Document No: D/D&P/PRO/G/06 rev 1 w.e.f.

16/9/98

GUIDELINES FOR BASIC ENGINEERING

(Incl. Design Calculations)
Basic engineering includes the following activities:
a) Study the DOB
b) Prepare the scheme
c) Batch time cycle finalization / Material balance
d) Prepare the PFD
e) Prepare the Conceptual layout
f) Prepare the equipment specifications
g) Prepare the P&ID
h) Prepare the IPDS (Incl. ON-OFF valves), RD/SRV data sheet
i) Prepare the other documents ( Linelist, service list, Drive list, Effluent data, RM / Utility
calculation etc)
j) Prepare the Operation manual and Logic
k) Commissioning of the plant

1. STUDY THE DOB.

Study the DOB. Witness the experiments carried out in the Lab and/or and discuss
with R&D / Customer to confirm the understanding.

Check for any mismatch between the data in the entire DOB and get it corrected
by R&D.

Check for any missing information as required for engineering and get it from
R&D

While preparing appraisal, for costing of Plant / individual equipment / item, LAN
data is used.

2. PREPARE THE SCHEME

List out the key equipment. List out the major operations / steps and sub-steps.
Allocate these operations to the key equipment.

3. BATCH TIME CYCLE FINALIZATION / MATERIAL BALANCE

- Assume batch time cycle for each key equipment.

- Find batch size considering Plant capacity and Yield ( Overall and per
pass) and Batch time cycle.
- Batch size change from one key eqpt to next should be studied carefully.
(Check whether such # size change will be practical, e.g. slurry
intermediate splitting from 3 batches to 4 batches.) If necesssary correct /
increase the batch size accordingly.
- Calculate the equipment size for the batch size from the DOB data and
select appropriate size from the standard equipment sizes available (Refer
equipment specification section for standard sizes). Select the required
size in case standard is not available /suitable.
- Calculate the actual time required for Batch taking into account 10 %
margin. ( for human factor, batch to batch performance variation etc.)
 - Check whether the assumed and calculated time cycle is matching. If
not, repeat the above steps till these match.
- In case the equipment size is too high, more equipment in parallel can be
considered.
- These calculations should preferably be done in EXCEL to facilitate
check of any other alternative.

- CONTINUOUS PROCESS.
In case of a continuous plant a detailed flowsheet with stream nos added
is to be used and material balance (Component wise) is to be prepared.

4. PREPARE THE PFD

Prepare the PFD as per the guideline and following information:

After the key equipment and their sizes are decided, prepare the PFD. Show all
equipment around these key equipment taking into account the following:

- Raw material supply packaging

- Product physical specs
- Collection, storage , handling and recycle/ disposal of all streams ( Try to
use gravity than pumping/ pressure or vacuum transfer)
- Solid handling ( Keep slurry / solid handling to the minimum by re-
slurrying, dissolving in situ after isolation, extraction etc)
- Vent scrubbers, condensers, high level venting / dilution
- Required catch pots / seals to avoid mixing / contact with eqpt with
different MOC.
- Intermediate cuts
- Storages for countercurrent washes
- Traps for the effluent
- Weighing / accounting requirements
- Dump reactor / tanks

5. CONCEPTUAL LAYOUT
Prepare the conceptual layout as per the guideline.

6. PREPARE THE EQUIPMENT SPECIFICATIONS

Process Data Sheets of various equipment to be prepared taking following into
account :
- Flexibility as required for any probable change in the DOB
- Uniformity in various equipment e.g. Pump model, GCL std reactors
/tanks to facilitate common inventory for easy replacement at a later date.
- System balancing for maximum possible batch size in key equipment
- Corrosion data as available from the DOB and following literature :
Corrosion Resistance Tables, Third Edition, Part A and B, Philip A.
Schweitzer.
Corrosion guide by Erich Rabald
- Physical properties data as given in DOB and following literature :
a) Lange’s Handbook of Chemistry, Eleventh Edition, John A. Dean.
b) Perry’s Chemical Engineers’ Handbook, Perry & Green.
c) Vapor Liquid Equilibrium Data at Normal Pressures, First Edition,
E. Hala, I. Wichturle, J. Polak & T. Boublin.
 d) Azeotropic Databook , Lee Horsely.
e) The Properties of Gases & Liquids , Fourth Edition, Robert C. Reid,
John M. Prausnitz, Bruce E. Poling.
f) Computer Aided Databook of Vapor Liquid Equilibrium , Hirata Mitsuho,
Ohe Shuzo & Naga Hama.
In case data is not available then study/ use the experimental data from Lab.

Design of various equipment should be done as per the reference / guideline

mentioned below :

TYPE OF EQPT REFERENCE / GUIDELINE FORMAT No

Reactors Annexure I D/D&P/PRO/F/05

Columns Annexure II
Centrifuges Annexure III D/D&P/PRO/F/04
Filter Annexure IV
Vacuum pump / Ejector Annexure V
Exchangers and Process Heat transfer D/D&P/PRO/F/06
Other Heat transfer calculations by D Q KERN
Pumps Annexure VI D/D&P/PRO/F/07
Tanks Use 80-90% filling max D/D&P/PRO/F/08
Other eqpt As per literature data D/D&P/PRO/F/04

Refer explanatory notes for the format also.

7. P&ID PREPARATION
P&ID should be prepared as per the guideline and after taking into account the
following :

- Legends document should be prepared for each project showing the

symbols used in the P&IDs. A legend drawing prepared for BP0095 to be
generally followed.
- MOC of piping etc should be judiciously decided to keep the cost to the
minimum and taking into account the corrosion data as available from the
DOB and Corrosion Resistance Tables, Third Edition, Part A and B,
Philip A. Schweitzer
Corrosion guide by Erich Rabald
- One system should preferably be provided with single MOC, however
vapor / vent equalisation lines can have a different spec.
- Take into account effect of moisture ingress by any means on MOC
selected.
- Piping specs should be selected from the available std specs and used. In
case of non-availability of suitable spec, FLUID service list should be
issued to PIP for their recommendation on suitable spec to be used.
- Refer MSDS and properties data to decide the Conductivity strip ( for
Static electricity problem) and Flange guard ( For hazardous / hot fluids)
requirement.
- Provide various safety gadgets like Flame arrestor, RD, SRV, TRV, Gas
detectors etc (Smoke and heat detectors, Fire alarm system are to be
provided but not to be mentioned in the P&ID)
- Provide adequate piping connections to take care of any probable
abnormal operation and recycle / reprocessing of the off spec materials.
- Provide adequate local / panel instruments, sight glasses, and proper type
of manual valves at proper location so as to facilitate smooth operations
from field / panel.
- Local instruments should be provided to crosscheck the Panel instruments
and to control of operations from the field in exceptional cases.
- Pressure gauge to be provided on steam in jacket, pressure and return
temp of cooling/heating media to assess the performance of heating /
cooling operation
- Adequate isolation valves to be provided for each equipment or a group
of equipment to isolate them in case of emergency.
- Slurry piping should be considered critically for routing specified, valves
and instrument or any other items specified in the piping. Provisions for
cleaning should also be considered.
- Adequate drains and vents should be provided to facilitate proper
cleaning / draining of the system.
- Jacketting / tracing requirements should be carefully studied / provided.
Action plan in case of jacketting / tracing is not used ( by mistake) should
also be thought of and incorporated.
- U seals (Incl. inverted) should be specified wherever required to avoid
mix-ups / backflow.
- Slopes and non-hold up lines requirement should be critically studied and
specified.
- Check valves / catch pots to be provided to avoid reverse flow.
- Strainer requirement to be studied to avoid foreign material ingress from
MS eqpt to SS / MSGL assemblies OR to trap solids.
- Venting locations should be properly specified for safety.
- Blinds should be provided on various nozzles / valves to avoid any
leakage to atmosphere of Toxic material.
- Nitrogen blanketing / Vacuum connections to be provided as required.

- Line sizing in a batch plant is not so critical due to cushion available in

the time cycle from other activities. However, it should be done as per
ANNEXURE VIII and the comments given below :

All lines (except steam) should preferably be 25 NB size minimum.

Transfer lines to be sized as per the timing available in time cycle and 2-3
M/sec (liquid) velocity. Gravity transfer and slurry transfer to be at ~0.6
M/sec velocity. Transfer time should generally be kept below 1 hr.
Steam lines to be sized at 10-15 M/sec velocity.

Vent equalisation / vent lines can be of 25/ 40 NB size. Steam lines to

limpets for temp maintenance to be of 15 NB / 25 NB.

Sizing of all other vapor /vacuum lines should be recorded.

8 PREPARE IPDS / RD/SRV DATA SHEET
RD/SRV data sheet to be prepared as per the procedure given in. Annexure VII
IPDS to be filled in the prescribed formats as per the explanatory notes.

9 PREPARE THE OTHER DOCUMENTS ( Linelist, service list, Drive list, .

. Effluent data, Energy balance, RM / Utility calculation etc)
Prepare these documents in prescribed formats as applicable as per explanatory notes.
RM calculations should have reference to DOB data. (Preferably to be done in EXCEL to check .
. the variations in DOB easily) . Following norms to be used for solvent losses calculations :

TYPICALSOLVENT > Methanol Toluene /

Operation involved Xylene
General handling 1 1
Vac distillation 5 5
Atm distillation 5 3
Filteration 5 3
Extraction 2 2

10 PREPARE OPERATION MANUAL AND LOGIC : To be prepared as per the guideline.

Logic requirement should be in line with the Operation manual.

11 COMMISSIONING OF THE PLANT : To be done as per the guideline

12 LIST OF ADDITIONAL IMPORTANT REFERENCES

( Available in Library)

a) Applied Process Design for Chemical & Petrochemical Plants Vol. 1/2/3,
Second Edition, Ernest E. Ludwig.
b) Mass Transfer Operations, Second Edition, Robert Treybal.
c) Coulson & Richardson’s Chemical Engineering Vol. 6 ( Design ) ,
Second Edition, R. K. Sinnott.
d) Encyclopedia of Chemical Technology, Fourth Edition, Kirk Othmer.
e) Ullmann’s Encyclopedia of Industrial Chemistry, Fifth Edition.

****************
 ANNEXURE I
(Prepared by KSS)

REACTOR
MS & SS reactors
MS/GL reactors
MS/LB, MS/LB/TL, MS/FRVE/TL, MS/CTL reactors

MS & SS reactors

a) Capacity For calculation click Capacity chart

Based on the batch size & reaction mass volume the capacity of the reactor is decided. The R.M volume is
available from R & D in the DOB. The # size is decided based on the capacity of the plant.
For a given # size the volume of R.M, lit. = Max R.M volume (lit/km) x # size (km).
From this volume of the reaction mass the suitable reactor is selected from the GCL standard.

Standard reactor in GCL

4400 lit.
5750 lit.
9400 lit. Nominal capacity
15700 lit.
19000 lit.
25000 lit.

After selecting a particular reactor the % filling is checked.

Vol. R.M *100

% Filling = ---------------------
Nominal cap. of reactor

The % filling chosen is 80 % to start with but it can be as high as 90 % based on the GCL standard reactor.
For reaction involving gas sparging the % filling should be more 75-80 % for gas space. For washing /
extraction it can be as high as 95 % even in some cases.

b) MOC selection

MOC is selected based on the information given in the DOB. If the information is ambiguous the
actual corrosion study should be done by corrosion lab (Activities to be co-ordinated by R & D ) &
actual corrosion data to be furnished. The reaction mixture contains so many components in it
that the reference can not be taken from books also in most of the cases. So corrosion study is
must. If the process is to be fitted in the existing plant which is having some equipment then the
corrosion study should be done for the MOC of the said equipment to certify its suitability.
c) Agitator design

The design of the agitation system involves:

Selection of type of agitator

Deciding sweep
Deciding RPM
Calculation of fluid power
Calculation of motor HP / Motor selection
Sizing / Selection of gear box
Selection of pulleys & belt

Selection of type of agitator

The type of agitator is decided based on the type of application. Various types of agitator & their
combinations are used.

Paddle:

2 bladed paddles are used mainly for blending operations. The sweep is normally 60-80 %. Paddle
may be used with or without baffles. If baffles are provided it will give good intense agitation &
vortex formation will be minimized. Either 2 or 4 baffles are used. Paddles may be straight or
pitched (45° pitched). Pitched paddles are used for solid suspension.

Gate anchor:

It may be single or double gate anchor. When the height of the vessel is less single gate anchor is
used. If height is more staggered double gate anchor is used. Gate anchors are used mainly for
crystallization / Distillation operation. Sweep of these agitators is 80-90 %. These agitators
provide very good scraping action on the wall thereby preventing the deposition of solid on the
wall (Preventing caking). This results in good heat transfer which is the main criteria for
crystallization / Distillation. Baffles are not used because baffle will help in solid deposition on the
wall of the vessel. This type of agitator operate at low RPM of the order of 20-40 depending upon
the size of the vessel.

3 bladed curved paddle:

These are used in chemical reaction, distillation etc. This type of agitator is available on glasslined
vessels operating at very high speed of the order of 96 RPM. It provides very intense agitation.
The sweep is 55-60 %.

Turbine:

There are 2 types of turbine agitator: Open turbine and Disc turbine.

In open turbines 6 / 4 bladed the blades are connected to the hub directly. In disc turbine the blades
are connected to disc which in turn is connected to the hub. As per the mounting of the blades the
turbines can be flat blade or pitched blade. In pitched turbine the blades are at an angle of 45°. Due
to this angle the flow pattern becomes axial in pitched turbine whereas it is radial in flat blade
turbines. The sweep is 33-43 %. These operate at very high speed. Agitation is very vigorous.
Normally baffles are used with turbines ( 2 or 4 baffles). Pitched turbines are used for solid
suspension. Disc turbines are used for 2-phase reaction, gas-liquid reaction.
 Multiple agitator:

In this various kind of agitation is possible depending upon the method of installation. In the
reactors used in GCL following combination can be achieved by proper fixing of impellers.
 60 % & 70 % paddle, no baffle.
 60 % & 70 % paddle, 2 baffles or 4 baffles.
 80 % staggered double gate anchor, no baffles.
 40 % disc turbine, 4 baffles.
 70 % paddle + single gate anchor, 2 baffles.

Examples of Some of the Typical Agitators used are:

 6 bladed flat blade disc turbine is used in GCL for IP condensation. The DMA gas is sparged
in p-cumidine. 4 baffles are used.
 6 bladed flat blade disc turbine is used for Cypermethrin condensation, Temephos
condensation, Napropamide condensation reaction in GCL. These reactions are between
aqueous & organic phase with the help of phase transfer catalyst. Very intense agitation is
required which disc turbine is quite capable of providing.
 80 % staggered double gate anchor is used for 2 CB crystallization in CMAC process, IP
crystallization etc.
 Paddles are used for distillation operation.
 Combination of flat blade disc & pitched turbine is used in hydrogenation of vegetable oil.
This reaction is done using raney nickel catalyst. Hydrogen gas is sparged into the vegetable
oil. The flat blade disc turbine is used for effective dispersion of hydrogen in oil. The pitched
turbine is used for suspending the raney nickel catalyst solid. The bottom pitched turbine is
upward flow type & top one is of downward flow type. The reaction is 3 phase reaction
involving gas-liquid-solid.
 In some of the applications the particular use is playing significant role in the very selection
of the agitator. For example in 2 CB preparation in CMAC process previously multiple
agitator was sued where disc turbine was extended with blades to make paddle. These blades
were connected with verticals to make gate anchor. In the reaction some tarry material was
generated. Before starting next batch the vessel needs to be washed very thoroughly. The tar
used to stick on the disc turbine & washing became real problem. For this reason the idea of
multipurpose agitator was dropped for this particular application. Standard double gate anchor
was made & installed. So in this case the very process is playing very significant role in
governing the agitator type.

Terminology’s used in agitation

Sweep:
It is the ratio of outside diameter of impeller to the inside diameter of the vessel expressed as the
percentage. If the vessel dia is 1500 mm and impeller dia is 600 mm the sweep is 40 %.

RPM:
It is the revolution per minute the shaft makes.

Power per unit volume:

It is the power input to the liquid per unit volume of the liquid.
Power number:
It is a constant & characteristic of a given type of impeller. The power number chart is given in the
calculation program.

P X gc X 75
 = --------------------------------------, where
 X 1000 X (N/60)3 X D5
F = Power number
P = Input fluid power, HP
gc = Acceleration due to gravity, 9.8 m/sec²
 = Density of liquid, gm/cc
N = RPM
D = Sweep of the impeller, m

Having known the power number for particular agitator the power required can be calculated.

L/D ratio:
It is the ratio of the liquid height to the dia of the vessel.

Deciding sweep

The sweep is decided by the agitator type.

Deciding RPM

It is decided using various scale up scale down criteria, type of application.

Some of the scale up criteria are:

 Constancy of power per unit volume.

This is the criteria very useful & used in more than 90 % of the cases. If we know the type of
agitator in one vessel & its RPM the RPM of other geometrically similar vessel can be
determined.

P P
--- = ---
V 1 V 2

From power number formula

P X gc X 75
 = --------------------------------------
 X 1000 X (N/60)3 X D5
.

. . P  N3 D5

.
. . N13D15 N23 D25
---------= ----------
V1 V2
1/3
N13 D15 V1
 N2 = ---------------
V1D25
 Constancy of tip speed.
This is used mainly for reaction. To start with some value of tip speed is taken such as 800-
900 ft/min. for reaction involving gas sparging in liquid. Based on this the RPM is calculated.
From this RPM scale down is done to lab scale. Experiments are conducted in lab varying
some parameters. The best results obtained in lab are again scaled up to plant scale.

V = rwhere

V = Tip speed in ft/min.

r = Sweep/2
 = Angular speed
2
 = ---- = 2
T

.
. . V = 2rN = DN, D = Sweep
N = RPM

V1 = V2
.
. . D1N1 = D2N2
D1
 N2 = ---- X N1
D2

 Constancy of Reynolds number

This criteria is used mainly for heat transfer purpose such as in applications like
crystallization, distillation etc. The overall heat transfer coefficient consists of inside &
outside h.t.c. The outside h.t.c is constant depending on the flow rate of utility fluid in the
jacket / limpet of the vessel. The inside h.t.c is directly proportional to some power of
Reynolds number.

D2 N 
Re = --------- , where

Re = Reynolds number
D = Sweep, cm
N = RPS ( revolutions per second )
 = Density, gm/cc
 = Viscosity, poise.

Re1 = Re2
.
. . D12 N1  D22 N2 
------------ = ------------


. D12 N11 2
. . N2 = -------------- RPM
D22 2 
Calculation of Fluid Power For calculation click---- Agitator calculation

Having calculated the required RPM based on the appropriate scale up criteria power can be
calculated.
Pgc X 75
 = -----------------------------
 X 1000 X ( N/60)3 D5

. f X r X 1000 X (N/160)3 X D5
. . P = --------------------------------------- HP , where
75 X 9.81

P = Fluid power, HP
= Power number
= Density, gm/cc
N = RPM
D = Sweep of impeller, m

Calculation of motor HP / motor selection

To the fluid power calculated various losses added such as frictional loss in bottom guide, shaft
transmission losses, gear box loss, belt friction loss etc. The calculation is shown in the calculation
program.
After calculating the motor HP suitable motor is selected for the required duty. The HPs of the
standard motor’s are:
0.5, 1.0, 1.5, 2, 3, 5, 7.5, 10, 12.5, 15, 20, 25, 30, 40,50, 60 etc.
The nearest higher size motor should be selected. For example if the motor HP comes to 13.5 by
calculation select 15 HP motor. The RPM of the motors used for agitators is normally 1440. If
2900 RPM motor is used lot of reduction will be required to attain the required RPM. Some times
where there is a chance of agitator jamming deliberately higher size motor is selected than
calculated to overcome the initial torque.

Sizing / selection of gear box

Having selected the motor, gear box is selected by referring to the gear catalogue. Two criteria
should be satisfied by the gear selected.

1. The actual output torque is lesser than the allowable input torque to gear.
2. The actual input HP is lesser than the allowable input HP to gear.

HP X 63000 KW X 9550
Torque = ---------------- or Torque = ---------------
(lb-inch) RPM (Nm) RPM

For turbine agitators power is more but RPM is also more so the output torque may not be that
high. For anchor agitators where RPM is very low, the output torque becomes very high. The gear
in such cases may be suitable from HP point of view but may not be o.k from torque carrying
capacity point of view. So such gears should not be selected.
The reduction ratio of gear should be selected such that the ratio of pulleys sizes required is <=
1.7. The standard reduction ratios of gears are:
5:1, 10:1, 15:1, 20:1, 25:1, 30:1, 40:1, 50:1 etc.
 Output RPM of gear
Reduction ratio of gear = ---------------------------
Input RPM to gear

In GCL most of the gears used are of Radicon make. Planetary gears have also been used of
Breveni make. The are high speed gears which give very high efficiency compared to worm gear.

Selection of pulleys & belt

From pulleys chart the power carrying capacity of pulleys is given. Number of grooves is decided
based on the power carrying capacity per belt after applying various correction factors such as
slippage etc. Based on power carrying capacity various sections of pulleys come.

Upto 10 HP : ‘A’ section pulley

10-20 HP : ‘B’ section pulley
20 HP & above : ‘C’ section pulley

Fenner make pulleys are very common & have got lot of flexibility. It has got two parts pulleys &
bush. By changing the bushes same pulley can be used for various applications. In ‘C’ section
pulley minimum 4 grooves come.

Pulleys are normally designated as: 200B3, 236C4, 280C4 etc.

Here 200, 236, 280 etc----- PCD of pulleys
B, C------ Sections
3, 4 ------ Number of grooves.

If the ratio of pulley on gear & motor is > 1.7 lot of slippage will occur.
Depending upon the HP of motor, its shaft dia is fixed & for that shaft dia some minimum size of
pulley will fix and not less than that. So this should be borne in mind while selecting the pulleys.
The calculation program shows power calculation, motor, gear, pulleys design / selection.

MS/GL reactors For calculation & chart click---- Glass lined vessels

For corrosive applications glass lined reactors are used. The design is totally by supplier. For various
capacities the details of GMM glass lined reactors are shown in the chart. The reaction mass volume for the
given # size is calculated as per the procedure mentioned for MS-SS reactors.

In conventional GMM glass lined vessels three bladed curved paddle is typical impeller used but now
various other types of agitators are also available such as turbine, anchor etc. Some of the typical uses of
MS/GL reactors in GCL are: Chlorination, Bromination, Acidification to low pH of the order of <=2 etc.

MS/LB, MS/LB/TL, MS/FRVE/TL, MS/CTL reactors

For corrosive applications these MOC reactors are also used. On the parent material which in MS bonding /
lining is done of lead, FRVE etc. Tiles are acid resistant, carbon etc. Care should be taken while deciding
the capacity of these lined vessels. Because of lining the actual ID becomes lower than the parent material
which is MS & hence there is reduction in the volume.

The lined vessels should not be used for some materials where contamination of foreign particles is not
acceptable at all. For example in polymer no impurity is allowed so tile lined vessels are out of question for
final polymer at least. In tile lined vessels teflon lined agitators are also possible to install. The biggest
disadvantage of these tile lined vessels is the heat transfer. In MS/LB the heat transfer is possible lead being
good conductor of heat. In AR tile lined vessels it is not possible. In carbon tile lined vessels it is possible
carbon being good conductor of heat. In some of the tile lined vessels where heat transfer is real problem
heat is removed by reflux condenser.
-----------------------------------------------------
ANNEXURE II
(Prepared by VR and PGB)

Design of Fractionation Systems

Fractionation systems represent a major component of chemical process plants .Since large volumes of
published data are available design of distillation systems seldom requires pilot plant studies.

Design of fractionation columns involves the following :

 Calculation of number of stages

 Calculation of height of column
 Hydrodynamics

I) Calculation of number of stages:

Determination of number of stages required to perform the required separations is first step in design
.Accuracy of design to a large extent is dependent on vapour-liquid equilibrium data on which calculations
are based.

A) Sources of vapour-liquid equilibrium (VLE) data:

1) Reported VLE data are available in books & Journals . A few selected referances are given
below :

a) Vapour Liquid equilibrium data by Hala

b) DECHEMA chemistry data series
c) Journal of chemical & Engineering data
d) Fluid Phase Equilibria

2) If reported VLE data is not available ,either determine VLE by experimental methods or
design based on relative volatility () calculated from vapour pressure data. vapour pressure data can be
obtained from books/literature or can be determined experimentally .

The following points are to be kept in mind while designing columns based on  from vapour
pressure data .
 Confirm that azeotrope is not formed. Aspen ; Azeotropic data book by Lee Horsley Perry,
Lange gives extensive data on azeotropes.
 For highly polar compounds calculation of  from vapour pressure data gives large errors &
is not to be used.

Vapour pressure in absence of reported data , can be be experimentally determined . Vapour

pressure apparatus is available in pilot plant which is suitable for measurement of vapour pressure < 760
mm Hg .For measurement of vapour pressure first determine the range in which vapour pressure
-temperature data are to be made .

Ensure that,
i) Data is be collected at several points (minimum five points) .
ii) For measurement of vacuum ,correct for barometric pressure.
iii) If we are measuring vapour pressure of two compounds with  <1.5 , ensure that
temperature & pressure measuring devices have accuracy of < 1 % of full scale .
 Standard mercury manometers do not give required accuracy .The following devices
improve the accuracy of pressure measuring devices
 Oil filled absolute pressure manometer for measuring pressure below 150 mm HgA.
 Kathetometer which magnifies the scale reading & gives level up to  0.01 mm .

3) Prediction of VLE data from thermodynamic phase equilibrium models using Aspen etc.

4) VLE data from experiment methods : The method followed in GCL till date is distillation
method in which small quantity of distillate is collected .This system gives large errors due to
condensation of vapours on cold walls of flask. Apparatus called ‘Jones still’ has been fabricated
& is available in pilot plant . In this system the returning vapours are bubbled through the liquid to
ensure adequate mixing & thus eliminate errors due to condensation of vapours on cold walls of
flask .

II) Calculation of height of column :

A)First step is Calculation of Number of Theoretical Stages ( N ) & Reflux Ratio ( R R)
For batch fractionations use programme -Batch
For continous fractionations use programme -Continous
These programme is based on fixed value of  .

Batch fractionations :

i) For fitting in existing column ‘N’ is fixed & R R is to be adjusted accordingly. The programme
has three alternatives :
Alternative 1 - for main cut : Input desired composition of distillate .Programme calculates
distillate & bottoms quantity ;final bottoms composition .
Alternative 2 - for inter cut : Input desired composition of bottoms .Programme calculates
distillate & bottoms quantity ; average composition of distillate .
Alternative 3 - : Input desired quantity of distillate .Programme calculates distillate & bottoms
composition ;quantity of bottoms .

ii ) For new column both N & R R is varied to obtain desired results.

Continous fractionation : Input data - Feed flow rate ,composition,condition ;  ; R R ;

desired distillate & bottom composition

Guidelines :
a)Take  at end conditions of the cut.This gives conservative design.
b) Check minimum R.R & minimum number. of plates .
Operating R.R = 1.25 to 1.50 times minimum .
N =1.5 to 2 times Nmin.

B) Column Height =N * HETP :

HETP is a function of packing size & type ,physical properties(surface tension & viscosity ), distribution
& capacity.For calculation of HETP of IMTP packing use Norton Formula (see appendix 1 ).
Norton has given formula for system base HETP which assumes uniform distribution of vapour & liquid in
the column.This concept is useful because it isolates the system’s effect on HETP away from distribution
conditions.The formula is valid for for systems which are non aqueous , non ionising and have  < 3

C)GCL has standardized height of column section for distillation = 4270 mm . Even if calculated height
is less ,take column of 4270 mm height. The packed height is 3900 mm and ‘N’ is to be taken accordingly.
Note : 1)M/S Kevin Ind. the manufacturer of IMTP has recommended to dry pack IMTP.
2) Thickness of IMTP is 0.5 mm & hence corrosion should be carefully checked .MOC of packing
should be superior to MOC of column thus providing galvanic protection .
III) Hydrodynamics :

A) Packing used are IMTP (available in SS316,SS316 L) and Ceramic intalox saddles
Ceramic preferentially wets aq. systems while metal preferentialy wets
organic systems.
IMTP is prefered except in cases where ceramic shows better performance.For e.g
aqueous NH3/DMA fractionation.
Prefered size of IMTP is 15 mm .As per Norton IMTP is effective at minimum wetting rate of 0.5
M3/M²/hr. For other packings minimum wetting rate is 1.25 M3/M²/hr .
B) For Columns with IMTP packing ,Column Diameter is calculated as per procedure & capacity/pressure
drop correlations given in Norton catalouge. (See appendix 2 )

Columns are to be designed for capacity factor of 60 to 70 % at top.

For columns with other packing ,column diameter is calculated as per generalized correlation of Leva &
Eckert given in a)Treybal page 160 (in FPS units )
b) Perry (VI th edition ) page no 18-22 fig 18-38. (in SI units )
Column are to be designed for capacity factor of  60 % at top.

Pressure drop (P) permitted is 10 % of flow the minimum head (h) required are as under : P across
column =geometric mean of P at top & P at bottom .
P across column ranges from 0.3 to 1.0 inch water/ft of packing .Flooding can occur when P exceeds
1.3 inch water/ft of packing . For fractionation below 15 mm Hg ,pressure drop across column is critical &
is calculated at several points over the height of column.

C) Design of column internals :

1)Liquid distributors Liquid distributor is the most important internal of packed column. Liquid
distributors are either gravity type or pressure type. Pressure type distributors consist of perforated pipes
and are used in absorbers where head available is much higher . Liquid distributors should be positioned 4
to 6 inch above the packed bed .

A) Gravity type liquid distributors:

Three types available are Orifice ; Weir flow and V notch Weir. Orifice type distributors are less than
sensitive to variation of head across the the distributor caused by improper levelling as explained by the
relation between flow & head as under :
Orifice : flow  (h)^ 0.5
Weir : flow  (h)^ 1.5
V notch : flow  (h)^ 2.5

If there is 10 mm variation of liquid head & maximum variation

Orifice distributor : h = 47.6 mm
Weir distributor : h = 152.3 mm
V notch distributor : h = 257.3 mm:
Liquid distributors in use in GCL are of orifice type .These comprise of flat plate having risers for vapour
flow & orifices for liquid flow.

a)Location of orifices ; See appendix 3.

b) Diameter of orifice : Diameter of orifice is calculated using equation

W = K*A**(2*g*h)^0. where W = flow rate in gms/sec

 = density of liquid in gm/cm3
h = height of liquid above orifice cms
g = 980 cm/sec²
 K = Orifice coeff. =0.7 for punched orifice

The diameter of orifice is selected such that height of liq. above orifice (h) is
i)@ TR is does not exceed 120 mm (see note)
ii)@ Operating reflux is not below 15 mm
iii) We take minimum diameter of orifice= 5 mm . Orifices smaller than 5 mm may result
in clogging .(However orifice of diameter < 5 mm are specified by vendors of internals
such M/S Kevin Enterprises. )

The actual liquid head (H) will be increased due to pressure drop(P ) across the distributor & the
aeration factor .

H= (h+P ) (density of liq.)

----------- * -------------------
aeration factor (density of liq-density of gas )

The increase due to above  20 mm of water .

Note : The height of gas riser is usually 150 mm . Hence liquid head at TR should not
exceed 120 mm .

c)Location/Diameter of gas risers :Locate of gas risers such that

i) gas riser area is distributed equally between inside and outside of 50 % of column
area .( riser area inside is  40-60 % of column cross-sectional area.)
ii)Area of gas risers is minimum 15 % of column cross-sectional area .

2) Packing support plate :

In packing support plate gas distribution is more important than liquid distribution .
Separate passages are provided for gas & liquid by locating gas inlets to packed beds above those points
from where liquid flows from bed.Open area for gas flow is equal or exceeds 100 % of column cross-
sectional area .Gas injection type packing support plates supplied by M/S Kevin have maximum thickness
of 3 mm (due to limitation of punching machine ) . Thickness of packing support plates fabricated as per
GCL design is 10 mm .

For fabricated Packing support plates :

a) Gas riser area is distributed equally between inside and outside of 50 % of column area .Gas riser area
inside the 50 % of column cross-sectional area should be 45-55 % of column area .This is the most
important criteria in design of packing support plates .Number of gas risers ,diameter ,pcd of gas risers are
varied to arraive at optimum design .

b) cross-sectional area of gas riser is minimum 25 % of column area. Gas risers are either circular pipes
or rectangular channels with cap at the top. The cap is supported on the riser by 5 mm stripes such that
flow area betwen riser & cap is equal to column cross-sectional area.

c)Liquid orifice location is not critical .Diameter of liquid orifice is selected such that liquid head above
orifice is minimum. Since IMTP 15 is widely used ,inorder that packings should not pass through orifice
select 10 mm diameter .

3) Redistributors :
Redistributors comprise of separate packing support plate & liquid distributors . The packing support plate
design is similar to the packing support plate at the bottom of column.
In the liquid redistributor , the orifice location ,number & diameter is same as in liquid. distributor at top.
The gas risers number & diameter is same as in liquid distributor at top except that gas flow is from sides
similar to gas risers on packing support plate.
 Appendix 1

Calculation of system base HETP :

System base HETP = A(20/)^0.16 *(1.78)^ for   0.4 cp

= B(20/)^0.19 *(/0.2)^0.21 for  > 0.4 cp

Where  is surface tension of liquid in dynes/cm .For  >27 use  =27

 = Liquid viscosity in CP
Values of ‘A’ and ‘B’ depends on size of packing and are as under
Values of A’ and ‘B’ are in mm .

Packing size A B
#15 272 296
#25 351 383
#40 412 452

Norton’s standard designs use operating HETP = 13 % above system base HETP

Rule of thumbs :
For ceramic intalox saddles HETP =24*Diameter of packing
 Appendix 2

Calculation of capacity & pressure drop in column .

Column C 3401
Function
Cut Toluene Toluene
Packing IMTP 15 IMTP 15
Vapour rate: V Kg/hr 1400.000 750.000
Reflux ratio: R 0.500 0.500
Liquid rate L= (R/R+1)*V Kg/hr 466.667 250.000
M3/M2/HR 5.401 2.893
Liquid density: l Kg/M3 870.000 870.000
Molecular weight: M 92.000 92.000
Surface tension S Dynes/cm 21.00 21

Pressure: P MM Hg 760.000 200.000

Temperature:T Deg C 50.000 50.000
Vapour density: g= (M*P*273)/ Kg/M3 3.471 0.914
(760*(t+273)*22.4)
Liquid. viscosity:  Cp 0.450 0.450
Column Diameter :D Inch 14.000 14.000
Area: A M2 0.099 0.099
Superficial velocity: U M/Sec 1.128 2.296
Capacity factor Cs M/Sec 0.071 0.074
Parameter X 0.021 0.011
Kinematic viscosity:  =/l CS 0.517 0.517
Capacity Co: From graph given in Noton M/Sec 0.115 0.120
catalouge
Efficient. capacity M/Sec 0.105 0.110
Csc=Co*(S/20)^0.16*(/0.2)^-0.11
Capacity Rating: (Cs*100)/Csc % 67.875 67.829
Ordinate value (from graph ) Y 2.620 2.849

Pressure drop:from graph given in Norton MM H2O/M 55.000 55.000

catalouge
Inch H2O/ft 0.660 0.660
MM Hg/ft 1.233 1.233
 APPENDIX 3

LIQUID DISTRIBUTORS:

The design is based on article by F MOORE & F RUKOVENA presented in 36 th Canadian conferance of
chemical engineers .

This article describes design of liquid & gas distributors based on concept of distribution quality `DQ'. DQ is a
measure of uniformity of liquid flow at the top of the bed. Each distribution point is representated by a point
circle, whose centre is located where liquid strikes the top of bed. Sum of area of point circles is equal to Cross
Sectional area of column.

DQ = 0.4 (100-A) + 0.6B - [0.33 * (C-7.5)]

where A = Cross sectional tower area not covered by point circles in %.

B = Least point circle area in 1/12th of tower area x 100

tower area / 12

OR

tower area / 12 x 100

most point circle area in 1/12th tower area

whichever is less

C = area of overlap of point circles x 100

tower area

An optimum distributor evaluated by this method will have all tower area on top covered by point circles. It is
geometric fact this is not possible & there will be overlap between point circles and/or point circle overlap
outside the tower area. It is also geometric fact, with this method of evaluation, tower area not covered by point
circle (A) is equal to point circle area which overlaps tower wall plus overlapping area of adjacent point
circles.

Both ‘A’ and ‘C’ values are a measure of uniformity of liquid distributed over the cross sectional tower area at
the top of bed .Factor `B' is a measure of deviation of liquid flow from average flow in any relative small area
of tower. A 1/12 tower area seems to be an appropriately small area to determine maximum deviation whether
flow in that area is high or low. This is most difficult value to determine and requires examination at several
locations in tower cross section.Points to be kept in mind while selecting 1/12 th of area are
a) 1/12 th of area should be a continous section.
b) Section can be rectangular or formed from arc of circle at any location across cross-section of
distributor.

It is generally found that if liquid hydraulic design will ensure that random flow variation across tower cross
section does not exceed 10 % ,each point circle can be assumed to be equal area to simplify distribution quality
evaluation procedure.

Norton defines three categories of liquid distributors :

High performance distributors : DQ > 90%.
Intermediate performance distributors : DQ = 75 -90 %
Standard : DQ = 30 -65 %
Procedure

1) Assume no. of orifices such that there are minimum 65 points per m 2 of column Xn Each orifice
corresponds to a point circle.
Area of each point circle = Cross sectional area of column
no. of points

2) Assume pitch and fix liquid orifices on a triangular pitch. It has been seen that DQ for orifices on
circular pitch is always less than DQ for orifices on triangular pitch .As a first trial ,pitch can taken =
diameter of point circle . Draw point circles with orifice as center .
Ensure that distance orifice from column wall = min 25 mm for column of diameter > 16 “ .

3) Draw point circles starting with either distribution point at centre of column or non central
distribution point.

4) Calculate A,B & C . Calculate ‘B’ at several locations across Xn of tower. Calculate DQ.

5) Repeat with various values of pitch and location of distribution point and find out alternative which
gives highest value of DQ.
6) Locate gas risers such that
a) X nal area of gas riser is not less than 15 % of tower X nal area .
b) Gas riser area inside 5 0 % of tower cross sectional area is 40 to 60 % of tower area
c) Distance between gas riser & orifice is minimum 20 mm.

Table I gives value of DQ of some of GCL columns.

Other points to be noted are :

a) Distributor or redistributor must be close to the bed. If there is considerable spacing between
distributor & top of bed, the flow stream position on top of packed bed becomes uncontrolled and
reduces distribution quality.

b) Bed limiter must not alter flow pattern from distributor.

Distribution quality VS tower performance

Sensitivity of tower performance to liquid distribution quality depends only on number of stages each bed of
packing could achieve at its system base HETP . Beds of packing designed for many stages will be more
sensitive to distribution quality .This is indicated infig 1 .

It has been found that in 1 meter diameter debutanizer number of stages increased from 8 to 15 when distributor
with DQ of 36 % was changed to distributor with DQ of 93 %. In 380mm dia. iso octane/toluene system no. of
stages increased from 7 to 8.5 when distributor with DQ of 55% was changed to distributor with DQ of 85%.

Conclusion : The importance of liq. distributor increases as number of stages per section of column increases .
Liquid & gas flow , packing size and type does not affect the performance of distributor.
 TABLE - I

DETAILS OF LIQUID DISTRIBUTORS

Diam No. of DQ Pitch Dia. No. of Ratio of PCD of No. of Riser area in
eter orifice % (MM) of Risers X’nal gas riser Dist. side :
Riser area of (MM) Point Outside of
(NB) riser to per M² 50% Col.
column area.
14” 3+3+6=12 76 100 65.0 6 20.0 242 125.0 68 : 32
18” 3+3+6=12 72 130 80.0 6 18.6 303 73.5 62 : 38
22” 1+6+12=17 87 126 90.0 1+6 18.0 0;396 80.0 61 : 39
28” 1+6+12+12 79 140 90.0 1+6+6 20.0 0;380;576 78.0 55 : 45
= 31
 ANNEXURE III
(Prepared by KSS)

CENTRIFUGE
A centrifuge is an equipment utilizing centrifugal force for separation of liquid from solids. It is
essentially a development of Gravity Filter wherein the force acting on the liquid, instead of being
restricted to gravity, is enormously increased by utilizing centrifugal force. Due to good performance and
high cost, centrifuges are often referred to as the Rolls – Royces of solid – liquid separation.

G- level :- Centrifugal acceleration (G) is measured in multiples of earth gravity:

G rb
------ = -------
g g

Where,
G : Centrifugal acceleration , m/s2
n)  Angular velocity
rb : Basket radius, m
g : Acceleration due to gravity, 9.81 m/s2
n : Revolutions per second, s-1

G- force vs. Throughput :- As stated above Centrifugal acceleration is,

G = rb
The throughput capacity (Q) of a machine, depending on the process need , is roughly
proportional to the nth power of basket radius:
Q = C1 (rb)n

Where n is normally between 2 to 3 , depending on the characteristics of slurry to be

centrifuged and specifications of centrifuged cake, viz., purity, LOD, absorbent value.
It follows that large centrifuges can deliver high flow rates but separation is @ lower G- force;
vice versa, smaller centrifuges can deliver lower flow rate but separation is @ higher G – force.

Centrifuges are classified according to the mechanism used for solids separation:
Sedimentation Centrifuges and Filtration Centrifuges
1. Sedimentation Centrifuges :- In these centrifuges the separation is dependent on a difference
in density between the solid and liquid phases (solid heavier). Decanter centrifuge (S-1002;
Model No.: S-3400, Make: Pennwalt India Ltd.) used at GCL, Panoli is a type of
Sedimentation centrifuge. S-1002 is the only Decanter Centrifuge in GCL.

Decanter Centrifuge:- They are generally applicable to particle size range 1 – 5,000 m.
A decanter centrifuge is basically a settling tank of circular form mounted on an axis
(horizontal) and spun at high speed to produce separation of solids in decanter bowl. A screw
type conveyor carried internally and rotated relative to the bowl provides continuous
discharge. The speed with which the cake transports is controlled by differential speed
(between bowl and conveyor). High differential speed facilitates high solid throughput where
the cake thickness is kept minimum so as not to impair filtrate quality due to entrainment of
solids. Also cake de - watering is improved due to reduction in drainage path with smaller
 cake height; however, this is offset by fact that the higher differential speed also reduces cake
residence time. Therefore, an optimum differential speed is required to balance filtrate clarity
and cake dryness.

2. Filtration Centrifuges :- They separate the phases (solid – liquid) by filtration. Such filters
essentially consist of a rotating perforated basket equipped with a filter medium. Similar to
other filters, filtration centrifuges do not require a density difference between the solids and
the suspending liquid. If such density difference exists sedimentation takes place in the liquid
head above the cake. This may lead to particle size stratification in the cake, with coarser
particles being closure to the filter medium and acting as precoat for the fines to follow. The
capacity of filtration centrifuges is very much dependent on the solids concentration in the
feed.

As a general rule, sedimentation centrifuges are used when it is required to produce a

clarified filtrate whereas filtration centrifuges are used to produce a pure dry solid.

It is convenient to classify the filtration centrifuges into two broad classes, depending
on how solids are removed : fixed bed and moving bed

In the fixed - bed type, the cake of solids remains on the walls of the perforated basket
equipped with filter medium until removed manually, or automatically by means of a knife
arrangement. They are essentially cyclic in operation. Top Discharge & Bottom Discharge
Basket Centrifuges and Peeler Centrifuges are fixed – bed centrifuges.

In the moving – bed type, the mass of solids is moved along the basket by a ram.
Washing and drying zones can be incorporated in the moving - bed type. It is essentially
continuous in operation. Pusher Centrifuge is moving – bed centrifuge.

Basket Centrifuges ( Top & Bottom Discharge):-

They are applicable to particle size range 10 – 8,000 m.
The basket housing is supported by a three – point suspension called the three – column
centrifuge. Suspension of machine on three columns provides extensive compensation of any
imbalances of the system, thus dispensing with concrete or damper foundations.

The simplest of the fixed – bed centrifuges is the perforated basket centrifuge which
has a vertical axis. They are equipped to handle feeding, washing and discharge requirements
in a discontinuous filtration process with minimum attrition of the solids. The suspension to
be separated enters the machine via a stationary feed pipe. When the basket is filled, the feed
valve is closed by automatic control. The subsequent treatment consists of drainage of mother
liquor, washing of solids, drainage of the wash liquid, and discharge of the cake.

In the case of bottom discharge, a knife removes the cake towards the open centre,
leaving a thin residual layer of cake in the basket. In the top discharge machine, the solids are
removed manually ( can be removed pneumatically, mechanically, or by withdrawal of the
entire filter bag).

After removal of the cake, the centrifuge is ready for another charge. Programming the
sequence of events can be accomplished by a fully automatic control unit.
These types of centrifuges are in use in GCL.

Peeler Centrifuges :- They are applicable to particle size range 10 – 8,000 m.
The vertical axis of the basket centrifuge may cause some non-uniformity due to the effects of
gravity, with the accompanying problems when cake washing is used. This can be eliminated
by making the axis horizontal. This is known as Peeler Centrifuge.
A peeler centrifuge is designed to deal with a wide range of suspensions
discontinuously in lots. Each lot is subdivided into the necessary operations – Feeding,
Spinning-I to drain off mother liquor, Washing, Spinning-II to drain off wash liquid, Scraping
the cake. This adjustable lot cycle is mostly controlled automatically. The various operations
within a lot can be performed at constant or varying speed of the centrifuge drum.

The principal application is for high output duties with non-fragile crystalline
materials giving reasonable drainage rates which requires good washing and de-watering.

The suspension to be separated is fed to the centrifuge through a feed pipe. In the
filtration process the liquid filters through a filter cloth under the effect of centrifugal force.
Filtrate is drained through the perforated basket into the filtrate tank. The solid material is
retained in the basket by filter cloth and forms a uniform layer. The wash liquid is sprayed on
the solid layer through a wash pipe to wash it. The resulting wash liquor leaves the centrifuge
in the same manner as the main filtrate and is stored in wash liquor tank. The solids are
centrifuged until the desired residual moisture / organic liquid content is reached, and scraped
out by a hydraulically operated scraper knife down to a residual layer – which remains on
filter cloth. The solid material is discharged by means of a chute or screw. The scraper knife
can not be allowed to contact filter medium, a residual layer of solids / products remains in
the basket after each unloading. This serves as a precoat to prevent loss of fines to the filtrate
through the filter medium during next cycle. The disadvantage is that it also adds resistance to
filtration similar to filter medium. The residual layer may become glazed and impervious
from the rubbing action of the knife and a rinse may be frequently required to restore the
permeability.

Disadvantages & Advantages :- (Based on experience in Polymer project @ GCL,

Panoli).
Centrifuge used: 1250 MM dia (Make: ANUP)
Filtering area: 2.46 M²

1. Peeler centrifuge has parts rotating at high speeds and require high engineering
standards of manufacture, high maintenance cost, and special foundations or
suspensions to absorb vibrations. It is a very sophisticated and critical equipment.
2. Overflow of basket due to malfunctioning of feed-controller can cause loss of
solids to the filtrate as well problems associated with processing of filtrate.
3. It can be operated even @ 200°C of process temperature using suitable lubricating
oil. (It is already established in case of salt filtration from PES Polymer solution @
200 °C)
4. Cake gets washed thoroughly even with 0.5 Cake Volume wash liquid). Here the
cake mentioned is final product (DCDPS plant).
5. Throughput obtained is 5-7 TPD wet cake with 5-7 % w/w LOD (with 45 – 30
minutes lot cycle & approx. 150 kg lot and 24 hrs operation basis)[DCDPS
plant ]. Whereas throughput in case of centrifugation of PES slurry in water
containing 4-5 % Sulfolane is 11 TPD wet cake with 45 % LOD (with 20 minutes
lot cycle & approx.150 Kg lot and 24 hrs operation basis) [PES plant ]. In other
words, throughputs obtainable per unit filtering area for equivalent products are
always greater in centrifuges than in conventional filters.
6. Filter cloth fixing / replacement is easier as well as less time consuming. Filter
cloth can be replaced in two hours.
 Following are the details of the Peeler Centrifuges being used in GCL:
Make : ANUP ENGG. LTD., AHMEDABAD
Model : HZ – 125
MOC : SS – 304
NO. OFF : 8 NOS. [5 NOS. at Panoli , 2NOS. at Lote &
1 NO. at Dombivli]
Basket I.D. : 1250 MM
Basket Height : 625 MM
Basket rpm (max.) : 1200
G – level (max.) : 1000 g
Filtering Area : 2.46 M²
Nominal volume : 320 litres
Direction of rotation : Clockwise (looking from front side)
Motor : 100 H.P. (1440 rpm; flameproof)
Cost : Rs. 40.0 lacs (including motor & inverter)
(12/02/96)

Pusher Centrifuges :- They are applicable to particle size range 100 – 10,000 m.
Pusher centrifuges utilize continuous filtration for separation of suspended, fast –
draining crystalline and granular solids from liquids. A pusher ring plate intermittently moves
the cake over the screen or filter area in axial direction up to the edge, over which it is
discharged. No cloth can withstand the abrasion due to the cake forced on the cloth and
pushed over its surface. So the particle size range they are applicable to is generally coarser
(larger than 100m). The material is handled more gently. The pusher plate is usually
powered hydraulically. The pusher frequencies are around 60 – 100 per minute. Pusher
centrifuges require high feed concentrations to enable formation of a sufficiently rigid cake to
transmit the thrust of piston. The capacities of biggest centrifuges are 60 - 80 TPH. They are
used mainly in the Potash and salt industries and for other fertilizers.
 ANNEXURE IV
(Prepared by KSS)

FILTER
Types of filters
Filer cloth & filter aid
Calculation of filtration time

Types of filters
Open nutche
Agitated closed nutche (Rosenmund)
Spiral Used in GCL
Candle
Sparkler
Plate & frame
Cartridge (For polishing filtration)
Rotary vacuum drum
Pressure leaf
Bag
Horizontal belt
Open nutche

In this filter the filtration is by vacuum. The slurry is dumped in the nutche. Vacuum is applied &
filtration done. Cake is washed by putting the wash liquid above the cake. For effective washing
of the cake the wash liquid is distributed over the entire surface of the cake. For removal of the
cake person has to go inside the filter & remove the cake by showel. The design is very simple.
This type of filter comes in MS, SS, PP etc. Moc. The washing of cake is problematic as well as
discharge of the cake. Also being open this filter can not be used for foul smelling chemicals.

Agitated closed nutche

As the name suggests this is also nutche filter but agitation is possible. This filter has got lot many
other advantages over conventional nutche filters.

Because of its closed design it can be used for nasty smelling chemicals. The contamination with
foreign particles is avoided. Pressure can be applied in addition vacuum & hence higher
pressure differential is possible for filtration. Because of agitation the cake can be reslurried in
wash liquid & hence the washing is very effective compared to conventional open nutche. The
discharge of the filtered cake is possible by rotating the agitator in reverse direction. In one
direction the agitator is pressing the cake & in another direction it is discharging the cake. The
final cake can be fed directly to dryer for drying or it can be reslurried & transferred to another
vessel for further processing. The very popular MOC is SS.

Spiral

This is pressure filter. There is a shaft around which the rod is wound in the form of spiral. Over
this spiral filter cloth is wound. The slurry is fed from the side entry of the filter. The filtrate comes
out from the bottom outlet & cake is deposited inside the filter around the spiral. After the
pressure drop across filter increases the filter is to be opened, the cake discharged & fitted back
again for next filtration. Spiral filters are used where cake volume is small. It comes in MS, SS,
Hast C MOC.

Candle

Candle filter is small in nature. This is used after spiral filter & before cartridge filter. The slurry is
fed by pressure. Major filtration is done by spiral filter & only small load comes to this filter. As the
pressure drop across it increases it indicates that the filter cloth has clogged, filter is opened &
cake discharged. Popular MOCs are MS, SS, Hast C etc.

Sparkler

In this filter there are filter plates mounted one above another. The slurry is fed under pressure by
pump. The filtrate comes out from bottom hollow shaft & cake is deposited on various filter plates.
As the pressure drop across filter increases it indicates that the plate has been loaded with cake.
Filter is opened & cake discharged.

The combination of sparkler, spiral,candle & cartridge is used for fine filtration such as charcoal
filtration where the carbon particles are not allowed in the filtrate. In the conventional nutche filter
these fine charcoal is very difficult to trap.

The method of operation is that the slurry is circulated through the spiral by pump & filtrate
pumped back to the reactor. As the clear filtrate starts coming out the filtrate is paseed through
spiral. When the clear filtrate comes out from the spiral it is passed through candle. When the
clear filtrate comes out from the candle the filtrate is passed finally through cartridge filter for final
polishing filtration. The MOCs are SS, Nickel, Hast C etc.

Plate & frame

As the name suggests this filter consists of plate & frame mounted over a horizontal shaft. The
slurry is fed by pump & clear filtrate is collected. Being open in nature it can not be used for foul
smelling chemicals. The cake deposits on individual plate. After filtration is over the plates are
dismantled & cake removed. The plates & frames can be washed individually & refitted for next
filtration. The MOCs available are SS, PP,Hast C etc. In GCL this type of filter has been used only
for one or two applications.

Advantages:
 Because of its basic simplicity it is versatile & may be used for wide range of materials under
varying operating condition of cake thickness & pressure.
 Maintenance cost is low.
 It provides a large filtering area on a small floor space and few additional associated units are
needed.
 Most of the joints are external & leakage is easily detected.
 High pressure operation is usually possible.
 It is equally suitable whether the cake or the filtrate is the main product.

Disadvantage:
 It is intermittent in operation & continual dismantling is apt to cause high wear on the cloths.
 Despite the improvements mentioned above it is fairly heavy on labour.

Cartridge

Cartridge filters are for very fine filtration at micron level. The cartridge is disposable after it has
clogged. For fine charcoal filtration cartridge filter is used at the last stage to trap any carbon
which has passed the candle. The MOCs available are SS, Nickel, Hast C, PP etc. In
pharmaceutical industries this type of filters are very common.

In GCL the cartridge filters are used for Oxyclosanide (Veterinary drug) purification with fine
charcoal. CPF (Insecticide) purification with fine charcoal.

Rotary vacuum drum

The arrangement consists of a trough in which slurry is fed. There is rotating drum on which filter
cloth is fixed. Vacuum is applied within the drum. The drum rotates at very very slow speed. By
vacuum the slurry is sucked to the top of drum, it is filtered & cake is deposited on the filter cloth.
The deposited cake is cut by knife blade & discharged. This filter is very useful for pasty sticky
cake. The disadvantage being that it is not very much air tight so there is smell if some foul
smelling chemical is handled. For CPF filtration this filter is used where other types of filters failed
to give satisfactory performance.

Other types of filter

Some other types of filters are pressure leaf filter, bag filter, horizontal belt filter etc. For details
refer books.

Filter cloth & filter aid

The filter cloths used in GCL are PP, Cotton, Terylene etc. These cloths are available in various
mesh sizes. For fine filtration higher size mesh cloth is used. In addition to cloth porous filtration
tiles (Grindwell Norton) have also been used. These tiles act as filter cloth. These are very fine in
nature.

For fine filtration, say filtration of fine powered charcoal from slurry Hyflo is used. The procedure
is to prepare bed of hyflo over filter cloth by circulating slurry of hyflo prepared separately. Once
the filter cloth is embedded with hyflo the slurry is fed to the filter. The hyflo bed is easy to
prepare. In spiral filter almost always hyflo bed is prepared first & then filtration done.

Calculation of filtration time

By the very nature of the cake two types of cakes are there. Compressible & Non compressible.
In filtration two resistances are encountered:

Cake resistance (  )
Filter medium resistance ( Rm )

The pressure drop across filter consists of:

P across cake
P across filter medium

Total pressure drop is the addition of these two pressure drops. If the cake resistance is
independent of the pressure drop it is called non-compressible cake. If the cake resistance is
dependent on pressure drop it is called compressible cake. The method is different for
compressible & non-compressible cake. By nature almost all the cakes are compressible to some
extent or other. It is their degree of compressibility that decides whether the cake is compressible
or non-compressible.

Non compressible cake For calculation click--- Filtration time

From lab study time (t) vs filtrate volume (v) data is given. Two methods are available for
analyzing these datas.
Method-1
The t/v on y-axis is plotted against v on x-axis. From the slope & intercept the cake
resistance & filter medium resistance are calculated respectively. The values are fed
along with some other values in the INPUT DATA in program & OUTPUT RESULT is
obtained.

Method-2
This is quicker method but it does work for non-compressible cake to the satisfactory
level. Here v on y-axis is plotted against t on x-axis. In the lab data the cake thickness is
mentioned at the end of filtration. For half of the filtration cake thickness will be half of the
total thickness. Time for half filtration is found from the plot. Respective values are put in
the INPUT DATA & OUTPUT RESULT is obtained. The filtration factor ‘n’ is calculated
which is very close to 2 for most of the non-compressible cakes. Hence filtration is said to
follow the square law.

Compressible cake

The filtration data time vs filtrate volume should be collected at various differential pressures in
lab. For individual P the plots of t/v vs v are made. From these plots the cake resistance  and
filter medium resistance Rm are calculated. Table is made of , Rm P.
From these datas two graphs are drawn:
 Log () vs Log (P)
 Rm vs P
The slope of first graph is ‘s’. The equation showing the relationship between cake resistance, 
& pressure drop P is:
= 0 (P)S
When ‘s’ = ‘0’ the cake is non-compressible. As the value of ‘s’ goes away & away from ‘0’ the
cake tends to become compressible. The slope of first plot is ‘s’. From the graph by putting the
value of & P into above equation 0 can be calculated.
From Rm vs P graph the value of Rm is taken in the equation for nearest value of P. The cake
resistance  & P can be plotted on Log-Log scale also instead of taking Log & then plotting.
 ANNEXURE V
( Prepared by KSS )

VACUUM EQUIPMENTS

Introduction

In chemical process industries vacuum is very oftenly used for various purposes e.g.
 Vacuum distillation of high boiling organic compounds. At atmospheric distillation, the
products may deteriorate due to higher temperature, but in vacuum distillation
temperature is lower.
 For transferring a material from on place to another, pressure differential can be
created by applying vacuum.
 In chemical reactions where gases are generated negative pressure that is vacuum is
applied for scrubbing gases so that leakage to atmosphere is minimised.
 In filtration and drying operations also vacuum is used very frequently.
For creating vacuum common devices used are:
Ejectors
Venturi scrubbers
Vacuum pumps
Ejectors

These are the most common equipments used for creating vacuum because of their simple
design, No moving parts.
Basic principle
Type of ejectors
Points to be considered while purchasing steam jet ejector
Performance factors
Steam flow through ejector nozzle
Information required for ejector selection
Cost factors

Basic principles of steam jet ejector

Steam ejectors are pumps without moving parts. Construction and operation are extremely simple
in as much as only three main processes are involved.
The main parts are head, the driving nozzle and the diffuser. The main processes are expansion
of driving steam in the driving nozzle, mixing of the steam jet thus produced with the medium to
be drawn off ( air, gasses or vapors) and the conversion of velocity of this mixture into pressure in
the diffusers.
Steam jet ejectors operate at very high velocities. The velocity of the driving steam jet is nearly
always many times that of the speed of sound. The large volumes under vacuum can therefore
be easily handled. This is the reason why stem ejectors are eminently suitable as vacuum pumps.
These are simple is construction, constructed from wide range of material of construction, simple
in operation, robust in design, long working life, extremely safe operation.

Type of ejector

 Single stage
To compress from about 80 torr to atmospheric pressure. It is used generally for compression
ration < 10. It is suitable as pre-evacuator.
 Two-stage
Without intercondensation, to compress from about 30 torr to atmospheric pressure. It is used
for small suction quantities and as two-stage pre-evacuator.
 Three- stage
With intercondensation to compress from about 10 torr to atmospheric pressure. It is mainly
used to evacuate large condensers. These ejectors have a smaller steam and water
consumption than the two stage steam jet ejectors with intercondensation, if the working
conditions are same for both the units.
 Four-stage
With surface condenser for vacuum from about 3 torr to about 30 torr when the drawn off
medium must not come into contact with the cooling water or if the condensate is to be
recovered.
These ejectors are used extensively in the mineral oil industries.
 Five stage
With mixing condensers for vacuum from about 0.1 torr to about 3 torr. For suction quantities
from about 0.5 kg/hr upto about 1000 kg/hr condensable and in condensable vapor and
gases.
These ejectors are used for freeze drying where large quantities of water vapor are drawn off
from a vacuum of about 1 torr. Also used in steel de-gassing where large quantities of
incondensable gases must be drawn off.
 Six-stage
With intercondenser and an after condenser, for vacuum from about 0.05 torr to about torr.
These are used in the manufacture of synthetic fibres where vacuum between about 0.1 torr
and 0.3 torr is required. An after condenser is always used, if the exhaust from the final stage
can not flow direct into the atmosphere.
 2 Steam stage + water ring vacuum pump
For vacuum from about 0.5 to 5 torr. This combined pump is particularly suitable if barometric
erection is not possible.
 1 Stem stage + water jet
This combination is used for vacuum upto 10 torr.
 2 Steam stage + 1 water jet
This combination is very famous in the chemical process industries. This gives vacuum in the
range of 2 to 5 torr.

Points to be considered while purchasing steam jet ejector:

Suction capacity
Vacuum
Pre-evacuator
MOC
Type of condenser
Method of installation
Suction capacity

The steam consumption of a stem jet vacuum pump does not depend upon whether the whole or
only part of the suction is ejected. No steam is saved when a large capacity pump is operating
below full load. Hence it is important to make the plant to be held under vacuum as air tight as
possible, to determine the suction capacity as accurately as possible and to design the pump for
this capacity. However the capacity should always be chosen with a safety margin so as not to
endanger safe working.

The suction capacity is best calculated by weight that is Kg/hr. The suction capacity be weight is
made up of:
 Air leakage For calculating the air leakage rate click here Air leakage rate
 This leakage is that which enters through gaps in sealing. ( A hole of 1 mm² lets in
approximately 1.8 lbs/hr ).
 Gases, vapor & air released from the material handled in the plant
The condensable part must be distinguished from the noncondensables ( Inert gas ). The
molecular weight & temperature must be taken into consideration. For discontinuous
operation, it is important to know at what vacuum the gas or vapor quantities are released.
 Gases & air from the cooling water
The air tightness of the vacuum plants can vary greatly depending upon whether the plant is
mainly welded or whether it has many flange connections, valves, cocks, inspection glasses,
stuffing boxes etc. The type of sealing material, the condition of the sealing surfaces and the
degree of the use of the fillings are important. However it is possible to give a rough guide for
a practical evaluation of the suction capacity ( Not including of course vapors or gases
released in the chemical process ).

Vacuum

The vacuum should not be chosen higher than the absolutely necessary. Too high a vacuum
leads to unnecessarily large suction pipelines, unnecessarily large steam jet ejectors, excessive
steam consumption and excessive cooling water consumption.

Pre-evacuator

If a vacuum plant is in constant operation, evacuating time on start-up is generally not an

important factor and it is not necessary to provide special pre-evacuator ( Start-up ejector ). If
however the vacuum plant has to be started up frequently, a short evacuating time is desirable. To
achieve this a pre-evacuator is used.
A pre-evacuator is normally a single stage steam jet ejector with a large suction capacity, which is
put to operation simultaneously with the steam jet ejector. Together with the ejector it evacuates
the plant very quickly to an intermediate vacuum, say upto 150 torr. At 150 torr almost 80 % of the
original volume of air in the plant has already been pumped out. The pre-evacuator is then shut-
off and the steam jet ejector alone evacuates in the remaining time to the required working
vacuum.
The pre-evacuator has a high steam consumption. However this is not of great importance since
it operates only for 10-15 minutes during the start-up of the plant. The steam consumption of pre-
evacuator decreases considerably with increasing stem pressure. The steam consumption at 10
atm is only 40 & of that at 3 atm.

MOC

A steam jet vacuum pump should be constructed of a material at least resistant to corrosion as
the plant to which it will be attached. The various MOCs available are: MS, SS-304, SS-316,
Graphite, MS/FRVE lines, PP etc.
The jet comes in various MOCs such as: SS-304, S-316, Titanium, PTFE etc.

Type of condenser

Mixing condensers are simple, reliable and inexpensive. Mixing condensers like steam jet
ejectors can be manufactured from many different materials and also make the best use of
cooling water.
With surface condenser, the cooling water does not enter the vacuum. The cooling water is
separate from the condensate. Surface condensers must be used when
 The cooling water must not be contaminated by the condensate of the vapors drawn off.
 The condensate is to serve as cooling water for units such as steam turbine condenser.
 The condensate is to be reclaimed.
 Ammonia vapors are contained in the drawn off vapors because insoluble condensate are
quickly formed in mixing the condensates which would cause breakdowns, that is, blocking
of the water outlet etc. Even small amount of ammonia are dangerous.

Method of installation

 Barometric erection
If possible steam jet ejectors should be erected barometrically because then the water flows
from the condensers without pumping. A column of 760 mm Hg ( 29.92” ) high balances the
pressure of the atmosphere. Since water is 13½ times lighter than mercury, the
corresponding height of water is 10.33 m. This height of 10.33 m is called the barometric
height. According to the pressure in difference condensers at reduced height may be
sufficient to ensure a free outlet of the cooling water. At a barometric height of 760 mm Hg
and a condenser pressure of 300 torr. A height of 6.3 m ( 20.7” ) is necessary.
Since the water may be mixed with same air and hence will no longer have a sp.gravity of 1,
it is safer to assume a full barometric height of 11 meters.
 Non barometric erection
In many cases it is not possible to install a steam jet ejector 11 m ( 36 ft ) above the water
drain. There are various alternatives depending on the height available.

Performance factors:
Driving steam
Suction pressure
Discharge pressure
Cooling water
Dry and saturated air
Gas and vapor densities
Driving steam

Steam jet ejectors can be designed for driving steam from about 15 psia to 600 psig. A steam jet
ejector operates at the driving steam pressure for which it is designed. It is necessary to ensure
that the designed pressure is maintained otherwise a breakdown is possible. On the other hand
operating above the design pressure results in steam wastage. In this case it is necessary to
install a pressure regulator.
It is essential to check what working steam pressure is available at the site of steam jet ejector.
The ejector should be designed for that pressure. ( This pressure is often much lower than the
boiler pressure ). Steam jet ejectors operate most efficiently with dry saturated steam or
superheated steam. Low superheat can be disregarded. Wet steam is not desirable at all.
The motive steam design pressure must be selected as the lowest expected pressure at the
ejector steam nozzle. The unit will not operate stably on steam pressures below the design
pressure.
Recommended steam design pressure = Minimum expected design pressure at the ejector
nozzle – 10 psi.

This design basis allows for stable operation under minor pressure fluctuations. An increase in
steam pressure over design pressure will not increase vapor holding capacity for the usual “
Fixed capacity ejector “. The increased pressure usually decreases capacity due to the extra
steam in the diffuser. The best ejector steam economy is attained when the steam nozzle and
diffuser are proportionated for a specified performance.
For a given ejector an increase in steam pressure over the design value will increase the steam
flow through the nozzle in direct proportion to the increase in the absolute steam pressure. The
higher the actual design pressure of an ejector the lower the lower the steam consumption. This
is more pronounced on one or two stage ejectors. When this pressure is above 350 psig, the
decrease in steam requirement will be negligible. As the absolute suction pressure decreases
advantage of high pressure steam becomes less pronounced. In very small units the physical
size of steam nozzle may place a lower ceiling on the pressure.
For the ejectors discharging to the atmosphere, steam pressure below 60 psig at the ejector are
generally uneconomical. If discharge pressure is lower in multistage units, the steam pressure at
inlet can be lower. Single stage ejectors designed for pressure below 200 mm Hg (abs) cannot
operate efficiently on steam pressures below 25 psig. The first stage for two of a multistage
system can be designed although perhaps not economically to use stem pressure below one
psig.
To ensure stable operations the steam pressure must be above a minimum value. This minimum
is called motive steam pick-up pressure, when the pressure is being increased from unstable
region.

Effect of wet steam

Wet steam erodes the ejector nozzle and interferes with performance by clogging the nozzle with
water droplets. The effect on performance is significant and is usually reflected in fluctuating
vacuum.
Effect of superheated steam
A few degree of superheat (5-15°C) is recommended, but if superheated steam is to be used its
effect must be considered in the ejector design. A high degree of superheat is of no advantage
because the increase in available energy is offset by the decrease in the steam density.

Suction pressure

The suction pressure of an ejector is expressed in absolute units. The suction pressure follows
the ejector capacity curve, varying with the non-condensable and vapor load to the unit.

Discharge pressure

All types of ejectors are sensitive to discharge pressure just as they are to the steam supply
pressure. Normally designed ejectors are suitable for operating against a pressure only slightly in
excess of atmospheric. In most cases ejectors can be designed to discharge at a pressure of 5
psi, provided a considerable increase in steam consumption can be permitted. It must be
appreciated, however, that the discharge of an ejector is contaminated with incondensable gases
and is therefore in many cases unsuitable for further use.
The pressure drop through discharge piping and aftercooler must be taken into consideration.
Discharge piping should not have pockets for condensation.

Cooling water

The cooling water temperature is of great importance.

Well water : 10 to 15°C

River water : 5 to 25°C
Re-circulated water : 10 to 28°C
Seawater : 15 to 30°C
Steam jet ejectors must be designed for the maximum cooling water temperature available.
Steam and water consumption are greatly dependent on the design temperature as can be seen
from the following example.

A unit to eject at 5 torr using steam at 45 psig.

The steam consumption at a cooling water temperature of 25°C is double that of 15°C. If the
cooling water temperature varies to a large extent throughout the year, it is advisable to adjust the
steam consumption in relation to the water temperature by suing different nozzles, or by altering
the steam pressure ( The lower in only possible at high pressure ). The pressure of the cooling
water should not oscillate as this may affect the vacuum.
With missing condensers it is not necessary to have any particular water pressure. With water jet
condensers a minimum pressure of 30 psig is required. With surface condensers the pressure
loss in the tubes must be overcome.
In general the cooling water outlet temperature must not exceed 45°C to 50 °C, otherwise chalk
deposits will cause breakdown.

Amount of gases released from cooling water, W = G/t 1, where

W = Air liberated lbs/hr
T1 = Inlet temperature °C
G = Cooling water, gal/min

Ws X L
GPM of cooling water required = ------------, where
tw X 500

Ws = Lbs of steam to condense

L = Latent heat of vaporisation, usually taken as 1000 BTU/lb for
process application and 950 BTU/lb for turbine exhaust steam.
tw = Cooling water temperature rise °F

Dry and saturated air

When an ejector is maintaining a condenser vacuum, the air extracted by the ejector is saturated
with water vapor. The amount of water vapor entrained by the dry air which leaks into the
condenser depends upon the temperature of the mixture and the vacuum at the ejector suction.

Gas and vapor densities

When an ejector or thermo-compressor is required to handle chemical gases or vapors, it is

necessary that the gas density ( For failing this the temperature & molecular weight ) be known.
The compression of a given weight of heavy gas requires less operating steam than the same
weight of light gas. As a simple example on pound of air is more easily handled by an ejector than
one pound of water vapor. The correct proportioning of the ejector & thermo-compressor
therefore necessitates detailed knowledge of the properties of the gas or vapor to be
compressed.

Steam flow through ejector nozzle

The quantity of steam passing through any ejector nozzle varies directly as the steam pressure,
but decreases with an increase of steam temperature. It is therefore essential that the maximum
steam temperatures be stated to enable the ejector to be correctly designed.
The weight of steam flowing through an ejector nozzle may be obtained from the following simple
formula.
S = KPD2, where

S = Steam flow in lbs/hr

P = Steam pressure at the upstream of nozzle in psi (abs)
D = Diameter of nozzle bore in inches.
K = A constant which depends on the steam pressure and temperature.

Information required for ejector selection

To enable the manufacturer to design most suitable type of vacuum system, it is essential that
correct data be available upon which to base proposals & designs. Whenever possible the
following information should be supplied with any enquiry.
 Brief description of purpose for which equipment is required.
 Weight of air or vapor to be handled in lbs/hr. If a mixture of air, vapor and /or other non-
condensable gases. Their approximate proportions should be stated
Lbs/hr of condensable vaporous & lbs/hr of non-condensable gases, either dissolved,
injected or carried in process formed by reaction, air leakage etc.
 The absolute pressure required at the ejector suction in inches mercury, or millimeters
mercury.
 If ejector is required to discharge at other than atmospheric pressure, the maximum
discharge pressure must be stated.
 If normal steam pressure and also the minimum steam pressure at which the plant is required
to operate. If steam is superheated give maximum temperature.
 Maximum temperature of cooling water.

Cost factors

Of all the classes of Chemical Engineering Equipments, ejectors are unique in that the actual
operating cost is usually much greater than the installed equipment cost. Ejectors have low first
cost because they are simple in design, have no moving parts and occupy a small space.
In general operating cost increases as the ejector steam pressure increases. If an ejector is
supplied with steam above the design pressure, steam consumption will be increased in direct
proportion to the ratio of actual operating pressure to the design pressure. It is good practice to
install a stem pressure regulating valve to maintain a uniform steam pressure at the ejector.
The other things which affect the performance of on ejector are.
 Discharge pressure
 Wet and superheated steam
 Load variation
 Back pressure
 Capacity and suction pressure
 Number of stages
 Condenser requirements
 Material of construction

Water jet ejectors

As the name suggests it water jet ejector the motive fluid is water instead of steam. Ejectors using
water as the motive fluid are designed for reasonable non-condensable load together with large
condensable flows. Water pressures as low as 10-20 psig are usable, while pressures of 40 psig
and higher will maintain a vacuum of 1-4 inches of Hg (abs) in a single stage unit. Combination of
water & steam ejectors are used to effectively handle a wide variety of vacuum situations. The
water ejector serves to condense steam from the steam ejector.
Water ejectors and water jet eductors are also used for mixing liquids, lifting liquids and pumping
+ mixing suspended solids & slurries ranging from ½” to 24. The ejectors are usually in pumping
air or gases while the eductors are used in pumping liquids.
Unlike steam jet ejector there is pump in water jet ejector which is the moving part.

Venturi scrubbers

These are used for scrubbing gases coming out from chemical reactions. For example gases
evolved such as S02, HCl are scrubbed in caustic soda solution. For venturi scrubbers slight
vacuum is required to ensure that the gases do not leak to atmosphere.
The venturi scrubbers used at GCL are of model: 25D, 40D, 60D of H. K. Industries make.
Vacuum pumps

Vacuum pumps are mainly used for handling large quantities of air leakage at relatively low
vacuum in operations such as filtration, drying etc.
The vacuum pump may be liquid ring type or dry pump. Amongst liquid ring more common are
water ring and oil ring vacuum pumps.

Water ring vacuum pump

Water ring machines are used principally as vacuum pumps, but can also serve as compressors
for special applications. Their main advantage lies in their simple and trouble free operations.
They may be used for handling moist air as well as aggressive gases and vapors, provided that
the material of construction is properly selected. Dust-laden gases can also be handled without
difficulty.
Water ring machines are absolutely oil-free and do not contaminate the medium passing through.
Compression takes place with a small temperature rise since most of the heat is absorbed by the
ring fluid which also acts as a seal. For special applications, an alternative sealing fluid instead of
water may be used.

Principle of operation

The casing of the machine is cylindrical with control discs on either side. Within this is an
eccentrically located impeller with curved blading. The liquid introduced, preferably water is set
into rotary motion by the blading, thus forming a water ring concentric with the casing. Cells are
formed between the ring and the blades of the impeller, which vary in volume from a maximum to
a minimum with each revolution. The conveying medium enters the working chamber through the
openings in the lateral control discs and is compressed by the decrease in volume of the cells. It
then flows together with the cooling water through discharge openings the control disc to the
delivery port. The impeller runs in bearings arranged in brackets outside the casing covers. At the
bottom of the casing an automatic drain valve is provided, through which the ring fluid can flow off
when the pump is at a standstill. The pump should be brought into operation without the ring fluid
whereby easy starting is ensured.

Drive

The pump is usually driven by a direct coupled electric motor. Pump with an automatic water
drain valve can be driven by squirrel-cage motors with star/delta starting.
Smaller units may depending on the regulations of the local electricity suppliers, be switched on
by means of direct on line starters.

Installation as a vacuum pump

With normal installation using water as a ring and cooling fluid, the water as well as the exhaust
air are delivered together into the drains. A silencer cum water separator can be provided as an
optional accessory if required. This would have to be fitted in the exhaust piping. A non-return
valve is normally supplied for fitting in the suction piping so that in case of operational failure, the
water does not flow back into the vacuum line. A vacuum relief valve can also be supplied if it is
desired to limit the vacuum. At a pre-set maximum value the valve opens automatically allowing
atmospheric air into the piping. The ultimate vacuum attinable by this type of machine will depend
upon the temperature and evaporation properties of the ring fluid.

Installation as a compressor

When ring compressors are equipped on the pressure side with a water separator in which the
ring water is separated from the compressed air. This is fitted with a float valve, which allows the
water to escape. It is possible to re-circulate this water with a closed system, similar to that for a
vacuum pump. The non-return valve for a water ring compressor is to be fitted to the pressure
piping. The working pressure is limited by fitting a safety valve in the delivery piping.

Material of construction

The standard material of construction is cast iron. However, for special application these can be
supplied in various materials such as bronze, Gunmetal, SS-304, SS-316, Rubber-lined etc.

Details of normal water ring vacuum pumps

Speed, RPM Max. Suction Recommended Water consumption

M3’hr Motor, HP Lit/min.

2850 49 4 7
2850 81 5 9
2850 123 7.5 14
1450 160 7.5 16
1450 220 10 20
1450 330 15 30
1450 440 20 40
980 725 30 60
980 850 35 80
980 1080 40 100
725 1500 65 150
725 2030 90 200
580 3650 200 350

Oil ring vacuum pumps

Oil ring vacuum pump is similar to water ring vacuum pump in operation except that the sealing
fluid used is oil instead of water. The advantage of oil ring vacuum pump is that high vacuum can
be obtained due to much lower vapor pressure of oil as compared to water. The disadvantage is
that after some usage oil gets contaminated with chemicals and vacuum starts dropping. The
recovery of oil is very expensive. Because of this reason the oil ring vacuum pumps are not very
common.

Dry pump

By using water ring/oil ring vacuum pump the chemicals/solvents get contaminated with water/oil.
The recovery of these solvents become very costly affair. In such cases dry pump scan be used
though the initial costs are high.
The principle of dry pumps is different from ring pump. In dry pump drip oil lubrication is used,
whereas in ring pump, there is continuous seal of liquid. In dry pump oil used is once through
pass. Heat exchanger can be installed on the discharge side of dry pumps to condense the
gases. The condensed solvent may be contaminated with little bit of oil, which is used for drip
lubrication of dry pump. The condensed solvent can be distilled out to separate it from oil. To
boost the capacity of dry pump mechanical boosters can be used.
Combination of one roots pump and two dry pumps will give capacity of 800 M 3/hr at 1 torr
absolute pressure. The roots pump is volumetric flow device. At high vacuum mass flow will be
lower. The oil consumption for both the dry pumps is 250 CC/8 hr.
For 800 M3/hr flow at 1 torr air = 1.22 kgs/hr.
The combination of two dry pumps and one roots pump can be used in place of 4-stage steam jet
ejector.
Advantages over steam jet ejector
 No steam consumption
 Low HP required
 Pollution free operation

There are dry pumps which are totally dry, no drip lubrication required. In steam jet ejectors the
last stage discharges to atmosphere, thereby contaminating the atmosphere with
chemical/solvent vapors, whereas in case of dry pumps chilled condenser can be installed to
recover any solvent loss to atmosphere, thereby improving environment as a whole.
ANNEXURE VI
(Prepared by KSS)

TRANSFER PUMPS
For transferring liquid from one place to another three methods are used. Pressure, Vacuum,
Pump. Pumping is very common & popular in chemical process industries. The various types of
pumps are as shown below:

Types of pump

Centrifugal pump Positive displacement pump

Coupled Mono block Magnetically Reciprocating Rotary

Pump Pump coupled pump Pump pump
Air operated Plunger Diaphragm Gear Shuttle block Lube Screw
Double pump type metering pump pump pump pump
diaphragm pump

Single diaphragm Double diaphragm

pump pump

Centrifugal pump

The centrifugal pump develops its pressure by centrifugal force on the liquid passing through the
pump and generally applicable to high capacity, low to medium head installations. For high
discharge head centrifugal pumps can be multistage instead of single stage. e.g multistage
centrifugal pump is used for boiler feed water.

 Basic parts of a centrifugal pump

 Arrangement of pump
 Hydraulic characteristics
 Pump affinity laws
 Selection of pump

Basic parts of a centrifugal pump

Impeller

The three common types of impellers used are:

 Fully closed: Used for high head & high pressure applications.
 Semiopen: Used for general purpose applications. It has got open vane tips at the
Entrance to break up suspended particles & prevent clogging.
 Open: Used for low heads, suspended solid s applications, very small flows. These
impellers are available in any nearly any material of construction as well as rubber,
rubber lined, teflon lined, glasslined. The lined impellers are of open type.
Casing

The casing may be constructed of a wide variety of metal, lined material, plastic materials etc.

Shaft

Care should be taken in selecting the shaft material. It must be resistant to the corrosive action of
the process fluid yet posses good strength characteristic for design. Generally sleeves are used
over shaft to prevent direct corrosion / mechanical danger to shaft. These sleeves may be metal,
ceramic, rubber etc.

Bearings

The bearings must be adequate to handle the shaft load without excessive wear provided
lubrication is maintained. In all cases bearings should be of the outboard type, that is not in the
process fluid unless special conditions prevail to make this situation acceptable.

Packing & seals on rotating shaft

Conventional soft or metallic packing in a stuffing box is satisfactory for many low pressure, non-
corrosive fluid systems. Special packings such as mechanical seals are commonly used for
corrosive fluids when the pressure becomes high above (about 50 psig) or the fluid is corrosive,
additional means of sealing the shaft must be provided. Particular care must be taken in handling
and using the mechanical seals.

Arrangement of pump

 Pumps in series
Sometimes it is advantageous or economical to use two or more pumps in series to reach the
desired discharge pressure. In this situation the capacity is limited by the smaller capacity of
any one of the pumps at its speed of operation. The total discharge pressure of the last pump
is the sum of individual discharge pressure of the individual pumps. The pump casing of each
stage (particularly the last) must be of sufficient pressure to withstand developed pressure.

 Pumps in parallel
Pumps are operated in parallel to divide the total load in two or more smaller pumps. The
individual pump operates on its own characteristic curve. The capacity gets added up.

Hydraulic characteristics :
 Capacity
 Total head
 Suction head or suction lift
 Discharge head
 NPSH & pump suction

1. Capacity
 It is the rate of liquid or slurry flow through a pump. For proper selection and corresponding
operation a pump capacity must be identified with the actual pumping temperatures of the
liquid in order to determine the proper power requirements as well as the effects of viscosity.
Pumps are normally selected to operate in the region of high efficiency and particular
attention should be given to avoiding the extreme rights side of the characteristic curve where
capacity & head may change abruptly.

2. Total head

The pressure available at the discharge of a pump as a result of change of mechanical input
energy into kinetic & potential energy. This represents the total energy given to the liquid by
the pump. Head is expressed as feet of liquid being pumped. The head is independent of the
fluid being pumped & is therefore the same for any fluid thorough the pump at a given speed
of rotation & capacity.

3. Suction head or suction lift

The total suction head is the difference in elevation between the liquid on the pump suction
side and the centerline of the pump + the velocity head. When the liquid level is below the
pump centerline the difference in elevation is known as suction lift.

Total suction head (TSH) = Static head – friction head loss.

Total suction lift (TSL) = Static lift + friction head loss.

4. Discharge head

The discharge head of a pump is the head measured at the discharge nozzle 7 is composed
of static head, friction losses through pipes, fittings, contractions, expansions, entrances &
exit, thermal system pressure.

5. Velocity head

As a component of both suction and discharge heads, velocity head is determined at the
pump suction & discharge flanges respectively and added to the gage reading. The actual
pressure head at any point is the sum of the gage reading + the velocity head. The values of
velocity heads are usually small and negligible often.

6. NPSH and pump suction

The net positive suction head (NPSH) is very important criteria for centrifugal pumps. There
are two types of NPSH. NPSHa and NPSHr.

NPSHa = Net positive suction head available = Pressure on suction side – friction loss –
vapour pressure of liquid being pumped at
pumping temp.

NPSHr = Net positive suction head required. It is specified by the pump manufacturers. The
NPSHr is very important consideration in selecting a pump, which might handle liquids at or
near boiling points or liquids of high vapour pressures. If this consideration of NPSH r is
ignored, the pump may well be inoperative in the system or it may be on the borderline and
become troublesome and expensive. The significance of NHSH R is to ensure sufficient head
of liquid at the entrance of the pump impeller to overcome the initial flow losses of the pump.
This allows the pump impeller to operate with a full bite of liquid essentially free of flashing
bubbles of vapour due to boiling action of the fluid poor suction condition to cavitation in
 pump impellers and this is a contributor at which pump can not operate for a very long without
physical erosion, damage to the impeller.
Cavitation in a centrifugal pump or any pump develops when there is insufficient head for the
liquid to flow into the inlet of the pump allowing flashing or bubble formation in the suction
system and entrance to the pump. Each pump design or “family” of dimensional features
related to the inlet and impeller entrance pattern requires a specific minimum value of NPSH
to operate satisfactorily without flashing, cavitating and loss of suction flow.
Under cavitating conditions a pump will perform below its head performance curve at any
particular flow rate. Although the pump may operate under cavitating conditions, it will often
be noisy because of collapsing vapour bubbles and severe pitting and erosion of the impeller
often results. This damage can become so severe as to completely destroy the impeller and
create excessive clearance in the casing. To avoid these problems, the following are few
situations to watch.
 Have NPSH available at least 2 feet of liquid greater than the pump manufacturer
 Requires under worst possible operating conditions.
 Internal clearance wears inside pump.
 Entrained gas (non condensables)
 Deviations or fluctuations in suction side pressure, temperature increases, low liquid
level.
 Piping layout on suction, low particularly tee intersections, globe valves, baffles etc.
 Liquid vortexing in suction vessel, thus creating gas entrainment into suction piping.
 Nozzle size on liquid containing vessel may create severe problems if inadequate. Liquid
suction velocities in general are held at 3-6.5 ft/sec. Usually as a guide the suction line is
at least one pipe size larger than the pump suction nozzle.

The NPSH required by the pump is a function of the physical dimensions of casing, speed
and type of impeller and must be satisfied for proper pump performance. The required NPSH
of a pump increases as the pump speed increases.

RPM  NPSHr0.75
For this reason many critical suction condition installations use relatively slow speed pumps.

Pump affinity laws

 For change in speed with fixed impeller design, diameter & efficiency the following conditions
and characteristics change simultaneously.

Q2 = Q1 X (N2/N1) Q2 = Flow rate at N2 RPM

H2 = H1 X (N2/N1)2 Q1 = Flow rate at N1 RPM

BHP2 = BHP1 (N2/N1)3 H = Head of pump

BHP = Break horse power

The relations do not hold good exactly if the ratio of speed change is greater than 1.2 to 2 nor
do they hold good if the suction conditions become limiting such as NPSH.

 For change in impeller diameter at a fixed speed and efficiency.

Q2 = Q1 X (D2/D1)
D1 = Impeller diameter
H2 = H1 X (D2/D1)2
 An impeller can be cut from one size down to another on a lathe and provided the change in
diameter is not greater than 20 percent, the conditions of new operation can be described by
the type of calculation shown in above formulae.

Selection of pump :
 Capacity
 Head
 Effect of viscosity
 Effect of temp. rise and minimum safe flow
 MOC
 Operating condition of fluid / nature
 System condition
 Motor details

Capacity
Capacity is decided by the batch time. Batch time is fixed by the capacity of the plant. For
example if 10000 litres of liquid is to be transferred from one vessel to another in 1 hr the capacity
becomes 10 M3/hr. For most of the normal operations the capacity of the pumps used is of the
order of 10,000 lit/hr.

Head
After deciding the capacity the head is to be calculated. For this the total pressure drop in the
suction & discharge piping is calculated. The elevation of discharge point & suction point is taken
into account. The pressure conditions at suction & discharge are also considered.
For discharge piping use velocity in pipeline as 6-8 ft/sec ( Source : Ludwig vol.1 ). For suction
piping use velocity <= 3-5 ft/sec. Using the flow rate & velocity fix up the line size. If the line size
is coming odd, select nearest size. The size of pipes available is 15, 20, 25, 40, 50, 65, 80, 100,
150, 200, 250, 300 NB etc. After deciding the pipe size calculate back the velocity in pipe.

Find out number of fittings, valves in pipe as well as total straight length. Calculate equivalent
length of fittings, valves, expansion, contractions etc. Add straight length to equivalent length. The
total length thus obtained is known as effective length L e.

Determine pressure drop in pipeline. For calculation click-- Pressure drop in liquid line
Calculation for suction & discharge piping is to be calculated.

Find out the static head of liquid on suction and discharge side. Static head on suction side is the
level difference between top of liquid on suction side and the centerline of pump. Static head on
discharge side is the level difference between centerline of pump and the discharge point, where
liquid disengagement takes place.

Find out the pressure at suction & discharge end.

Apply Bernoulli’s equation & calculate the head to be developed by the pump.

P1 + V12 + g Z1 + H = P2 + V22 + g Z2 + P
---- ----- ------ ---- ----- ------ , where
 2gc gc  2gc gc
r = Density of liquid, kg/m3
P1 = Pressure on suction side of pump, kg/m2(abs)
Z1 = Static head on suction side, m
V1 = Velocity on suction side, m/sec
 H = Total developed head by pump, m
P2 = Pressure on discharge side of pump, kg/m2(abs)
V2 = Velocity on discharge side, m/sec
Z2 = Static head on discharge side, m
DP = Total pressure drop in suction & discharge piping, m

After having calculated capacity and head refer to the characteristic curves of pumps given by
pump manufacturers. By referring to the various characteristic curves select pump to give best
efficiency. In the characteristic curves the brake horsepower is also given. Select particular
impeller size. Note what is the break horsepower for that particular impeller at cut off point. Use
nearest higher HP motor for your service. Find out the NPSH r from the characteristic curves and
make sure that the NPSHa is greater than that required by at least 2 mlc.
For critical services, where vacuum is there on suction side or for pumping highly volatile liquids
such as propylene, ammonia, DMA, Isobutylene, the NPSH becomes very important.
In such cases pumps are to be selected where the NPSHr is very low.
The characteristic curves are given for water. If any other liquid is pumped density & viscosity
effects should be taken into account. For liquids other than water

Brake horse power = BHPwater X liquid in gm/cc.

Effect of viscosity
When viscous liquids are handled in centrifugal pumps, the brake horsepower is increased, the
head is reduced and the capacity is reduced as compared to the performance with water. The
corrections are negligible for the liquids whose viscosity’s are in the range of water but it becomes
significant above 10 centistokes for heavy materials.

For agrochemicals product the effect of viscosity is negligible but for polymeric solutions where
the viscosities are very high the effect in tremendous.

 For pumping viscous liquids the viscosity correction factor is to be applied. When the
required capacity & head are specified for a viscous liquid, the equivalent capacity
and head can be determined using the viscosity correction factors shown below.
Qvis
Qw = ------ Qw = Capacity of equivalent water
Cq Hw = Head of equivalent viscous liquid
w = Efficiency while pumping equivalent water
Hvis Qvis = Capacity of viscous liquid
Hw = ------ Hvis = Head of viscous liquid
Ch vis = Efficiency of viscous liquid
Cq= Capacity correction factor
vis Ch = Head correction factor
w = ------ C = Efficiency correction factor
C

For a viscous liquid the capacity & head has been calculated as per the procedure
mentioned previously. Look at the viscosity correction chart. For the viscous capacity
Qvis & viscous head Hvis read the value of capacity correction factor Cq and head
correction factor Ch and efficiency correction factor C. For finding out the correction
factors the kinematic viscosity of the viscous liquid is required.
After having found the correction factors calculate the equivalent capacity & head for
water. Refer to various characteristic curves of pump & select the pump giving best
efficiency. While selecting the pump keep in mind the types of pumps already in use
in the company. This is necessary to avoid so many types of pump otherwise keeping
spare parts will be problematic.
 After finding out the efficiency on water, w calculate the efficiency while pumping
viscous liquid, vis using above mentioned formula.

Now having calculated the efficiency for pumping viscous liquid vis the horsepower
for pumping viscous liquid is calculated as below.

BHPvis = QVIS X Hvis X vis

The chart below shows how important are the correction factors

For water For viscous liquid. = 1.8 gm/cc

 = 160 cp

At operating condition At operating condition

1. Capacity = 18.3 m3/hr 1. Capacity = 15 m3/hr
2. Head = 29.4 m 2. Head = 30 m
3. Efficiency = 61 % 3. Efficiency = 34.16 %
4. BHP = 3.25 HP 4. BHP = 8.78 HP

At cut off point At cut off point

1. Capacity = 40 m3/hr 1. Capacity = 38.4 m3/hr
2. Head = 18.5 m 2. Head = 16.65 m
3. Efficiency = 57 % 3. Efficiency = 38.76 %
4. BHP = 4.9 HP 4. BHP = 11 HP
Hence recommended motor = 5 HP Hence recommended motor = 15 HP
Or 10 HP with relay.

For above mentioned viscous liquid in table the horsepower required is 170 % more than
that required for pumping equivalent water. The viscous efficiency is 44 % lower than
equivalent water efficiency.

Effect of temp. rise & minimum safe flow

When a pump operates near shut-off capacity and head or is handling a hot material at suction, it
may become overheated and create serious suction as well as mechanical problems. To avoid
overheating due to low flow, a minimum flow should be recognized as necessary for proper heat
dissipation. However it is not necessarily impossible to operate at near shut-off conditions
provided

 It does not operate long under these conditions as temp. Rise per minute vary from less than
1°F to 30-40°F.
 A bypass is routed or recycled from the discharge through a cooling arrangement and back to
suction to artificially keep a minimum safe flow through the pump while actually withdrawing a
quantity below the minimum.

MOC

The pumps come in metallic & non metallic material of constructions. The MOCs are:

Metallic: CI, CS, SS-304, SS-316, Hast B, Hast C, Zirconium, Alloy 20

Non metallic: PP, PVDF, PTFE, CI/PP, CI/PVDF, CI/PFA, GRP, Graphite,
The efficiency of metallic pumps is higher than non metallic pump because in metallic pumps
closed as well as semi-open impellers are possible. In non metallic pumps only open impellers
are possible because of basic problem of moulding. The efficiency of open impellers are lower
than closed / semi-open. The metallic pumps are more sturdy so mechanical damage chances
are low. Many times the pump manufactures do not take the responsibility of MOC. In such cases
the MOCs are to be mentioned by the customer himself.

Operating condition of fluid / Nature

The operating condition details such as temp, pressure, density, viscosity, % slurry if it is
applicable, nature of solids whether abrasive or soft must be clearly mentioned in the data sheet.
The corrosive nature of the pumping fluid also must be clearly mentioned. Incomplete information
may lead to wrong selection of pump which may ultimately fail at plant level.

System condition

The pump manufacturer must know if the suction side of the pump is associated with vacuum
equipment or is to lift the liquid. This can make a difference as to the type of impeller, Suction
opening to be provided. If the system operates intermittently it should be noted. A piping diagram
is often helpful in obtaining full benefit of the manufacturer’s special knowledge.

Pumps operating under suction lift

As centrifugal pumps need priming, there are various methods to prime. Foot valve can be used
so that the liquid does not drain out after pump stoppage. Auto-priming pot can be used. The
auto priming pot is installed above the pump centerline. One line from top of pot goes to the
suction tank; another line from bottom of priming pot goes to the pump suction. Initially some
liquid is filled in priming pot, pump is started. The liquid from pot gets sucked into pump, vacuum
gets created in the pot. This vacuum causes the liquid to get lifted from suction tank into auto-
priming pot. This common method of priming is used for pumping water from underground tanks.

Motor details

The details of the electric motors should be clearly mentioned in the data sheet if the scope of
supply of motor is in client’s scope. The frame size of motors should be also very clearly
mentioned. If the frame size is not mentioned in the data sheet ensure that the pump
manufacturer knows about it. Also ensure that latest frame details are available for motor. Some
times the motor manufacturer changes the frame details and the latest information may not be
available with client as well vendor. In that case wrong frame size will be supplied & when the
pump is delivered the motor may not fit exactly & modifications may need to be carried out at site.
The RPM of motor must also be mentioned clearly in the data sheet whether 1440 or 2900 RPM.

Positive displacement pump

Under this category two types of pump come: Rotary pumps

Reciprocating pumps.
Rotary pumps

There are many types of rotary pumps e.g. Cam, Screw, vane, Lobe, Shuttleblock etc. The
majority of this type is capable of handling a clean solution essentially free of solids. These
pumps handle materials of a wide range of viscosity ( upto 500.000 ssu ) and can develop quite
high pressures ( over 1000 psi ). In addition the unit s can handle some vapour or dissolved
gases with the liquid being pumped. The capacity is generally low per unit and at times are used
for metering.

These pumps are low in cost, require small space and are self-priming.
Some can be operated in either direction, have close clearances, require over-pressure relief
protection on discharge and have low volumetric efficiency.

The performance characteristics are:

 Performance characteristics
 Flow is proportional to speed and almost independent of pressure differential.
 Internal slip reduces efficiency.
 Entrained gases reduce liquid capacity and cause pulsations.
 BHP varies directly with pressure and speed.
 For speed & pressure constant, BHP varies directly with viscosity.

 Selection
Suction and discharge heads are determined the same way as that for centrifugal
pumps. Total head and capacity are used in selecting the proper rotary pump from a
manufacturer’s data or characteristic curves since viscosity is quite important in the
selection of these pumps. It is sometimes better to select a large pump running at low
speeds than a smaller pump at high speed when dealing with viscous materials. As a
general guide, speed is reduced 25-30 percent below rating for each ten fold
increase in viscosity above 100 ssu. Mechanical efficiency of pump is decreased 10
percent for each ten fold increase in viscosity above 1000 ssu.

Reciprocating pumps

In this category the pumps are: Plunger pump

Diaphragm type metering pump

Plunger pump

These can be simplex, duplex or multiplex types. These are gland pack pumps. The liquid being
pumped is coming in contact with the plunger. There is always some leakage through gland.
These pumps cannot be used for dangerous, poisonous liquids. These are used generally for
slow feeding of chemicals.

Diaphragm type metering pump

This type of pump is the most commonly used pump in the category of reciprocating pumps. The
advantage is that zero leakage is possible with these pumps. The pumping liquid is separated
from the plunger via teflon diaphragm. The plunger is operating in hydraulic oil chamber. On the
suction stroke plunger moves backward, thereby creating vacuum in hydraulic chamber. This
pulls the diaphragm back, thereby sucking the liquid. In forward stroke the piston moves forward,
thereby pushing the diaphragm in the forward position. This causes pressure to develop on
discharge side, thereby opening discharge non return valve and forcing liquid out. The action is
repeated. Since there is no leakage these pumps can be used for pumping toxic chemicals as
well. Diaphragm type metering pumps can be simplex, duplex, triplex, multiplex etc. The liquid
coming in contact with the plunger is hydraulic oil ad not the process fluid. There is some leakage
of hydraulic oil to atmosphere, which is acceptable.

For handling extremely corrosive chemicals double diaphragm type teflon coated metering pumps
are used. As the name suggests there are two diaphragms. One diaphragm separates the
hydraulic oil from intermediate fluid. Another diaphragm separates intermittent fluid from process
fluid. The fluid used in the intermediate chamber should be compatible with the pumping fluid.
These pumps come in metallic as well as non metallic MOC used for very corrosive liquids. The
metallic MOCs are CS, SS-304, SS-316, HastC etc. The non metallic MOCs are PP, PTFE,
CI/PP, CI/PTFE etc.

ANNEXURE VII
Prepared by VR

Relief systems
Items covered in this Annexure are :

1 Relief devices on process side

2 Relief devices on limpet/jacket
3 Relief devices on hold tanks
4 Types of R D & sizing
5 Types of SRV & sizing

A) Relief devices on process side of reactors:

As per factories act under rule 65 relief devices are to be provided on all equipments Which are operating
under pressure above atmospheric pressure .GCL has applied stricter criterion by which all vessels having
a valve on vent line even if vessel is not opreating under pressure are considred as pressure vessels & are
to be provided with relief devices . Type and number of relief devices depend on operating conditions .

1. Reactions under section 73 H + Pressure + Runaway possible :

Section 73 H of factories act refers to special safety precautions for certain highly hazardous chemical
process as listed under :.
a. Nitro or Amino processes
b. Halogenation reactions
c. Aromatization & Isomerisation process .

For the processes covered under section 73 H factory act specifies that R.D is to be provided in addition to
usual spring loaded SRV.

Runaway reaction : In a runaway reaction heat generated due to exothermicity is not removed at the rate of
generation resulting in increase in temp. & further acceleration of reaction.The exotherm may be in main
intended reaction or side reaction .Whether a reaction is runaway or not is determined by lab. either based
on expt. /reference to literature or by generating data from outside lab. such as CISRA.

It is proposed to install on separate nozzles

i) R.D ( R D 1 )& SRV in series
ii) Two R.D in series ( R D 2 & R D 3)

Pressure rating of R D 1 = Max. operating pressure + 10 % of Max. operating pressure + burst tolerance
of disk (which is equal to 5% of burst pressure ).
Pressure rating of R D 2 & R D 3 : MWAP of vessel - burst tolerance of disk
Pressure rating of SRV = Max. operating pressure + 10 % of Max. operating pressure ).

Note : 1) RD below SRV should of non fragmenting type.

2) Reactions under 73 H + Pressure + no Runaway :
R.D & SRV in series is to be installed .
 Pressure rating of R D & SRV : Max. operating pressure + 10 % of Max. operating
pressure + burst tolerance of disk (which is equal to 5% of burst pressure ).
3) Reactions under 73 H at atm . pressure + Runaway present or absent: It is proposed to
install one safety device provided
a) that vessel is connected to scrubber which is provided with R D or SRV . The
scrubber’s SRV or R D is effectively vessel’s relief device during operation.
e.g acid chloride prep.
b) with open vent e.g Cumene Nitration

Note :Rule 65 which refers safety measures for vessels operated under pressure above atmospheric
specifies that “ suitable safety valve or other effective device (OED ) “ is to be provided on pressure plant
or vessel .We can provide R.D in place of safety valve since R. D qualifies as OED.

In all above cases vapour load is calculated based on combined heat generated due to
i) Exothermicity : Maximum heat liberated is calculated either from lab. data taking into account possible
deviations in operating condiions or from studies at CISRA .In absence of data maximum heat liberated is
taken = 10 times the average rate .
ii) Heat input to limpet coil .This is calculated based on steam pr., & pressure drop in line.
iii) Heat absorbed by reaction mass due to external fire.

Note :

2. Other reactions :
For reactors carrying out reactions other than those specified under section 73 H and also for reactors used
for distillation ,mixing , crystallisation e.t.c it is proposed to install only R.D .
-
Burst pressure of disk = MAWP of reactor – burst tolerance – mfg. Tolerance (whereever applicable )

Vapour load is calculated on the basis of combined heat input to reactor due to
i) Heat of reaction :
ii) Heat input to limpet coil .This is calculated based on steam pr., & pressure drop in line.
iii) Heat absorbed by reaction mass due to external fire.

Relief devices on limpet/jacket of reactors

Reactors other than GLV:
In case valves on CW lines are closed ,any raise in temp. of process medium will result in increase in CW
temp. & hence increase in CW pressure . SRV/TRV is to be installed on CW return line. Since only
pressure relief (no specific capacity ) is required ¾ * 1 “ SRV is adequate.
SRV set pressure = MAWP of limpet of reactor .

For GLV in case stem pr. < MAWP of jacket , valve is required to relieve hydraulic pressure only & ¾*1
SRV is adequate .
Pressure rating = MAWP of jacket

In case steam pressure exceeds MAWP of jacket

,SRV is to be designed for flow rate of steam calculated as per line size .SRV is to be provided with
lifting lever in case of IBR approved lines .

Relief devices for hold tanks :

A. Tanks containing pressurized liquid :
Safety devices to be installed are:
1. R D
2. SRV
3. R D @ the outlet of SRV
R D is installed before & after SRV to protect SRV from corrosion , solid particles e.t.c
The pressure rating of R D = MAWP of tank less burst pressure margin .
The pressure rating of SRV = MAWP of tank
R D is to be of non fragmenting type .
R D & SRV are to be sized based on external fire conditions.

Heat absorbed due to external fire

Q = 21000*( Aw) ^0.82* F Ref: Ludwig Vol. 1 page 260
Where Q is heat absorbed in Btu/hr
Aw is total wetted area in sq. Feet
F is the environmental factor .
F= 1 for uninsulated tank
F = 0.3 for tank with 1 “ thk. insulation
F = 0.15 for tank with 2 “ thk. insulation
F = 0.075 for tank with 4 “ thk. insulation

Venting rate V = Q /Heat of vapourisation

Tank is not be insulated unless required for purpose other than relief device sizing.

B. Tanks operating at atmospheric pressure/under vacuum

R. D to be sized for external fire conditions . Tank is not insulated unless required for purpose other than
relief device sizing. R D to be suitable for vacuuum conditions.

C. Relief devices on limpet/jacket :

For limpet/jacket provided with CW/CHW (no steam) 1/2*3/4 “ S RV is to be installed .For limpet/jacket
provided with no steam .relief valve is not required if max. Pressure of steam is less than MAWP of
limpet/jacket .

If steam pr. is > limpet/jacket ,SRV is to be sized for maximum flow rate possible which is calculated
based on line size ,length & number of fittings.

Types of R D :

1. Conventional Prebulged tension loaded solid metal disks .

a. Max. operating of vessel /tank = 70 % of rated pressure of disk for non pulsating service. For
pulsating service max.. operating of vessel /tank = 50 % of rated pressure of disk .Disk is installed
with dome facing away from process side .
b. Disk can be used for liquid service also.
c. Least expensive of all disks.
d. Drawbacks : i)Disk fragments on bursting & hence cannot be installed below SRV
ii) Vacuum support is required if vacuum exceeds 600 mm Hg. Because of vacuum
support if disk is installed upside down it will burst at pressure higher than rated
pressure.

2. Prescored tension loaded disks :

a. his disk is a solid metal disk prescored in a specific pattern to weaken the disk.
b. Max.. operating of vessel /tank = 80 to 85 % of rated pressure of disk
c. They avoid disadvantage of conventional tension loaded disk viz.fragmentation & need for vacuuum
support. Since there is no vacuum support ,disk with PTFE on process side is available. However they
are costlier than conventional solid metal disks.
d. Disk installed incorrectly will burst at rated pressure or lower.

3. Composite disk :
a. This is prebulged disk & has a slotted top metal section wherein burst pressure is controlled by size &
location of slots & perforations.The disk is isolated from process by PTFE membrane . Since burst
pressure is not controlled by thickness , disk is available at lower burst pressure than conventional
disks .
b. Max.. operating of vessel /tank = 80 % of rated pressure of disk.
c. Drawbacks : Similar to conventional metal disks viz.need for vacuum support ; fragmentation & bursts
at higher pressure if installed upside down.

4. Reverse buckling disks

Installed with dome facing towards process side & hence disk is in compression. The bursting action of
disks is due to either knife blades on downstream or prescoring on disks.
Advantage over tension loaded disks :
b. Disk does not fragment & hence can be installed below SRV .
a. Vacuum support is not required .Disk is available with PTFE seal facing process & hence can be used
for corrosive applications.
b. Disk can be used up to 90 % of operating pressure.
c. Manufacturing range is zero.
d. Disadvantage :Cost is higher than conventional tension loaded disks.Except RLS other types are not
suitable for liquid service .

5. Non metallic disks:

Rupture disks in graphite are available which are useful in highly corrosive services .The drawbacks of
graphite disks are :
a. Cannot be installed below SRV because of fragmentation ;
b. lower shelf life due to resin bonding .

Effect of temperature :
The burst pressure of disk is function of temperature .The temp. :pressure relation for RD is different than
that for parent metal .In order to minimise effect of temp. it is recommended to install R D at least one
metre away from vessel nozzle so that disk faces much lower temp. than the vessel. In this case operating
temp. of disk can be specified as 50 ° C.

Sizing of R D :
For vapour or gas service area of R D is calculated from formula :

W = C*K*A*P*(M/T)^0.5
Ref: Loss prevention in process industries by F Lees Page 12/67

Where W is venting rate in Kg/hr

A is area of disk in sq. mm
P is absolute pressure in bars
M is molecular weight of gas or vapour
T is temperature in ° K
K is Coeff. of discharge = 0.6
C is constant = 2.7

From A , diameter of disk is calculated .

Types of SRV:
a. Conventional relief valve :
Conventional relief valves operate satisfactorily when builtup back pressure is less than 10 % of set
pressure .Changes in back pressure affects both set pressure & capacity.

b. Balanced relief valve:

Balanced relief valves have bellows & they operate satisfactorily under varying back pressure .In addition
bellows protect SRV from corrosion from discharge side .

Definition of terms :
Accumulation : Pressure increase over MAWP during discharge expressed as %.Accumulation =
10 % of set pressure or 3 psi whichever is higher .
Blowdown:Blowdown is differance between set pressure & reseating pressure .This is adjustable
from 7 % to 12 % of set pressure.

Sizing of SRV :
For gases & vapours at constant backpressure orifice area is calculated as under :

A = W*(Z*T/M)^0.5
---------------------
C*K*P*Kb

Ref: LUDWIG Page 254

Where W is flow rate in lb/hr

A is area of orifice in sq. inch
P is inlet pressure in psia
M is molecular weight of gas or vapour
T is temperature in ° R
Kb = 1 for backpressure below 55 % of absolute relieving pressure
K is orifice Coeff. = 0.97 for nozzle type SRV
C is constant depending on ratio of specific heats.

SRV are available in with specific orifice area which are designated by by code D to H ; J to N ; P to Q .
**********
 ANNEXURE VIII

PIPELINE SIZING

(Prepared by KKV)

Pipe Sizing: Pipe size is generally chosen as the minimum size able to carry the desired maximum flow
rate required at a pressure loss equal to the pressure available. In addition to friction and velocity-head
losses in the piping and fittings, the pressure loss owing to flow meters and valves in the line, must be
included as a part of the pressure loss in the piping system. Appropriate Safety margin must be considered
for off-design operating conditions that may involve changes in the available pressure differential between
the terminals of the piping system and changes in the required flow rate to ensure that the line size is
adequate for all anticipated operating conditions. Secondly piping economics shall be considered as,
increase in the pipe size reduces the pressure loss, but increase piping cost.

The important equations for calculation of pressure drop are determination of the Reynolds number (Re)
and the head loss ( hf)

Re = d * v *  /  and hf = 4( f * L / d ) ( v2 / 2gc)

Where d = pipe inside diameter in meter, v = fluid velocity in m/s,  = fluid density in kg/m 3,  = fluid
dynamic viscosity in Pa.s, f = Fanning friction factor (non-dimensional), L = length of line in meter, gc =
acceleration of gravity in m/s2.

For laminar flow (Re < 2000), generally found only in circuits handling heavy oils or other viscous fluids, f
= 16 Re. For turbulent flow, the friction factor is dependent on the relative roughness for the pipe and on
the Reynold number. In such case it is preferable to use charts of friction factor versus Reynolds number
and proper equation must be used with the proper friction factor. An approximation of the Fanning friction
factor for turbulent flow in smooth pipes, reasonably good up to Re = 150000, is given by f = (0.079) / (4 *
Re1/2).

The flow resistance of pipe fittings (elbows, tees, etc) and valves is expressed in terms of either an
equivalent length of straight pipe or velocity head loss (head loss = K v2 / 2gc). Most handbooks and
manufacturers’ publications dealing with fluid flow incorporate either tables of equivalent length for
fittings and valves or K values for velocity head loss.

Suggested Fluid Velocities in Pipe and Tubing for liquids, gases and vapors at low/moderate pressure to 50
psig and 50°C to 100°C are given in table on next page :
Fluid Velocity Pipe Material Fluid Velocity Pipe Material
fpm fpm
Acetylene 4000 Steel NaOH upto 30% 360 Steel
Air 0 – 30 psig 4000 Steel NaOH 30 – 50% 300 Steel
Ammonia (Gas) 6000 Steel NaOH 50 - 73% 240 Steel
Ammonia (Liquid) 360 Steel NaCl Solution 300 Steel
Benzene 360 Steel NaCl slurry 430 Steel
Bromine (Liquid) 240 Glass Perchlorethylene 360 Steel
Bromine (Gas) 2000 Glass Steam Steel
Calcium Chloride 240 Steel 0-30 psi saturated 5000 Steel
CCl4 360 Steel 30-150 psi satu. Or 8000 Steel
superheated
Chlorine (Liquid) 300 Steel,sch.80 Above 150 psi 12500 Steel
superheated
Chlorine (Gas) 3500 Steel Steam(Short Line) 15000 Steel
Chloroform (Liq) 360 Cu & Steel H2SO4 88-93% 240 Lead,SS316
Chloroform (Gas) 2000 Cu & Steel H2SO4 93-100% 240 Steel,sch80
Ethylene Gas 6000 Steel SO2 4000 Steel
Ethylene Dibromide 240 Glass Styrene 360 Steel
Ethylene Dichloride 360 Steel Trichloroethylene 360 Steel
Ethylene Glycol 360 Steel Steel
Hydrogen 4000 Steel Steel
HCl (Liquid) 300 Rubberlined Steel
HCl (Gas) 4000 RL, Saran, Steel
Haveg
Oxygen (ambient temp) 1800 Steel 300 psig
max.
Oxygen (Low temp) 4000 SS304l Steel