Estimation of Over-All Heat Transfer

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ...

Page 1 of 10

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline Heat Loss/Gain Introduction Overall Heat Transfer Coefficient Inside Film Resistance Pipe Wall Resistance Pipe Coating/Insulation Resistances Surroundings Resistance Diameter Basis for Overall Coefficient Controlling Resistance References

INTRODUCTION The average rate at which heat is lost from or gained by fluid flowing through a section of pipe, due to steady state heat transfer between the fluid and the pipe surroundings, is generally calculated using an expression of the form,

where Q = average heat transfer rate (BTU/hr or kJ/hr) Uo = overall heat transfer coefficient based on the area Ao (BTU/ft2-hrF or W/m2-K) Ao = area of heat transfer surface (ft2 or m2) Ts = average temperature of the pipe surroundings (F or C) Tf = average temperature of the flowing fluid in the pipe (F or C) For a particular length of pipe L, the heat transfer area is given by,

where Dot = overall outside diameter of the pipe, including all layers of insulation and/or other coatings (ft or m) L = length of pipe section (ft or m) As it is normally used, Equation (1) can thus be written as,

Note that when Tf > Ts, heat will be lost from the fluid to the surroundings and Equation 3 will compute a positive heat transfer rate. The parameter Q can thus be viewed as the rate of heat loss, and a negative value would indicate that the fluid is gaining heat from the surroundings. Note also that if L is relatively small, Equation 3 essentially defines the local heat transfer rate. If Equation 3 is solved, along with appropriate relations for the other energy changes in the pipeline, for successive L lengths of the pipe, the procedure will generate the flowing temperature profile in the pipeline. back to top OVERALL HEAT TRANSFER COEFFICIENT It is clear from Equation 3 that the accuracy with which the overall heat transfer coefficient Uo can be determined will directly dictate the accuracy of the heat transfer calculations. In fact, Uo is a complex function of many factors, including the properties and flow rates of the fluid(s) in the pipeline, the nature of the surroundings, the thickness and properties of the pipe coatings and insulation (if any). A completely rigorous calculation of Uo, taking into account all of the effects, is seldom possible because either key data are unknown or because reliable models/correlations are not available. In practice however, a reasonable estimate of Uo can generally be obtained from the relationship,

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 2 of 10

Each term on the right hand side of Equation 4 represents the heat transfer resistance due to a particular part of the system, and can be dealt with on an individual basis. back to top INSIDE FILM RESISTANCE The first term in the right hand side of Equation 4, Rfilm, is the resistance to heat transfer that results from the layer of fluid adjacent to the inside surface of the pipe wall. It is calculated using the expression,

where hi = inside convective heat transfer coefficient (BTU/ft2-hr-F or W/m2-K) Di = inside diameter of the pipe (ft or m) The coefficient hi relates to the transfer of heat between the flowing fluid and the inside wall of the pipe. The calculation of this parameter depends on the nature and properties of the fluid(s), the flow rate, and the diameter of the pipe. Most heat transfer textbooks contain a variety of correlations that can be used for single phase fluids in laminar or turbulent flow, condensing vapours, flashing fluids, etc. A particularly useful relationship for turbulent flow of single phase fluids can be written as,

where kf = thermal conductivity of the flowing liquid (BTU/ft-hr-F or W/m-K) Vf = average velocity of the fluid (ft/s or m/s) = average density of the fluid (lb/ft3 or kg/m3) μf = average viscosity of the fluid in appropriate units (for Re, lb/ft-s or Pa.s; for Pr, lb/ft-hr or Pa.s) Cpf = specific heat capacity of the fluid (BTU/lb-F or J/kg-K) Re = the Reynolds number Prf = the Prandtl number
f

If n = 0.333, Equation 6 is generally referred to as the Colburn equation. Dittus and Boelter (1930) recommended that one should set n = 0.3 if the fluid is being heated, and n = 0.4 if the fluid is being cooled. It should be noted however that these recommendations were based on data from heat exchangers, where heat transfer rates tends to be much higher than those encountered in most pipelines. All fluid properties are assumed to be evaluated at the average fluid temperature. Equation 6 is generally valid for turbulent single phase flow, where Re > 104 0.6 < Prf < 120

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 3 of 10

If the flow is laminar (i.e. Re < 2100), one should estimate hi using the Hausen (1943) equation,

where Lo = distance from the pipe inlet to the point of interest (ft or m) In most pipelines of interest, Di/Lo 0 and Equation 9 simplifies to

For the transition region (2100 < Re < 104), several arbitrary and relatively complex procedures have been proposed for calculating hi. As noted by Parker et al (1969) however, the expected heat transfer behaviour in this region is always uncertain because of the unstable nature of the flow. This, of course, is especially true for pipeline systems where multiphase flow or other fluid property complications may exist. It is then not even clear, for example, how to compute the Reynolds number. It should also be noted that the procedure referred to above follows a totally conjectured path through the transition region in any case. As a conservative practice (assuming that higher heat losses represent a worst case), it is suggested that one use the following procedure in the transition region. (a) for 2100 Re 3000, set Re = 3000 (b) for Re > 3000, use Equation (6) Alternatively, as a somewhat less conservative estimate, one can simply compute hi,1 using Equation 10 and hi,2 using Equation 6 with Re = 104. The estimated value of hi is then computed as a weighted average, using the relation,

Table 1 shows typical ranges of values for the inside film coefficient when the flow is turbulent. In most pipelines involving multiphase flow, Rfilm will in fact be small compared to the other terms in Equation 4 and an approximate value of hi based on Table 1 will often be adequate. Table 1 Typical Values of the Inside Film Coefficient (hi) for Turbulent Flow Fluid Water Gases Oils BTU/ft2-hr-F 300 - 2000 3 - 50 10 - 120 W/m2 - K 1700 - 11350 17 - 285 55 - 680

Tables 2 through 5 give values of thermal conductivity and specific heat capacity for a variety of typical oils and gases. In cases where measured values of these parameters are not known, appropriate values from those tables should be sufficiently accurate for most purposes. Table 2 Typical Thermal Conductivity (kf) Values for Crude oils and Other Hydrocarbon Liquids

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 4 of 10

0 F (-18 C) API Gravity BTU/ft-hr- F .068 .074 .078 .083 .088 .093 .103 .111 W/mK .118 .128 .135 .144 .152 .161 .178 .192 Table 3

200 F (93 C)

10 20 30 40 50 60 80 100

BTU/ft-hr- F .064 .069 .074 .078 .083 .088 .097 .105

W/mK .111 .119 .128 .135 .144 .152 .168 .182

Typical Thermal Conductivity (kf) Values for Hydrocarbon Gases 50 F (10 C) Gas Gravity BTU/ft-hr-F W/m-K BTU/ft-hr-F W/m-K BTU/ft-hr-F W/m-K .016 .028 .018 .031 .023 .039 .014 .013 .012 .011 .024 .022 .021 .019 .016 .015 .014 .013 Table 4 Typical Specific Heat Capacity (Cpf) Values for Crude Oils and Other Hydrocarbon Liquids 0 F (-18 C) API Gravity BTU/lb-F kJ/kg-K BTU/lb-F kJ/kg-K BTU/lb-F kJ/kg-K 0.320 1.340 0.355 1.486 0.400 1.675 0.325 0.330 0.335 1.361 1.382 1.403 0.365 0.370 0.375 1.528 1.549 1.570 0.415 0.420 0.430 1.738 1.758 1.800 100 F (38 C) 200 F (93 C) .028 .026 .024 .022 .021 .019 .018 .017 .036 .033 .031 .029 100 F (38 C) 200 F (93 C)

0.7 0.8 0.9 1.0 1.2

10 30 50 70

Table 5 Typical Specific Heat Capacity (Cpf) Values for Hydrocarbon Gases

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 5 of 10

Gravity 0.7 0.8 0.9 1.0 back to top

0 F (-18 C) 100 F (38 C) 200 F (93 C) BTU/lb-F kJ/kg-K BTU/lb-F kJ/kg-K BTU/lb-F kJ/kg-K 0.47 1.97 0.51 2.14 0.55 2.30 0.44 0.41 0.39 1.84 1.72 1.63 0.48 0.46 0.44 2.01 1.93 1.84 0.53 0.51 0.48 2.22 2.14 2.01

PIPE WALL RESISTANCE The second term on the right hand side of Equation 4, Rpipe, is the resistance to heat transfer due to the pipe wall. It is computed from the Equation,

where Do = outside diameter of the pipe (ft or m) kpipe = thermal conductivity of the pipe material (BTU/ft-hr-F or W/m-K) For carbon steel, kpipe is about 26 to 30 BTU/ft-hr-F (45 to 52 W/m-K). In most cases, Rpipe will be very small compared to the sum of the other resistance terms in Equation 4 and it is not unusual practice to ignore it altogether. It can, however, be significant for some materials (e.g. plastic, concrete), especially for greater wall thicknesses. Typical values of thermal conductivity are given in Table 6 for some common pipe materials. Table 6 Thermal Conductivity (kpipe) of Common Pipe Materials Material Carbon Steel Ductile iron Aluminum Copper Polyethylene H.D. PVC Concrete back to top PIPE COATING/INSULATION RESISTANCES BTU/ft-hr-F 26-30 30 115 220 0.21 0.11 0.5-1.0 W/m-K 45-52 52 199 380 0.36 0.19 0.86-1.73

The third term on the right hand side of Equation 4, Rc, is the sum of the heat transfer resistances due to any coatings, insulation, concrete, etc. that may have been applied to the pipe. Each term in the summation is of the form,

where tc = thickness of the coating (ft or m) Dci = inside diameter of the particular layer of coating (ft or m) kc = thermal conductivity of the coating material (BTU/ft-hr-F or W/m-K)

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 6 of 10

The heat transfer resistance due to coatings can be very significant, and is often the major factor affecting the overall heat transfer coefficient. Table 7 gives typical values of kc for a number of common pipe coating materials. Table 7 Thermal Conductivity (kc) of Typical Pipe Coating Materials Material asphalt concrete (insulating) concrete (weight) glass fibre neoprene PVC PVC foam polystyrene foam polyurethane foam back to top SURROUNDINGS RESISTANCE The final term on the right hand side of Equation 4 is the resistance to heat transfer due to the pipe surroundings. The determination of Rsurr depends on whether the pipe is buried, above ground and exposed to atmosphere, or submerged and exposed to water, as well as the nature of the surroundings themselves. (i) Buried Pipe For buried pipe, Rsurr is given by, BTU/ft-hr-F 0.17 0.06 - 0.35 0.50 - 1.0 0.023 0.145 0.11 0.023 0.017 - 0.020 0.014 - 0.020 W/m-K 0.29 0.10 - 0.60 0.86 - 1.73 0.040 0.25 0.19 0.040 0.029 - 0.035 0.024 - 0.035

where Zb = depth of cover to the center line of the pipe (ft or m) ks = thermal conductivity of the pipe surroundings material (BTU/ft-hr-F or W/m-K) Values for ks for some typical soil types and other surroundings materials are given in Table 8. Table 8 Thermal Conductivity (ks) of Typical Pipe Surroundings Material Peat (dry) Peat (wet) Peat (icy) Sandy soil (dry) Sandy soil (moist) Sandy soil (soaked) Clay soil (dry) Clay soil (moist) Clay soil (wet) Clay soil (frozen) Gravel Gravel (sandy) Limestone Sandstone Ice (0C) Ice (-40C) BTU/ft-hr-F 0.10 0.31 1.09 0.25 - 0.40 0.50 - 0.60 1.10 - 1.40 0.20 - 0.30 0.40 - 0.50 0.60 - 0.90 1.45 0.55 - 0.72 1.45 0.75 0.94 - 1.20 1.27 1.54 W/m-K 0.17 0.54 1.89 0.43 - 0.69 0.87 - 1.04 1.90 - 2.42 0.35 - 0.52 0.69 - 0.87 1.04 - 1.56 2.51 0.95 - 1.25 2.51 1.30 1.63 - 2.08 2.20 2.66

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 7 of 10

Snow (loose) Snow (hard packed) (ii) Above-ground Pipe

0.05 - 0.13 0.40 - 0.70

0.09 - 0.23 0.69 - 1.21

For above-ground pipelines (i.e. exposed to air), Rsurr is computed using a relationship based on the Hilpert (1933) equation as follows:

where ksf = thermal conductivity of the surrounding fluid (BTU/ft-hr-F or W/m-K) Vsf = velocity of the surrounding fluid perpendicular to the pipe axis (ft/s or m/s)
sf

= density of the surrounding fluid (lb/ft3 or kg/m3)

sf = viscosity of the surrounding fluid (lb/ft-sec or Pa.s) C1, n = parameters that depend on the value of Resf

Values of C1 and n for Equation 15 are given in Table 9 for various ranges of the Reynolds number, Resf. (iii) Submerged Pipe The Hilpert equation, and therefore also Equation 15, is only applicable to air, for which the Prandtl number is about 0.7. For other fluids (e.g. seawater) Parker et al (1969) recommend use of an empirically modified form of Equation 15, namely,

where Cpsf = specific heat capacity of the surrounding fluid (BTU/lb-F or kJ/kg-K) C2, n = parameters that depend on the value of Resf Values for C2 and n for Equation 17 are also given in Table 9. Table 10 contains some values for ksf and sf for air and water (i.e. the most usual surroundings fluid) at various temperatures. Table 11 gives some conversion factors that may be useful when estimating the velocity of the surroundings fluid normal to the pipe. Table 9 Parameters for Equations 15 and 17 Resf 1-4 4-40 40-4000 4000-40,000 C1 0.891 0.821 0.615 0.174 C2 0.989 0.911 0.683 0.193 n 0.330 0.385 0.466 0.618

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 8 of 10

40,000-250,000

0.0239 Table 10

0.0266

0.805

Thermal Conductivity and Viscosity of Air and Water for Various Temperatures (ksf or sf) Thermal Conductivity Viscosity BTU/ft-hr-F W/m-K cP or mPa.s Air (-20C) 0.0131 0.0227 0.016 Air (0C) 0.0137 0.0237 0.017 Air (20C) 0.0145 0.0251 0.018 Air (40C) 0.0153 0.0265 0.019 Water (0C) 0.321 0.555 1.80 Water (20C) 0.349 0.604 1.00 Table 11 Some Useful Conversion Factors for Velocity Original Units ft/s knots km/hr mile/hr m/s (iv) Partially Buried Pipe Sometimes a submerged pipeline is known to be partially covered because of silting or sedimentation. Similarly, an aboveground pipeline may become partially buried over a period of time. A rigorous calculation of the heat transfer resistance due to the surroundings is quite impractical in most cases because of both irregular and variable geometry of the system. It would, however, seem reasonable in such cases to use the following procedure: (a) compute (Rsurr)1 using Equation 14 with a burial depth of, say, 0.55 Dot (i.e. pipe is just covered) (b) compute (Rsurr)2 using either Equations 15 or 17, as appropriate. (c) compute Rsurr using the relation, Multiplying Factor to Get (ft/s) (m/s) 1.0 0.3048 1.688 0.5144 0.9113 0.2778 1.466 0.447 3.281 1.0 Fluid

where fe = fraction of outside surface of the pipe assumed to be exposed to the surrounding fluid. The accuracy of this procedure should be satisfactory for most cases. However, where heat transfer considerations are a critical factor in the design, it is strongly recommended that the calculations be performed based on the worst case (i.e either buried or not buried, rather than partially buried). back to top DIAMETER BASIS FOR OVERALL COEFFICIENT Equation 3 is a convenient expression for computing the heat transfer rate, but it does not permit a direct comparison of the effect on the heat transfer coefficient itself of say, adding or removing insulation. This is because the presence or absence of a coating layer also affects the value of the overall outside diameter Dot. Note however, that for a given temperature difference (Tf - Ts) and length of pipe section L, the heat transfer rate is directly proportional to the product of Uo and Dot. Thus, one could define another overall heat transfer coefficient, say, Uop, that is related to Do, the outside diameter of the pipe itself. We could then write Equation 3 as,

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipeline ... Page 9 of 10

If Equation 3 and Equation 20 are solved for Uop, we obtain the relation,

It can be seen that Equation 21 provides a means for adjusting the value of Uo, calculated using the procedures described in this Note, to a constant and fixed diameter basis. Any change that is observed is Uop as a result of changing any design value (i.e. insulation type or thickness, pipe material, flow rate, burial depth) will be directly proportional to the expected change in the overall heat transfer rate. back to top CONTROLLING RESISTANCE To illustrate the concept of a controlling resistance, we will examine the values of the various resistances for a particular case of an oil pipeline having the following specifications: Pipe O.D. = 2.75 inches Wall Thickness = 0.406 inches Thermal Conductivity = 28 BTU/ft-hr-F Oil Gravity = 30API Oil Viscosity = 3.5 cP Thermal Conductivity = 0.076 BTU/ft-hr-F Specific heat capacity = 0.35 BTU/lb-F Flow rate = 25,000 bbl/day The pipe is assumed to be buried 4 feet in a moist clay soil (ks = 0.45 BTU/ft-hr-F). It has an 0.25 inch coating of asphalt (kc = 0.40 BTU/ft-hr-F), covered by a 0.125 inch yellow jacket material (kc = 0.10 BTU/ft-hr-F). In the interests of brevity, the results of the calculations will simply be stated here and it is left to the reader to verify them. (i) Inside film resistance Re = 48,300 Pr = 51.8 n = 0.333 hi = 36.6 BTU/ft2-hr-F Rfilm = 0.0275 ft-hr-F/BTU (ii) Pipe wall resistance Rpipe = 0.0012 ft-hr-F/BTU (iii) Coatings resistance asphalt: Rc = 0.0481 ft-hr-F/BTU yellow jacket: Rc = 0.0935 ft-hr-F/BTU (iv) Surroundings resistance Rsurr = 2.9442 ft-hr-F/BTU Thus, from Equation (3), for Dot = 1.125 ft.

or Uo = 0.285 BTU/ft2-hr-F Note that if all of the resistance terms are ignored except Rsurr. we obtain Uo = 0.302, which differs from the correct answer by only about 6%. Furthermore, if higher heat loss is considered to be a worst case, ignoring the other resistances makes the

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

Estimation of The Overall Heat Transfer Coefficient For The Calculation of Pipelin... Page 10 of 10

estimate more conservative. In any case, it is apparent that, for this particular pipeline, the surroundings resistance is the controlling factor in the heat transfer rate. Clearly, even a large error in calculating (or estimating) the inside film resistance would have negligible effect on the heat transfer calculations. Consider now, however, the case where 2 inches of insulation ( kc = 0.025 BTU/ft-hr-F) is applied over the yellow jacket. This will have no effect on the values of Rfilm, Rpipe or Rc for the asphalt and yellow jacket layers. However, for the insulation, Rc = 5.1902 ft-hr-F/BTU Also, since the overall outside diameter has changed, we must re-calculate Rsurr. This gives, Rsurr = 2.6521 ft-hr-F/BTU and Uo = 0.0856 BTU/ft2-hr-F Clearly, the insulation is now the controlling resistance, although the surroundings resistance is still significant. However, all of the other terms now account for less than 2% of the total heat transfer resistance. To properly compare the effect of adding the insulation, we should examine the values of Uop, relative to the pipe outside diameter. This gives, based on Equation (21), No insulation: Uop = 0.302 BTU/ft2-hr-F With insulation: Uop = 0.117 BTU/ft2-hr-F The effect of the insulation will thus be to reduce the heat loss by about 61%. back to top CONCLUDING REMARKS Overall heat transfer coefficients can be either specified for or calculated automatically, using the procedures described in this Note, by SPT Group's (formerly Neotec's) comprehensive pipeline design and analysis module PIPEFLO. All of the methods described are also contained in the Technical Utility program HTCOEFF. This program provides a particularly convenient means of examining the sensitivity of the overall heat transfer coefficient to various system parameters. back to top REFERENCES Dittus, F.W., and Boelter, L.M.K., Univ. Calif. (Berkeley) Pub. Eng., Vol. 2., p. 443 (1930) Hausen, H., "Darstellung des Warmeuberganges in Rohren durch verallgemeinerte Potenzbeziehungen", Zeitschr. V.D.I. Beihefte Verfarenstechnik, No. 4, p. 91, (1943) Hilpert, R., "Warmeabgabe von geheizen Drahten und Rohren", Forsch. Gebiete IIngenieurw., Vol. 4, p. 220, (1933) Parker, J.D., Boggs, J.H., and Blick, E.F., Introduction to Fluid Mechanics and Heat Transfer, Addison-Wesley Pub. Co. Inc., Reading, Mass. (1969) back to top

mk:@MSITStore:C:\Program%20Files\SPT%20Group\WELLFLO%208.3.0\help\WF...

3/28/2013

You might also like