat medium)
Tc = Critical temperature, F (mole fraction average for mixtures)
Tsat = Saturation temperature, F (outlet saturation temperature for
mixtures)
Twall = Tube wall temperature, F (average for sensible heat medium)
Uc = Overall clean heat transfer coefficient, Btu/hrFft2

December 1989  1989 Chevron USA Inc. All rights reserved. D-3
Appendix D Shell Side Hydrocarbon Boiling Correlations Heat Exchanger and Cooling Tower Manual

D-4  1989 Chevron USA Inc. All rights reserved. December 1989
Appendix E Analysis of Horizontal Shell Side Steam Generators
With Very High Temperature Process Gas

This appendix supplements Section 350. Required heat transfer surface, steam
generation profiles, and dryout heat flux are covered.

Contents Page
E1.0 Heat Exchanger Surface Area E-2
E2.0 Axial Steam Generation Profile E-3
E3.0 Actual Peak Heat Flux E-3
E4.0 Dryout Heat Flux E-4
E5.0 Recommended Maximum Operating Heat Flux E-4
E6.0 Sample Heat Flux Calculations E-4
E7.0 Nomenclature E-6

December 1989  1989 Chevron USA Inc. All rights reserved. E-1
Appendix E Analysis of Horizontal Shell Side Steam Generators With Very High Temperature Process Gas Heat

E1.0 Heat Exchanger Surface Area

Heat transfer surface area can be calculated given the following process require-
ments: inlet gas temperature (Tgi), outlet gas temperature (Tgo), average gas
temperature (Tga), gas mass flow rate (Mg), and steam temperature (Ts). The
average gas temperature (Tga) in the tubes is lower than the numerical average of
Tgi and Tgo because of the shape of the temperature profile (the gas temperature
asymptotically approaches the water temperature). Tga can be approximated as the
gas outlet temperature plus one third of the total gas temperature drop.
Required clean surface area (Arc):
Q
Arc = -------------------------------
U  Tga – Ts 
(Eq. E-1)
Recommended heat exchanger surface area (A):
A = 1.25 Arc
(Eq. E-2)
where:
Q = Mg Cp (Tgi - Tgo)
(Eq. E-3)

1
U = ----------------------------------------------------------------------
Dto Dto  ln  Dto  Dti 
----------------- + ----------------------------------------------
Dti  hi 2 Kt
(Eq. E-4)
hi = NuK / Dti
(Eq. E-5)

Nu = 0.025 Re0.79 Pr0.42

(Eq. E-6)

m
Re = ----------------------------
2827Dti  
(Eq. E-7)

  C p 3600
Pr = --------------------------
-
KG
(Eq. E-8)
Kt = Thermal conductivity of tube
KG = Thermal conductivity of the gas
Gas properties are calculated at Tga.
Other variables are defined in the nomenclature section at the end of this appendix.

E-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix E Analysis of Horizontal Shell Side Steam Generators With

E2.0 Axial Steam Generation Profile

The steam generation profile is similar to the process gas temperature profile and
can be approximated by the following method. S(x) is steam production in pounds
per hour per linear foot of heat exchanger at x, where x is the distance from the hot
tube sheet.

qx  A
S(x) = --------------------
L  Hfg
(Eq. E-9)
where:

q(x) =U (T(x) - Ts)

(Eq. E-10)

U = U as calculated above

– U  Ax
T(x) = Ts + (Tgi  Ts) exp [ ---------------------------- ]
Mg  Cp  L
(Eq. E-11)

E3.0 Actual Peak Heat Flux

The actual peak heat flux, qpe, occurs just downstream of the ferrules where gas
from the ferrule impinges on the tube I.D. This impingement region is a few inches
long. The enhanced heat flux in this region is from two to three times the heat flux
with fully developed flow. The expression for the variable “hie,” below, represents
the enhancement of the internal heat transfer coefficient in this region. Inlet fluid
properties are used to calculate qpe.

qpe = Ue (Tgi - Ts)

(Eq. E-12)
where:

1
Ue = -------------------------------------------------------------------------
-
Dto Dto  ln  Dto  Dti 
-------------------- + ----------------------------------------------
Dti  hie 2 Kt
(Eq. E-13)

hie = hi 22.5 (Dti/Dfi)0.68 Re-0.21

(Eq. E-14)
hi, Nu, Re, and Pr are given by Equation E-5 through E-8.
Gas properties are defined at Tgi.

December 1989  1989 Chevron USA Inc. All rights reserved. E-3
Appendix E Analysis of Horizontal Shell Side Steam Generators With Very High Temperature Process Gas Heat

E4.0 Dryout Heat Flux

Dryout heat flux, qdo, is a fraction of maximum bundle heat flux, qmb. Maximum
bundle heat flux is calculated using correlations from Heat Transfer Research Inc.
(HTRI) experimental bundle data.

qdo = qmb  F2
(Eq. E-15)
where:
qmb = 664 Pc (P/Pc)0.28 (1-P/Pc)0.73 F1
(Eq. E-16)
F1 = exp(-0.22 [(1-Gb)/Gb]0.667)
(Eq. E-17)
F2 = 33.3 Gb - 0.632
(Eq. E-18)
With limits: 0.5  F2  0.7
Gb = Pi  Db  L / A
(Eq. E-19)
The total heat transfer area, A, should include the area of an internal bypass if it is
present and uninsulated.
Accuracy of these correlations is about 30%.

E5.0 Recommended Maximum Operating Heat Flux

The recommended maximum operating heat flux, qro, is 70% of the dryout heat
flux, qdo.

qro = 0.7 qdo

(Eq. E-20)

E6.0 Sample Heat Flux Calculations

Find the recommended maximum operating heat flux and the actual peak heat flux
for the following horizontal steam generator.
Geometry:
Tube O.D. = 0.125 ft
Tube I.D. = 0.09167 ft
Ferrule I.D. = 0.0687 ft
Bundle O.D. = 4.624 ft

E-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix E Analysis of Horizontal Shell Side Steam Generators With

Tube length = 23.71 ft

Number of tubes in bundle = 499
Total heat transfer area = 4646 ft2
Thermal conductivity of tubes = 17.5 Btu/hrft°F
Tube layout = rectangular (90°)
Operating conditions:
Inlet gas temperature = 1560 °F
Gas mass flow rate per tube = 393 lbm/hr
Gas specific heat = 0.66797 Btu/lbm°F
Gas thermal conductivity = 0.11337 Btu/hrft°F
Gas viscosity = 2.6155E-5 lbm/ftsec
Water/steam pressure = 890 psia
Water/steam temperature = 536.3 °F
Recommended maximum operating heat flux:
Gb =  DbL / A = 3.14159 (4.624) 23.71 / 4646 = 0.07413
F2 = (33.3 Gb) - 0.632 with limits 0.5  F2  0.7
= (33.30.07415) - 0.632 = 1.837; therefore F2 = 0.7 (upper limit)
F1 = exp (-0.22 [(1-Gb)/Gb]0.667)
= exp(-0.22[(1-0.0742)/0.0742]0.667) = 0.3059
qmb = 664 Pc (P/Pc)0.28 [1-(P/Pc)]0.73 F1
= 664(3208)(890/3208)0.28 [1-(890/3208)]0.73 (0.3059)
= 359,000 Btu/hrft2
qdo = QmbF2 = 359,000 (0.7) = 251,300 Btu/hrft2
qro = 0.7 qdo =0.7(251,300) = 175,900 Btu/hrft2
Actual peak heat flux:
Pr = Cp 3600 / K = 2.6155E-5 (0.66797) 3600 / 0.11337 = 0.5548
Re = m / (2827 DtiV) = 393 /[ 2827 (0.09167) 2.6155E-5] = 57,981
Nu = 0.025 Re0.79 Pr0.42 = 0.025 (57981)0.79 (0.5548)0.42 = 113.1
hi = NuK/Dti = 113.1(0.11337)/0.09167 = 139.9 Btu/hrft2°F
hie = hi 22.5 (Dti/Dfi)0.68 Re-0.21
= 139.9 (22.5) (0.09167/0.0687)0.68 (57,981)-0.21
= 382.7 Btu/hrft2°F

December 1989  1989 Chevron USA Inc. All rights reserved. E-5
Appendix E Analysis of Horizontal Shell Side Steam Generators With Very High Temperature Process Gas Heat

1
Ue = ---------------------------------------------------------------------------
Dto Dto  ln  Dto  Dti -
-------------------- + ----------------------------------------------
Dti  hie 2 Kt
(Eq. E-21)

1
= ------------------------------------------------------------------------------------------------------------
0.125 - + 0.125
------------------------------------- ln  0.125  0.09167 -
--------------------------------------------------------------
0.09167  382.7  2  17.5 
(From Eq. E-21)
= 214.1 Btu/hrft2°F
qpe = Ue (Tgi - Ts) = 214.1 (1560 - 536.3) = 219,200 Btu/hr
The actual peak heat flux is within the data scatter band of predicted dryout heat
flux and is much higher than the recommended maximum operating heat flux. This
helps explain tube failures in this unit between l979 and l982. In l983 the unit was
modified by plugging 36 tubes in the lower central region of the bundle. This
increases actual peak heat flux about 5% but increases dryout heat flux much more.
Calculation of dryout heat flux in bundles with selected tube plugging or removal is
beyond the scope of this appendix.

E7.0 Nomenclature

A = Total heat transfer area of steam generator, ft2

Arc = Required clean heat transfer area of steam generator, ft2
Cp = Specific heat of tube side fluid, Btu/lbm°F
Db = Diameter of tube bundle, ft
Dfi = Ferrule I.D., ft
Dti = Tube I.D., ft
Dto = Tube O.D., ft
exp = exponential function, e to the power that follows in parentheses
F1 = Multiplier for maximum bundle heat flux
F2 = Multiplier for dryout heat flux
Gb = Bundle geometry factor
gc = 32.17 lbm  ft/lbfsecsec
Hfg = Heat of vaporization for steam/water, Btu/lbm
hi = Heat transfer coefficient inside tube, Btu/hrft2°F
hie = Enhanced heat transfer coefficient inside tube, Btu/hrft2°F
K = Thermal conductivity of tube side fluid, Btu/hrft2°F
Kt = Thermal conductivity of tube material, Btu/hrft2°F
L = Length of pipe or tube, ft

E-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix E Analysis of Horizontal Shell Side Steam Generators With

ln = Log to the base e

log = Log to the base 10
m = Mass flow rate of tube side fluid per tube, lbm/hr
Mg = Total mass flow rate of tube side fluid, lbm/hr
Nu = Nusselt number
P = Steam/water pressure, psia
Pc = Steam/water critical pressure, 3208 psia
Pi = 3.14159
Pr = Prandtl number
Q = Heat duty, Btu/hr
q(x) = Heat flux at distance x along the tube, Btu/hrft2
qdo = Dryout heat flux, Btu/hrft2
qmb = Maximum bundle heat flux, Btu/hrft2
qpe = Actual maximum operating heat flux, Btu/hrft2
qro = Recommended maximum operating heat flux, Btu/hrft2
Re = Reynolds number
S = Steam production, lbm/sec
S(x) = Steam production at distance x along tube, lbm/hrft
Tga = Average tube side fluid temperature, °F
Tgi = Inlet tube side fluid temperature, °F
Tgo = Outlet tube side fluid temperature, °F
Ts = Shell side steam/water temperature, °F
T(x) = Tube side fluid temperature at distance x along tube, °F
U = Overall heat transfer coefficient, Btu/hrft2°F
Ue = Enhanced overall heat transfer coefficient, Btu/hrft2°F
 = Viscosity of tube side fluid, lbm/ftsec
x = Distance along tube measured from hot tube sheet, ft

December 1989  1989 Chevron USA Inc. All rights reserved. E-7
Appendix E Analysis of Horizontal Shell Side Steam Generators With Very High Temperature Process Gas Heat

E-8  1989 Chevron USA Inc. All rights reserved. December 1989
Appendix F Flow Rates and Shock Pressures Through Tube
Ruptures

Methods to calculate tube rupture flow rates and shock pressures are given here for
shell and tube exchangers containing gases, two-phase mixtures, and liquids.

Contents Page
F1.0 High-pressure Gases and Two-phase Mixtures F-2
F2.0 High-Pressure Liquids F-6
F3.0 Effective Bulk Modulus F-8
F4.0 Two-Phase Density F-9
F5.0 Applications F-9
F6.0 Nomenclature F-10

April 1998  1998 Chevron USA Inc. All rights reserved. F-1
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

F1.0 High-pressure Gases and Two-phase Mixtures

When the high-pressure fluid is part or all gas, the flow through the rupture may or
may not be choked (limited by the gas reaching sonic velocity). If it is choked, flow
through the rupture is independent of the low-pressure side. Otherwise, the response
of the low-pressure side affects the rupture flow. The rupture flow may be choked
on one side of the break and subsonic on the other side. Rupture flow from both
sides of a break is maximum for a break at one end.
Rupture flow rates have been calculated for the usual range of conditions in heat
exchangers and are given below in Equation F-1 and Figure F-1. For definitions, see
the nomenclature section at the end of this appendix.

k H + 1 0.5
 --------------
-
· = 2C  A   g  P  k  ---------------
M
2
-   k H – 1
R F T H H H  k + 1
H
(Eq. F-1)
Equation F-1, with CF = 1, is the equation for isentropic (frictionless) choke flow
from both sides of the break. CF is the ratio of actual flow with friction to the flow
rate for choked isentropic flow with a break at one end and is given in Figure F-1.
Two-phase flow through a rupture is similar to all-gas flow. Initially, the two-phase
rupture flow is mostly gas. The gas usually accelerates rapidly to near sonic speeds
by virtue of its internal energy, whereas liquid is dragged along by viscous shear and
accelerates much more slowly. The presence of liquid reduces the gas flow area and
adds to frictional resistance of the gas flow. Treating liquid-gas mixtures on the
high-pressure side as if they were all gas is conservative and is recommended.
The high-pressure fluid on the low-pressure side displaces and compresses the low-
pressure fluid. The pressure rise on the low-pressure side depends on the volume
flow rate into the low-pressure side, the compressibility of the low-pressure fluid
and the flow area of the connected piping. The fluid compressibility and pipe sizes
on the low-pressure side are usually the same for inlet and outlet lines for single
phase fluids but may be significantly different for liquid-gas mixtures.
Equation F-2 below is a volume balance: volume flow rate of high-pressure fluid
into the low-pressure side equals rate at which volume is made available by the
compression waves in the inlet and outlet piping.

· = 2A  G   g - 0.5
HFL   P LR – P L  
M -----------------------
L1  K L1
R L1 L

(Eq. F-2)

F-2  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

where:

 A L2   L1 K L1
0.5
1
GL = --
- 1+ ---------
-  --------
- ---------
- 
2  A L1   L2 K L2
(Eq. F-3)
For single phase service on the low-pressure side, fluid properties and pipe sizes for
inlet and outlet lines are usually equal. Then GL = 1 and Equation F-2 becomes:

· = 2A   g 0.5
HFL  P LR – P L    K
M R L
------------------
L L
(Eq. F-4)(From Equation F-3)
Solving Equations F-1 and F-2 simultaneously, with HFL = H (PLR/PH) for high-
pressure gas, results in:

2 0.5
P LR 1 PL   PL 
---------- = --- ------- +   ------- + 4 X
PH 2 P H   P H 
(Eq. F-5)
where:

0.5
 k--------------
H + 1
C F  A T    L1  K L1 2  k H – 1
-
X = -------  ----------  ---------  ---------- k H ----------------
G L  A L1   H   P H  kH + 1

(Eq. F-6)
Equation F-6 is plotted in Figure F-2.
Figure F-2 applies to any low-pressure fluid. Liquid cases are toward the right side
of the figure, gas cases toward the left, and two-phase cases in the middle.

April 1998  1998 Chevron USA Inc. All rights reserved. F-3
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

Fig. F-1 Factor for Friction and Flow Regime

F-4  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

Fig. F-2 Rupture Pressure on Low-Pressure Side Due to High-Pressure Gas

April 1998  1998 Chevron USA Inc. All rights reserved. F-5
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

F2.0 High-Pressure Liquids

Tube rupture with high-pressure liquid involves decompression waves in the high-
pressure system similar to the compression waves in the low-pressure system.
Choked liquid flow does not occur at conditions encountered in heat exchangers.
The relationship between rupture flow and decompression is given by:

g   H
· = 2A  P – P   -------------
0.5

HR  K 
M R H H
-
 H 
(Eq. F-7)
Equal pipe sizes and fluid properties are assumed for the high-pressure inlet and
outlet lines.
The rupture flow rate and pressure drop across the break for a tube break at one end
(worst case) are related as follows:

· = 2C  A   4  3 g    P – P   0.5
M R FL T H HR LR
(Eq. F-8)
where:

32 0.5
C FL = --- 1 +  ------------------------------------
1
2  3  2 + F  L  D
(Eq. F-9)
Solving Equations F-7, F-8, and F-3 simultaneously results in:

P LR – P L Z  1 + SH   4  SH 
2 0.5

---------------------- = --- ---------------------  1 + ---  ---------------- – 1
PH – PL 2 SH  2 Z  1 + S H 
(Eq. F-10)

P H – P HR  P LR – P L
----------------------- = S H  ----------------------
PH – PL  PH – PL 
(Eq. F-11)

· 0.5
M R  P LR – P L
- = 1 –  1 + S H   ----------------------
-------------------------------------------------------------------------------------------
2C FL  A T   4  3 g   H  P H – P L   0.5  PH – PL 
(Eq. F-12)

F-6  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

where:

 A L1  H  K H 
0.5
SH = GL  ---------
-  --------
-  ---------
- 
 A H   L1  K L1
(Eq. F-13)
and

2
 AT  KH 
Z = 4  3 C FL2  --------  -------------------
 A H  P H – P L
(Eq. F-14)
Equations F-10 and F-12 are plotted in Figures F-3 and F-4 respectively.

Fig. F-3 Rupture Pressure on Low-Pressure Side Due to High-Pressure Liquid

April 1998  1998 Chevron USA Inc. All rights reserved. F-7
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

Fig. F-4 Liquid Rupture Flow Rate

F3.0 Effective Bulk Modulus

Effective bulk modulus is the pressure change per unit of volume change.

P
K = ----------------
V  V
(Eq. F-15)
Expansion or contraction of the pipe, as well as compression or decompression of
fluid, contribute to the effective bulk modulus.
The effective bulk modulus, including pipe flexibility, is:

KF
K = ---------------------------------
-
KF  D  C
1 + ------------------------
E  tp
(Eq. F-16)
where C = 0.91 for restrained piping.
The effective bulk modulus is typically 10% less than the liquid bulk modulus.
Equation F-16 is appropriate for liquid systems. Bulk modulus data for liquids are
available in various reference books including the Fluid Flow Manual.

F-8  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

Bulk modulus for gases and two-phase flow may be calculated as follows:

R + 1-  P LR 
 R   P
------------ ---------- – 1
KL L 
------- = --------------------------------------------
PL –1  kL
 P LR
1 –  ----------
 PL 
(Eq. F-17)
where:

R =   1   g   -------------
Y
 1 – Y
(Eq. F-18)
Only gas phase compressibility is included in Equation F-17. Liquid phase
compressibility and pipe flexibility are insignificant in comparison to gas phase
compressibility.

F4.0 Two-Phase Density

Two-phase density of low-pressure fluid may be calculated as follows:
g
 = -------------------------------------------------
-
Y +  g  1   1 – Y 
(Eq. F-19)

F5.0 Applications
The use of the equations and graphs in this appendix is illustrated in the examples
following the nomenclature section below.
The units indicated in the nomenclature section are consistent with the equations.
The equations, however, are presented in terms of dimensionless parameters or
ratios where practical. Any consistent units may be used in those cases. Bulk
modulus, elastic modulus, and fluid pressure all have the same units. Where these
variables appear as ratios in the examples, psi units are used; otherwise psf units are
used.
Example 1 is for high-pressure gas and low-pressure liquid. Post-rupture pressure
on the low-pressure side is 83% of the initial pressure of the high-pressure fluid and
governs the design of the low-pressure side.
Example 2 is for high-pressure gas and low-pressure steam generation. Post-rupture
pressure on the low-pressure side is insignificant and does not affect design pres-
sure.
Example 3 is for high-pressure liquid and low-pressure liquid. Post-rupture pres-
sure on the low-pressure side is about 40% of the initial pressure of the high-pres-
sure liquid and governs the design of the low-pressure side.

April 1998  1998 Chevron USA Inc. All rights reserved. F-9
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

F6.0 Nomenclature
A = Flow area of tube or pipe, ft2
C = Pipe restraint factor
CF = Actual tube rupture flow/isentropic choke flow
CFL = Actual tube rupture flow/frictionless liquid flow
D = Inside diameter of pipe, ft
E = Modulus of elasticity of pipe, psf
F = Darcy-Weisbach friction factor
g = 32.17 lb/slug
GL = Factor for dissimilar inlet and outlet piping
k = Ratio of specific heats (Cp/Cv)
K = Effective bulk modulus of fluid and pipe, psf
KF = Bulk modulus of fluid, psf
L = Flow length through broken tube, ft
·
M R = Rupture flow rate, lb/sec
P = Pressure, psfa
P = Increment of pressure, psf
R is defined by Equation F-18.
SH is defined by Equation F-13.
tp = Pipe wall thickness, ft
V = Volume, ft3
V = Increment of volume, ft3
X is defined by Equation F-6.
Y = Weight fraction vapor
Z is defined by Equation F-14.
 = density, lb/ft3

Subscripts:
1 - Inlet piping, low pressure side (LPS)
2 - Outlet piping, LPS
g - Gas, LPS
H - High-pressure side, normal conditions
HR - High-pressure side, rupture conditions
HFL - High-pressure fluid on low pressure side
l - Liquid
L - Low-pressure side, normal conditions
LR - Low-pressure side, rupture conditions
R - Rupture conditions
T - Tube

F-10  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

Example 1

High-Pressure Hydrogen Gas and Low-pressure Cooling Water

Gas
PH = Pressure = 2000 psia = 288,000 psfa
H = Density = 0.459 lb/ft3
kH = Ratio of specific heats = 1.4

Cooling Water
PL = Pressure = 65 psia = 9360 psfa
L = Density = 62 lb/ft3
KF = Bulk modulus = 315,000 psi = 45,360,000 psf

Tube
Inside diameter = 0.482 in. = 0.0402 ft

Low-Pressure Piping
Inside Diameter = 4.026 in.
tp = Wall thickness = 0.237 in.
E = Elastic modulus = 29  106 psi = 4.2  109 psf
GL = 1 (similar inlet and outlet)

KF 315,000
K L = ---------------------------------
- = -----------------------------------------------------------------------
- = 269,700psi
KF  D  C  315,000   4.026
1 + ------------------------ 1 + --------------------------- ------------- 0.91
E  tp  29,000,000  0.237
(Eq. F-20)
Assume CF = 0.5

0.5
 k--------------
H + 1
C F  A T    L1  K L1 2  k H – 1
-
X = -------  ----------  ---------  ---------- k H ----------------
G L  A L1   H   P H  kH + 1

(Eq. F-21)

0.5
 2.4
-------
2  0.4
= -------  -------------  ------------
0.5 .482 62   269,700 2
- ------------------- 1.4 ------- = 0.662
1  4.026  0.459  2000  2.4
(Eq. F-22)

April 1998  1998 Chevron USA Inc. All rights reserved. F-11
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

2 0.5
P LR 1 PL   PL  65 2 0.5
= --- ------------ +   ------------ + 4  0.662 
1 65
- +   ------- + 4X
---------- = --- ------ = 0.83
PH 2 P H   P H  2 2000   2000 
(Eq. F-23)
PLR = 0.83 (2000) = 1660 psia
CF from Figure F-1 is 0.5; therefore assumed value above is okay.

k H + 1 0.5
 - 
 --------------
· = 2C  A    g P  k ---------------
2  k H – 1 
M T H
-
R F H H k +1 
 H 
(Eq. F-24)
0.5
 2.4
-------
  3.14   0.0402  2  2  0.4
= 2  0.5  -------------------------------------------- 0.459  32.17   288,000   1.4  -------
4 2.4

= 1.79lb/sec Total Flow from Rupture

Example 2

High-Pressure Hydrogen Gas and Low-Pressure Steam Generator

Gas
PH = Pressure = 2000 psia = 288,000 psfa
H = Density = 0.459 lb/ft3
kH = Ratio of specific heats = 1.4

BFW/Steam
PL = Pressure = 165 psia = 23,760 psfa
l = Liquid Density = 55 lb/ft3
g = Vapor density = 0.36 lb/ft3
kL = Ratio of vapor specific heats = 1.28
KF = Liquid bulk modulus = 183,000 psi = 26.4  107 psf

Tube
Inside diameter = 0.482 in. = 0.0402 ft

BFW Piping
Inside diameter = 2.067 in.
tp = Wall thickness = 0.154 in.
E = Elastic modulus = 29  106 psi = 4.2  109 psf
Y1 = Weight fraction vapor = 0 (R1 = 0)

F-12  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

KF 183,000
K L1 = ------------------------- = -----------------------------------------------------------------------
- = 169,000 psi
K F DC  183,000   2.067
1 + ---------------- 1 + --------------------------- ------------- 0.91
Et p  29,000,000  0.154
(Eq. F-25)

Steam Piping
Inside diameter = 7.981 in.
Y2 = Weight fraction vapor = 1 (1/R2 = 0)
Assume PLR/PL = 1.13

 1 + ----- 1-  ---------

P LR 
  - – 1
R 2  P L 
K L2  1   1.13 – 1 
---------- = ---------------------------------------------- = ----------------------------------------- = 1.43
PL – 1  k 1 –  1.13  – 1  1.28
 P LR L

1 –  ----------
 PL 
(Eq. F-26)
KL2 = 1.43 (165) = 236 psi

 A L2   L1 K L1
0.5
7.981 2 55 169,900 0.5
= --- 1 +  ------------- ----------  -------------------
1 1
G L = --- 1 +  ----------  --------- ---------- = 2472
2  A L1   L2 K L2 2  2.067 0.36  236 
(Eq. F-27)
Assume CF = 0.675

k H + 1 0.5
 --------------
C F  A T    L1  K L1 2  kH – 1
-
X = -------  ----------  ---------  ---------- k H ----------------
G L  A L1   H   P H  kH + 1
(Eq. F-28)

0.5
 2.4
-------
0.675  0.482 2
 ------------
55   169,900 2  0.4
= ------------- ------------- - ------------------- 1.4 ------- = 0.001025
2472  2.067  0.459  2000  2.4
(Eq. F-29)

P LR 1 PL PL 2 0.5
165- 2 0.5
+  ------ + 4X
1 165   -----------
--------- = --- ------ = --- -----------
-+ + 4  0.001025  = 0.0935
PH 2 PH PH 2 2000   2000
(Eq. F-30)
PLR = 0.0935 (2000) = 187 psia
PLR/PL = 187/165 = 1.13; therefore assumed value above is okay.

April 1998  1998 Chevron USA Inc. All rights reserved. F-13
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

CF from Figure F-1 is 0.675; therefore assumed value above is okay.

0.5
 k--------------
H + 1
-
2
· = 2C  A   g  P k ---------------  k H – 1
M R F T H H H k +1
-
H
(Eq. F-31)
0.5
 2.4
-------
  3.14   0.0402  2  2  0.4
= 2  0.675  -------------------------------------------- 0.459  32.17   288,000   1.4  -------
4 2.4
(Eq. F-32)

= 2.42 lb/sec Total Flow from Rupture

Example 3

High-Pressure Liquid and Low-Pressure Cooling Water

High-Pressure Liquid
PH = Pressure = 400 psia = 57,600 psfa
H = Density = 50 lb/ft3
KFH = Bulk modulus = 200,000 psi = 28,800,000 psf
Viscosity = 0.000672 lb/ftsec = 1 CP

Cooling Water
PL = Pressure = 65 psia = 9,360 psfa
L = Density = 62 lb/ft3
KFL = Bulk modulus = 315,000 psi = 45,360,000 psf

Tube
D = Inside Diameter = 0.56 in.
L = Length = 480 in. = 40 ft

High-Pressure Piping
Inside diameter = 4.026 in.
tp = Wall thickness = 0.237 in.
E = Elastic modulus = 29  106 psi = 4.2  109 psf

K FH 200,000
K H = ----------------------------- = ------------------------------------------------------------------------ = 180,700 psi
K FH DC
1 +  ---------------------------  ------------- 0.91
200,000 4.026
1 + --------------------   
Et p 29,000,000 0.237
(Eq. F-33)

F-14  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix F Flow Rates and Shock Pressures Through Tube Ruptures

Low-Pressure Piping
Inside Diameter = 4.026 in.
tp = Wall thickness = 0.237 in.
E = Elastic modulus = 29  106 psi = 4.2  109 psf
GL = 1 (Similar Inlet and Outlet)

K FL 315,000
K H = ----------------------------
- = ------------------------------------------------------------------------ = 269,700 psi
K FL DC
1 +  ---------------------------  ------------- 0.91
315,000 4.026
1 + -------------------  29,000,000  0.237
Et p
(Eq. F-34)

 A L1  H  K H 
0.5
4.026 2 50 180,700 0.5
SH = GL  ---------
-  --------
-  ---------
-  = 1  ------------- ------  ------------------- = 0.735
 A H   L1  K L1  4.026 62  269,700
(Eq. F-35)
Assume Friction Factor F = 0.0126

32 0.5 32 0.5

C FL = --- 1 +  -------------------------------- = --- 1 +  -------------------------------------------------------------
1 1
= 0.675
2  3  2 + FL  D 2  3  2 + 0.0126  480   0.56
(Eq. F-36)
2
 AT  KH  0.56 4 180,700
Z =  4  3 C FL2 --------  ------------------- =  4  3   0.675  2  -------------  --------------------- = 0.123
 A H  P H – P L  4.026  400 – 65
(Eq. F-37)

P LR – P L
---------------------- = 0.32 from Figure F-3 (0.318 from Equation F-10)
PH – PL
PLR = 65 + 0.32 (400 - 65) = 172 psia

M ·
R
--------------------------------------------------------------------------------------- = 0.67
2C FL A T   4  3 g   H  P H – P L   0.5

from Figure F-4 (0.669 from Equation F-12)

(Eq. F-38)

0.56 2
3.14  ----------
 12 
· = 0.67  2   0.675  -----------------------------
M R   4  3   32.17   50   400 – 65 144  0.5
4

(Eq. F-39)
= 15.7 lb/sec Total Flow from Rupture
Check assumed friction factor (above value is okay).

April 1998  1998 Chevron USA Inc. All rights reserved. F-15
Appendix F Flow Rates and Shock Pressures Through Tube Ruptures Heat Exchanger and Cooling Tower Manual

F-16  1998 Chevron USA Inc. All rights reserved. April 1998
Appendix G Flange Reference Information

April 1998  1998 Chevron USA Inc. All rights reserved. G-1
Appendix G Flange Reference Information Heat Exchanger and Cooling Tower Manual

Fig. G-1 Gasket Materials and Contact Facings (From ASME BPVC, Section VIII-I, Para. UCS-66. Table UA-49.1. Used
with permission.)

G-2  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix G Flange Reference Information

Fig. G-2 Effective Gasket Width (Courtesy of Lamons Gasket Co.)

April 1998  1998 Chevron USA Inc. All rights reserved. G-3
Appendix G Flange Reference Information Heat Exchanger and Cooling Tower Manual

Fig. G-3 Bolt Dimension Requirements (Courtesy of TEMA)

G-4  1998 Chevron USA Inc. All rights reserved. April 1998
Heat Exchanger and Cooling Tower Manual Appendix G Flange Reference Information

Fig. G-4 Code Stresses and Elastic Modulus

ASME SECTION VIII DIVISION 1 CODE STRESSES
COMPOSITION Cr-Mo C-Si C-0.5Mo 1.25Cr-0.5Mo-Si

FORM Bolts, 2.5" Max. Forging Plate Forging Plate Forging Plate

SPEC. NO. SA-193 SA-105 SA-516-70 SA-182-F1 SA-204 SA-182-F11 SA-387-11

SA-181-2 SA-515-70

B7 B7M B16 Gr. B Cl. 1 Cl. 2 Cl. 1

MIN. YIELD, 105 80 105 36 38 40 40 30 40 35 ELASTIC

KSI MODULUS

TEMP. LIMIT, F 1000 1000 1100 1000 1000 1000 1000 1200 1200 1200 C-0.5Mo
C-Si & 1.25 Cr

TEMP, F MAXIMUM ALLOWABLE DESIGN STRESS, KSI PSI x 10-6

-20 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 29.0 29.9

100 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 29.0 29.9

150 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 28.9 29.7
200 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 28.7 29.5

250 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 28.5 29.3

300 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 28.2 29.0

350 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 27.9 28.8
400 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 27.6 28.6

450 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 27.2 28.3

500 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 26.8 28.0
550 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 26.4 27.7

600 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 25.9 27.4

650 25.0 20.0 25.0 17.5 17.5 17.5 17.5 15.0 17.5 15.0 25.2 27.0
700 25.0 20.0 25.0 16.6 16.6 17.5 17.5 15.0 17.5 15.0 24.5 26.6

750 23.6 20.0 25.0 14.8 14.8 17.5 17.5 14.8 17.5 15.0 23.8 26.2

800 21.0 18.5 25.0 12.0 12.0 17.5 17.5 14.4 17.5 15.0 23.0 25.7

850 17.0 16.2 23.5 9.3 9.3 17.1 17.1 14.0 17.1 14.6 22.0 25.1

900 12.5 12.5 20.5 6.5 6.5 13.7 13.7 13.6 13.7 13.7 21.0 24.5

950 8.5 8.5 16.0 4.5 4.5 8.2 8.2 9.3 9.3 9.3 19.6 23.8

1000 4.5 4.5 11.0 2.5 2.5 4.8 4.8 6.3 6.3 6.3 18.1 23.0

1050 NA NA 6.3 NA NA NA NA 4.2 4.2 4.2 NA 21.7

1100 NA NA 2.8 NA NA NA NA 2.8 2.8 2.8 NA 20.4

1150 NA NA NA NA NA NA NA 1.9 1.9 1.9 NA 18.0

1200 NA NA NA NA NA NA NA 1.2 1.2 1.2 NA 15.6

April 1998  1998 Chevron USA Inc. All rights reserved. G-5
Appendix G Flange Reference Information Heat Exchanger and Cooling Tower Manual

Fig. G-5 Code Stresses and Relaxation Stresses

G-6  1998 Chevron USA Inc. All rights reserved. April 1998
Appendix H PCFLANGE Program User’s Guide

H1.0 Scope of the Program

The PCFLANGE program uses the Chevron method and the ASME Code method to
calculate stresses in integral-type heat exchanger body flanges. The user defines the
geometry, pressures, gasket factors, and Code allowable stresses. The program
compares calculated stresses to allowable stresses, enabling the user to trouble-shoot
existing flanges and size new flanges. The calculations, and the differences between
the Chevron and Code methods, are explained in Appendix G of the Heat
Exchanger and Cooling Tower Manual.

H2.0 Equipment Requirements

The program will run on IBM compatible PCs. A Macintosh version is planned, but
is not available as of this printing. If you are interested in the Macintosh version,
call CTN-242-3632 to check on its availability.

H3.0 Starting the Program

PCFLANGE is provided on a floppy at the back of the Heat Exchanger and Cooling
Tower Manual. The disk provided is write-protected, and, thus, will not allow data
or output files to be saved. Therefore, the program file (PCFLANGE.EXE) must be
copied to your hard disk or to another floppy before running.
To start the program type “PCFLANGE” at the DOS prompt.

H4.0 Running the Program

The program gives the user four options with the F1 through F4 function keys.
F1 starts the program’s data editor. This option allows data files to be retrieved,
edited, and saved. The first time the program is used there will not be any data files
saved. Pressing ENTER without entering a data file name will cause the sample data
to be displayed. This data can be saved, or edited and then saved, under any accept-
able DOS filename.
F2 starts the flange calculations. After F2 is pressed, the user is prompted for a data
file name. Pressing ENTER without entering a data file name will run the calcula-
tions using the sample data built into the program. The results of the calculations are
shown on the screen and saved in an output file.

March 1994  1994 Chevron USA Inc. All rights reserved. H-1
Appendix H PCFLANGE Program User’s Guide Heat Exchanger and Cooling Tower Manual

F3 starts bolt circle calculations using TEMA clearances. The user is prompted for a
required total bolt area. Then the number of bolts, bolt circle diameter, and other
useful dimensions are displayed in a table for all bolt diameters between one-half
and four inches.
F4 exits the program.

H5.0 Using the Data Editor

PCFLANGE has a data editor that can be accessed by pressing F1. However, once a
data file has been saved, any editor may be used to edit it. Users with DOS 5.0 or
higher may find the DOS editor useful. To use the DOS editor, type EDIT (file-
name) while in DOS. Any word processor may also be used to edit data files.
Remember to save the data file as an ASCI (text) file, and not in the word
processor’s format.
When F1 on the PCFLANGE main program screen is pressed the program enters
the data editor. In the data editor, the function keys are redefined as discussed below.
F1 reads in a data file. The user is prompted for a data file name. If ENTER is
pressed without a data file name being entered, the program’s sample data is
displayed.
F3 saves a data file. The user is prompted for a file name. If no data has been read in
before saving, the sample data is saved in the file.
F4 exits the data editor and returns back to the main program screen.
F5 and F6 are used to scroll the data up and down since the 28 lines of data will not
all fit on the screen at the same time.

H6.0 Understanding the Input Data

Most of the input values in the data file are self explanatory. The input values used
in the calculations are defined in Appendix G. A few of the data lines will be
discussed below.
Line 3 allows the user to enter a case title that will be displayed at the top of the
output screen and the output file.
Line 12 defines the nominal bolt diameter. The program supplies root diameter and
root area for standard bolts. If the user wants to specify non-standard bolts, zero
should be entered for bolt diameter in the data file. The program then prompts the
user for the bolt root diameter when F2 is pressed to start the calculations.
Line 26 is for actual bolt stress. This value should normally be zero. When the value
on line 26 is zero, the program calculates the recommended bolt stress for both the
Chevron and the ASME Code methods. Entering a non-zero bolt stress on line 26
allows the user to investigate the effect of various bolt stresses on flange stresses.
Line 27 is the number of gasket partitions (gasket segments that seal pass partitions
in the channel of multipass exchangers). The partitions are assumed to be one gasket

H-2  1994 Chevron USA Inc. All rights reserved. March 1994
Heat Exchanger and Cooling Tower Manual Appendix H PCFLANGE Program User’s Guide

diameter in length. The value on line 27 does not have to be an integer, therefore the
total length of short partition segments can be entered as an appropriate number of
gasket diameters.
If line 28 is non-zero, the results printed on the screen and in the output file will
contain several lines of intermediate Code parameters used in the flange calcula-
tions. These allow the user to analyze the details of the calculations or double-check
hand calculations.

H7.0 Understanding Calculation Results

The calculation output first displays the results of the Chevron method and then the
ASME Code method. These two methods are described in detail in Appendix G.
The two basic differences are that the Chevron method uses hydrotest pressure
instead of design pressure as the design basis and the Chevron method adds the
hydrostatic end load to the bolt load for gasket seating. This causes the Chevron
method to produce thicker flanges than the Code method.
The difference between the Chevron and Code calculation methods shows up in the
results in several ways. The top part of the output shows bolt loads for hydrotest,
operating, and bolt-up conditions. For the Code case, the hydrotest loads and bolt-up
end loads are zero since these values are not used in the calculations. The middle
section of the output shows a comparison between calculated flange stresses and
allowable stresses. An asterisk highlights stresses that are above allowable stresses.
The Code method shows zero hydrotest stresses since these values are not calcu-
lated. The Code operating stresses are usually lower than the Chevron method
stresses because the Code method gives lower recommended bolt loads for most
cases.
Line 16 of the output expresses flange rotation due to the maximum moment in both
angle of deflection (in radians) and in deflection of one flange at its OD (in frac-
tions of an inch).

H8.0 Technical Assistance

User suggestions, comments, and questions concerning the use of the PCFLANGE
program are welcomed. If you have comments or questions, or would like assis-
tance with the program please call CTN-242-3632.

March 1994  1994 Chevron USA Inc. All rights reserved. H-3
Appendix H PCFLANGE Program User’s Guide Heat Exchanger and Cooling Tower Manual

H-4  1994 Chevron USA Inc. All rights reserved. March 1994
Appendix I CTI Code Tower Standard Specifications, 
Nomenclature for Industrial Water-Cooling Towers

Courtesy of the Cooling Tower Institute

December 1989  1989 Chevron USA Inc. All rights reserved. I-1
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-3
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-5
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-7
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-9
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-11
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix I CTI Code Tower Standard Specifications, Nomenclature for

December 1989  1989 Chevron USA Inc. All rights reserved. I-13
Appendix I CTI Code Tower Standard Specifications, Nomenclature for Industrial Water-Cooling Towers Heat Ex-

I-14  1989 Chevron USA Inc. All rights reserved. December 1989
Appendix J Cooling-Water Treatment for Control of Scaling,
Fouling, Corrosion

Courtesy of “Power Magazine,” June 1984

December 1989  1989 Chevron USA Inc. All rights reserved. J-1
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-2  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-3
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-4  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-5
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-6  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-7
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-8  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-9
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-10  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-11
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-12  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-13
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-14  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-15
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-16  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-17
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-18  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-19
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-20  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-21
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-22  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-23
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-24  1989 Chevron USA Inc. All rights reserved. December 1989
Heat Exchanger and Cooling Tower Manual Appendix J Cooling-Water Treatment for Control of Scaling, Fouling,

December 1989  1989 Chevron USA Inc. All rights reserved. J-25
Appendix J Cooling-Water Treatment for Control of Scaling, Fouling, Corrosion Heat Exchanger and Cooling Tow-

J-26  1989 Chevron USA Inc. All rights reserved. December 1989