Você está na página 1de 28

S E C T I O N

O N E

Equipment Design

Fluid Flow
Velocity Head Full Plant Piping Partially Full Horizontal Pipes Equivalent Length Recommended Velocities Two-phase Flow Compressible Flow-Short (Plant) Lines Compressible Flow-Long Pipelines Sonic Velocity Metering Control Valves Safety Relief Valves

............................................................... 3 ......................................................... 4 ................. 5 ....................................................... 6 ............................................ 7 ........................................................... 9 .................12 ............ 18 ............................................................... 21 ....................................................................... 21 ............................................................. 22 ..................................................... 25

Fluid Flow

Velocity Head
Two of the most useful and basic equations are Ah=2g Au2 AP(V)+-+AZ+E 2g where Ah = Head loss in feet of flowing fluid u = Velocity in ft/sec g = 32.2 ft/sec2 P = Pressure in lb/ft2 V = Specific volume in ft3/lb Z = Elevation in feet E = Head loss due to friction in feet of flowing fluid Applications In Equation 1 Ah is called the velocity head. This expression has a wide range of utility not appreciated by many. It is used as is for 1. Sizing the holes in a sparger 2. Calculating leakage through a small hole 3. Sizing a restriction orifice 4. Calculating the flow with a pilot tube With a coefficient it is used for 1. Orifice calculations 2. Relating fitting losses, etc. Why a Coefficient? For a sparger consisting of a large pipe having small holes drilled along its length Equation 1 applies directly. This is because the hole diameter and the length of fluid travel passing through the hole are similar dimensions. An orifice, on the other hand, needs a coefficient in Equation 1 because hole diameter is a much larger dimension than length of travel (say 1/8 in for many orifices). Orifices will be discussed under Metering in this chapter.
=0
U2

Sonic velocity (l) For the situations covered here, compressible fluids might reach sonic velocity. When this happens, further decreases in downstream pressure do not produce additional flow. Sonic velocity occurs at an upstream to downstream absolute pressure ratio of about 2 : 1. This is shown by the formula for sonic velocity across a nozzle or orifice. critical pressure ratio = P2/P1= [2/(K+ l)]K/(K-I) when K = 1.4, ratio = 0.528, so P1/P2= 1.89 To determine sonic velocity, use V, = (KgRT? where V, = Sonic velocity, ft/sec K = C,/C,, the ratio of specific heats at constant pressure to constant volume g = 32.2 ft/sec2 R = 17544/mol.wt. T = Absolute temperature, O R PI, P2 = Inlet, outlet pressures, psia Critical flow due to sonic velocity has practically no application to liquids. The speed of sound in liquids is very high. For sonic velocity in piping see the section on Compressible Flow. Bernoulli Equation Still more mileage can be gotten out of Ah = u2/2g when using it with Equation 2, which is the famous Bernoulli equation. The terms are 1. The PV change 2. The kinetic energy change or velocity head 3. The elevation change 4. The friction loss These contribute to the flowing head loss in a pipe. However, there are many situations where by chance, or

(2)

Rules of Thumb for Chemical Engineers

on purpose, u2/2g head is converted to PV or vice versa. We purposely change u2/2g to PV gradually in the following situations: 1. Entering phase separator drums to cut down on turbulence and promote separation 2. Entering vacuum condensers to cut down on pressure drop We build up PV and convert it in a controlled manner to u2/2g in a form of tank blender. These examples are discussed under appropriate sections. Example Given: Methane (M, = 16) Line @ 100 psia and 60F Hole in the line of 1/8 in diameter Hole discharges to atmosphere (15 psia) Assume Z = compressibility = 1.O Find: Flow through the hole in lb/hr

Calculations: Use AH = u2/2g. Flow is sonic, so use AP = 100 - 50 = 50 psi (2 : 1 pressure drop). Hole diameter = 1/8 in = 0.125 in Hole area = 7c x (0.1252)/4 = 0.0123 in2 = 0.0000852 ft2 Density of methane = (161b/76ft3) x (100/76) x [(460 + 76)/(460 + 60)] = 0.285 lb/ft3 (See Approximate Physical Properties in Section 25, Properties, for the rule-of-76.) AH = 50 lb/in2x 144 in2/ft2x ft3/0.285lb = 25,263 ft u2 = 25,263(64.4) = 1,626,900 u = 1275 ft/sec Flow = 1275 ft/sec x 0.0000852 ft2 x 0.285 lb/ft3 x 3600 sec/hr = 111 lb/hr Source Branan, C.R. The Process Engineer k Pocket Handbook, Vol. 1, Gulf Publishing Co., Houston, Texas, p. 1. 1976.

Full Plant Piping


A handy relationship for turbulent flow in full commercial steel pipes is: exchanger tubeside pressure drop calculations), a constant of 23,000 should be used instead of 20,000. The equation applies to: Liquids Compressible fluids at: Non-critical flow AP less than 10% of inlet pressure It was derived from the Fanning equation: APF = (2f u2 L p)/(32.2 D) and the approximate relationship:2 This relationship holds for a Reynolds number range of 2,100 to lo6. For smooth tubes (assumed for heat f = 0.054/Re0

AP~ = w1.8po.2 /20,00od~.~p


where: APF = Frictional pressure loss, pd100 equivalent ft of Pipe W = Flow rate, lb/hr p = Viscosity, cp p = Density, lb/ft3 d = Internal pipe diameter, in.

Fluid Flow

where: u = velocity, ft/sec L = length, ft f = Fanning friction factor = Moodys / 4 D = diameter, ft = Reynolds Number 1. Branan, C. R., Rules of Thumb for Chemical Engineers, Buttenvorth-Heineman, 2002, p. 4. 2. Simpson, L.L., Sizing Piping for Process Plants, Chemical Engineering, June 17, 1968, p. 197.

Partially Full Horizontal Pipes


The equations in the previous section are, of course, intended for use with full pipes. Durand provides a rapid way to estimate whether a horizontal pipe carrying liquid is full. The criteria are If Q/d2.5 2 10.2 the pipe is full. If Q/d2,5< 10.2 do a partially full flow analysis as follows. and find the height of liquid in the Let x = In (Q/d2,5) pipe by: H/D = 0.446 + 0 . 2 7 2 ~ +0 . 0 3 9 7 ~ -~ 0.0153~~ -0.003575~~ Find the equivalent diameter by: Calculations: De/D = -0.01130+3.040(H/D)-3.461 (H/D)2 +4.108 (H/Dy -2.638 (H/Dy [This is an empirical way to avoid getting D, from D, = 4 (cross-sectional flow area/wetted perimeter)] Note that for 1.0 > H/D > 0.5, D,/D > 1.0. My calculations and all references confirm this. D, is substituted for D in subsequent flow analysis. Nomenclature D = pipe diameter, ft D, = equivalent diameter, ft H = height of liquid in the pipe, ft Q = flow rate, gpm d = pipe diameter, in q = flow rate, ft/sec u = velocity, ft/sec Q/d25 = 100/32 = 3.125 Not full since Q/d25 < 10.2 x = ln(3.125) = 1.1394 H/D = 0.779 H = 0.779 (4) = 3.12 in D,/D = 1.227 D, = 1.227 (4) = 4.9 1 in Source Durand, A. A. and M. Marquez-Lucero, Determining Sealing Flow Rates in Horizontal Run Pipes, Chemical Engineering, March 1998, p. 129. Example Given: Horizontal pipe d = 4 in ID Q = 100 gpm Find: Is the pipe full? If not, what is the liquid height? Also, what is the pipes equivalent diameter?

Rules of Thumb for Chemical Engineers

Equivalent length
The following table gives equivalent lengths of pipe for various fittings.
Table 1 Equivalent Length of Valves and Fittings in Feet

Enlargement

Contraction Std. red. Sudden Std. red.

f i .m
.E

L Q

02
-x

E '5
O Q

$$

$8
0s

.n Q

> m > Q

m
al
Y
Y O

O
0)

P O

0ag I

s O

8
0

m m
0)

45" ell

jhorl Long iard soft rad. rad. T. T. ell ell

90" miter bends

Sudden

L L
+ I

Equiv. L in terms of small d

b
al E $ 2 $ E ;E ; e .8 $ E s s 3 5 Lr- Lr- 5 5 hl al c
Q c .-

'5 fn

z
d

ti *

s
1

s
\

s
II

x
II

(3

- - 1'h 2 2% 3 4 6 8 10 12 14 16 18 20 22 24 30 36 42 48 54 60 55 70 80 100 130 200 260 330 400 450 500 550 650 688 750

26 33 40 50 65 100 125 160 190 210 240 280 300 335 370

- 13 17 20 25 32 48 64 80 95 105 120 140 155 170 185 7 14 11 17 30 70 120 170 170 80 145 160 210 225 254 312

- -- - - 1 2 2 2 3 4 6 7 9 10 11 12 14 15 16 21 25 30 35 40 45

- - -

12 23 2 .. 2 3 4 6 7 9 10 11 12 14 15 16 21 25 30 35 40 45

35 45 5 .. 6 7 11 15 18 22 26 29 33 36 40 44 55 66 77 88 99 110

23 34 3 .. 4 5 8 9 12 14 16 18 20 23 25 27 40 47 55 65 70 80

8 E 10 1' 12 14 19 28 37 47 55 62 72 82 90 100 110 140 170 200 220 250 260

23 34 3 .. 4 5 8 9 12 14 16 18 20 23 25 27 40 47 55 65 70 80

- - - 5 7 8 10 12 18 25 31 37 42 47 53 60 65 70 3 4 5 6 8 12 16 20 24 26 30 35 38 42 46 1 1 2 2 3 4 5 7 8 9 10 11 13 14 15

II

n
U

II

II

n
U

II

- 3 3 4 5 6 9 12 15 18 20 24 26 30 32 35
2 3 3 4 5 7 9 12 14 16 18 20 23 25 27

s s
1 1 2 2 3 4 5 6

28 32 38 42 46 52 56 70 84 98 112 126 190

21 24 27 30 33 36 39 51 60 69 81 90 99

20 22 24 28 32 34 36 44 52 64 72 80 92

7
8 9 10 11 12 13

- - -

Sources
1. GPSA Engineering Data Book, Gas Processors Suppliers Association, 10th Ed. 1987. 2. Branan, C. R., The Process Engineer5 Pocket Handbook, Vol. 1, Gulf Publishing Co., p. 6, 1976.

Fluid Flow

Recommended Velocities
Here are various recommended flows, velocities, and pressure drops for various piping services.
Sizing Cooling Water Piping in New Plants Maximum Allowable Flow, Velocity and Pressure Drop
LATERALS GPM 3 4 6 8 10 12 14 16 18 20 24 30 100 200 500 900 1,500 2,400 3,100 4,500 6,000 ft/sec. 4.34 5.05 5.56 5.77 6.10 6.81 7.20 7.91 8.31 AP ft1100 4.47 4.29 3.19 2.48 2.11 2.10 2.10 2.09 1.99 Flow GPM 70 140 380 650 1,100 1,800 2,200 3,300 4,500 6,000 11,000 19,000 MAINS Vel. Wsec. 3.04 3.53 4.22 4.17 4.48 5.11 5.13 5.90 6.23 6.67 7.82 8.67 AP ftI100 2.31 2.22 1.92 1.36 1.19 1.23 1.14 1.16 1.17 1.17 1.19 1.11

Sizing Steam Piping in New Plants Maximum Allowable Flow and Pressure Drop
Laterals Pressure, PSlG Density, #/CF LIP, PSI/lOO Nominal Pipe Size, In. 3 4 6 8 7.5 15 40 76 130 190 260 360 600 0.91 1.0 175 0.41 0.70 30 0.106 0.50 600 0.91 0.70 Mains 175 1.41 0.40 30 0.106 0.30

.... ....

.... ....

.... ....

....

Maximum Lb/Hr x 3.6 7.5 21 42 76 115 155 220 300 1.2 3.2 8.5 18 32 50 70 100 130 170 6.2 12 33 63 108 158 217 300 2.7 5.7 16 32 58 87 117 166 227 0.9 2.5 6.6 14 25 39 54 78 101 132

....

....

Sizing Piping for Miscellaneous Fluids


Dry Gas Wet Gas High Pressure Steam Low Pressure Steam Air Vapor Lines General Light Volatile Liquid Near Bubble Pt. Pump Suction Pump Discharge, Tower Reflux Hot Oil Headers Vacuum Vapor Lines below 50 MM Absolute Pressure 100 Wsec 60 Wsec 150 ft/sec 100 ft/sec 100 ft/sec Max. velocity 0.3 mach 0.5 psi/lOO ft 0.5 ft head total suction line 3-5 psi/lOO ft 1.5 psi/lOO ft Allow max. of 5% absolute pressure for friction loss

70
12 14 16 18 20

...

...

...

... ...

...

Note: ( I ) 600 PSlG steam is at 75015 175 PSlG and 30 PSlG are saturated. (2) On 600PSIG flow ratings, internal pipe sizes for larger nominal diameters were taken as follows: 18116.5: 14112.8: 12l11.6: 1019.75: (3) If other actual l.D. pipe sizes are used, or if local superheat exists on 175 PSlG or 30 PSlG systems, the allowable pressure drop shall be the governing design criterion.

Rules of Thumb for Chemical Engineers

Suggested Fluid Velocities in Pipe and Tubing (Liquids, Gases, and Vapors at Low Pressures to 50ps i g and 50"F-1OO0F) The velocities are suggestive only and are t o be used to approximate line size as a starting point for pressure drop calculations. Fluid Acetylene (Obsewe pressure limitations) Air, 0 to 30 psig Ammonia Liquid Gas Benzene Bromine Liquid Gas Calcium Chloride Carbon Tetrachloride Chlorine (Dry) Liquid Gas Chloroform Liquid Gas Ethylene Gas Ethylene Dibromide Ethylene Dichloride Ethylene Glycol Hydrogen Hydrochloric Acid Liquid Gas Methyl Chloride Liquid Gas Natural Gas Oils, lubricating Oxygen (ambient temp.) (Low temp.) Propylene Glycol Suggested Trial Velocity 4000 fpm 4000 fpm 6 fps 6000 fpm 6 fps 4 fps 2000 fpm 4 fps 6 fps 5 fps 2000-5000 fpm 6 fps 2000 fpm 6000 fpm 4 fps 6 fps 6 fps 4000 fpm 5 fps 4000 fpm Pipe Material Steel Steel Steel Steel Steel Glass Glass Steel Steel Steel, Sch. 80 Steel, Sch. 80 Copper & Steel Copper & Steel Steel Glass Steel Steel Steel Rubber Lined R. L., Saran, Haveg Steel Steel Steel Steel Steel (300 psig Max.) Type 304 SS Steel The final line size should be such as to give an economical balance between pressure drop and reasonable velocity. Fluid Sodium Hydroxide 0-30 Percent 30-50 Percent 50-73 Percent Sodium Chloride Sol'n. No Solids With Solids Suggested Trial Velocity 6 fps 5 fps 4 5 fps (6 Min.15 Max.) 7.5 fps 6 fps
4000-6000 fpm

Pipe Material Steel and Nickel Steel Monel or nickel Steel Steel

Perchlorethylene Steam 0-30 psi Saturated* 30-1 50 psi Saturated or superheated* 150 psi up superheated *Short lines Sulfuric Acid 88-93 Percent 93-1 00 Percent Sulfur Dioxide Styrene Trichlorethylene Vinyl Chloride Vinylidene Chloride Water Average service Boiler feed PumD suction lines Maximum economical (usual) Sea and brackish water, lined pipe Concrete

6000-1 0000 fpm 6500-1 5000 fpm 15,000 fpm (max.) 4 fps 4 fps 4000 fpm 6 fps 6 fps 6 fps 6 fps
3-8 (avg. 6) fps 4-1 2 fps 1-5 fps 7-1 0 fps S. S.-316, Lead Cast Iron & Steel, Sch. 80 Steel Steel Steel Steel Steel

6 fps 4000 fpm 6000 fpm 6 fps 1800 fpm Max. 4000 fpm 5 fps

Steel Steel Steel Steel R. L., concrete, asphalt-line, saranlined, transite

5-12 5-8fps) fps (Min.)

Note: R. L. = Rubber-lined steel.

Fluid Flow

Typical Design Vapor Velocities* (ft./sec.)


Fluid Saturated Vapor 0 to 50 psig Gas or Superheated Vapor 0 to 10 psig 11 to 100 psig 101 to 900 Dsig
~~~~~~ ~

Typical Design* Velocities for Process System Applications


~

5 6

Line Sizes 8-1 2 50-1 25 90-1 90 75-1 65 60-1 50

214

Service Average liquid process Pump suction (except boiling) Pump suction (boiling) Boiler feed water (disch., pressure) Drain lines Liquid to reboiler (no pump) Vapor-liquid mixture out reboiler Vapor to condenser Gravity separator flows

Velocity, ft./sec. 4-6.5 1-5 0.5-3 4-8 1.5-4 2-7 15-30 15-80 0.5-1.5

30-1 15 50-1 40 40-1 15 30-85

60-1 45 110-250 95-225 85-1 65

*Values listed are guides, and final line sizes and flow velocities must be determined by appropriate calculations to suit circumstances. Vacuum lines are not included in the table, but usually tolerate higher velocities. High vacuum conditions require careful pressure drop evaluation.

*To be used as guide, pressure drop and system environment govern final selection of pipe size. For heavy and viscous fluids, velocities should be reduced to about b values shown. Fluids not to contain suspended solid particles.

Usual Allowable Velocities for Duct and Piping Systems*


ServiceIApplication Forced draft ducts Induced-draft flues and breeching Chimneys and stacks Water lines (max.) High pressure steam lines Low pressure steam lines Vacuum steam lines Compressed air lines Refrigerant vapor lines High pressure Low pressure Refrigerant liquid Brine lines Ventilating ducts Register grilles Velocity, ft./min. 2,500-3,500 2,000-3,000 2,000 600 10,000 12,000-1 5,000 25,000 2.000 1,000-3,000 2,000-5,000 200 400 1,200-3,000 500

Suggested Steam Pipe Velocities in Pipe Connecting to Steam Turbines


Service-Steam Inlet to turbine Exhaust, non-condensing Exhaust, condensing Typical range, ft./sec. 100-1 50 175-200 400-500

Sources
1. Branan, C. R., The Process Engineer5 Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976. 2. Ludwig, E. E., Applied Process Design for Chemical and Petrochemical Plants, 2nd Ed., Gulf Publishing

co.
3. Perry, R. H., Chemical Engineer 5 Handbook, 3rd Ed., p. 1642, McGraw-Hill Book Co.

*By permission, Chemical Engineers Handbook, 3rd Ed., p. 7642, McGraw-Hill Book Co., New York, N. Y

Two-phase Flow
Two-phase (liquidhapor) flow is quite complicated and even the long-winded methods do not have high accuracy. You cannot even have complete certainty as to which flow regime exists for a given situation. Volume 2 of Ludwigs design books and the GPSA Data Book2 give methods for analyzing two-phase behavior. For our purposes, a rough estimate for general twophase situations can be achieved with the Lockhart and Martinelli3 correlation. Perrys4has a writeup on this correlation. To apply the method, each phases pressure drop is calculated as though it alone was in the line. Then the following parameter is calculated:

where: APL and APG are the phase pressure drops The X factor is then related to either YL or YG. Whichever one is chosen is multiplied by its companion pressure drop to obtain the total pressure drop. The following equation5 is based on points taken from the YLand YG curves in Perrys4 for both phases in turbulent flow (the most common case):

10

Rules of Thumb for Chemical Engineers

d
U

Flowrate, Ib/h

Sizing Lines for Flashing Steam-Condensate

The X range for Lockhart and Martinelli curves is 0.01 to 100. For fog or spray type flow, Ludwig' cites Baker's6 suggestion of multiplying Lockhart and Martinelli by two. For the frequent case of flashing steam-condensate lines, Ruskan7 supplies the handy graph shown above. This chart provides a rapid estimate of the pressure drop of flashing condensate, along with the fluid velocities. Example: If 1,000lb/hr of saturated 600-psig condensate is flashed to 200 psig, what size line will give a pressure drop of l.Opsi/lOOft or less? Enter at 6OOpsig below insert on the right, and read down to a 2OOpsig end pressure. Read left to intersection with 1,000lb/hr flowrate, then up verti-

cally to select a 1y2in for a 0.28psi/lOOft pressure drop. Note that the velocity given by this lines up if 16.5 ft/s are used; on the insert at the right read up from 6OOpsig to 2OOpsig to find the velocity correction factor 0.4 1, so that the corrected velocity is 6.8 ft/s. Lockhart and Martinelli Example Given: Saturated 600 psig condensate flashed to 200 psig 1'/* in line, sch. 80 (ID = 1.500 in) Flow = 1000 lb/hr

Fluid Flow

11

Steam Sat. 615 psia T, O F V, ft3/lb H, btu/lb Sat. 215 psia T V H

Condensate

489 489 0.7504 0.0202 1203.0 474.7

Vapor .5p(0.468)] APF = (135)l (0.015)02/[20,000(1 = 0.045 psi/100 ft Cameron = 0.05 Liquid 1.5) (55.5)] APF = (865)l(0.14~*/[20,000( = 0.017psi/100ft Cameron = 0.02 Crane = 0.01 Total Pressure Drop

P,CP
Find:

388 2.135 1199.3 0.015

388 0.018 361.9 0.14

X = [APL/APG]o5 = [0.017/0.045]05= 0.615 +12.5(0.615)-068 +0.65 = 29 YL = 4.6(0.615)-17 Total AP = 29(0.017) = 0.49psi/lOOft Ruskan7 = 0.28 Sources 1. Ludwig, E. E., Applied Process Design For Chemical and Petrochemical Plants, Vol. 1, Gulf Publishing Co. 2nd Edition., 1977. 2. GPSA Data Book, Vol. 11, Gas Processors Suppliers Association, 10th Ed., 1987. 3. Lockhart, R. W., and Martinelli, R. C., Proposed Correlation of Data for Isothermal Two-Phase, TwoComponent Flow in Pipes, Chemical Engineering Progress, 45:3948, 1949. 4. Perry, R. H., and Green, D., Perry5 Chemical Engineering Handbook, 6th Ed., McGraw-Hill Book Co., 1984. 5. Branan, C. R., The Process Engineer5 Pocket Handbook, Vol. 2, Gulf Publishing Co., 1983. 6. Baker, O., Multiphase Flow in Pipe Lines, Oil and Gas Journal, November 10, 1958, p. 156. 7. Ruskan, R. P., Sizing Lines For Flashing SteamCondensate, Chemical Engineering, November 24, 1975, p. 88. 8 Cameron, Hydraulic Data, Ingersoll-Rand Co., 17th Ed. 9 Flow of Fluids, Technical Paper No. 410, Crane CO., 1981.

The flash amounts of steam and condensate, lb/hr Individual pressure drops if alone in the line, psi/lOO ft Total pressure drop, psi/lOO ft

Calculations: Flash Let X = lb/hr vapor Y = lb/hr liquid X + Y = 1000 1199.3X+361.9Y = 474.7(1000) Solving: X = 135 lb/hr Y = 865 lb/hr Individual pressure drops d4,p) (See Full Plant Piping APF = W. p0.2/(20,000 in Section 1, Fluid Flow) where APF = psi/lOO ft W = lb/hr P=C P d = in p = lb/ft3

12

Rules of Thumb for Chemical Engineers

Compressible Flow-Short

(Plant) lines
2. Determine fL/D. 3. Obtain Z. Figures 5, 6, and 7 are provided for convenience. 4. Calculate P2/P1. 2. Same. 3. Same.

For compressible fluid flow in plant piping, one can use Maks Isothermal flow chart (Figure 1). Maks chart was provided originally for relief valve manifold design and adopted by APL2 The relief valve manifold design method, and its derivation, is discussed in Section 20, Safety. Maks methods can be applied to other common plant compressible flow situations. Since Maks Isothermal flow chart is intended for relief manifold design, it supports calculations starting with P2, the outlet pressure, that is atmospheric at the flare tip, and back-calculates each laterals inlet pressure, P1.These inlet pressures are the individual relief valves back pressures. The chart parameter is M2,the Mach number at the pipe outlet. Having M2 is very useful in monitoring proximity to sonic velocity, a common problem in compressible flow. For individual plant lines the following cases are easily solved with Figure 1 and the tabulated steps. Given: P2 and P1 Find: W Steps: 1. Get f from GPSA graph (Figure 4 ) . P2 and W P1 1. Same.
Based on outlet pressure

4.Calculate M2.
See Equation 3. If M2 > 1 flow is choked, so set M2 at 1 and determine a reduced W. 5. Get P2/P1 from Figure 1. Read at the reset value of M2 = 1 if applicable. 6. Calculate P1. Note: This case (given P2 and W) is the same as an individual lateral in relief manifold design.

5 . Get M2 from

Figure 1. If below the critical flow line, use M2 = 1. 6. Calculate W. See Equation 3.

LIVE GRAPH Figure 1. Isothermal flow chart based on MZ.


Click here to view

a c. P
fL/D

LIVE GRAPH Figure 2. Flow chart based on MI.


Click here to view

14

Rules of Thumb for Chemical Engineers

MI Graph

LIVE GRAPH
Click here to view

Critical pressure at the pipe outlet in psia


= (W/408d2)(ZT/Mw)0'5 Pcfit

(5)

For comparison the author has generated an Excel@ plot (Figure 3) using the data from Figure 2. This is for those readers who work with this popular spreadsheet.
Example
Figure 3. Excel@version of M I chart.

Given (see sketch following):

The Mak Isothermal flow chart is such a useful tool that the author has used it for cases where PI is known instead of P2 with a trial and error approach. The author has now generated a graph (Figure 2) based upon MI using Equation 2. The Isothermal flow chart (Figure 1) based on M2 uses Equation 4. Figure 2 facilitates the following case. Given: Find: Steps: PI and W P2 1. Get f from GPSA graph (Figure 4). 2. Determine fL/D. 3. Obtain Z. Figures 5, 6, and 7 are provided for convenience. 4. Calculate M1. See Equation 1. 5. Get P2/P1 from Figure 2. The critical curve indicates where MI = P2/P1.When this happens M2 = 1 since M2 = M1(P1/P2). The design pipe diameter might have to be changed to provide a possible set of conditions. 6. Calculate P2.

Calculate how much gas will flow to the vessel through the 1 in line with the normally closed hand valve fully opened. Use the psv full open pressure of 136 psia as the vessel pressure. The equivalent length of 200 ft includes the fully opened hand valve. The 1 in pipe's inside diameter is 1.049 in. Assume Z = 1.0.
Calculations:

Note that if the AP was across a restriction orifice, sonic velocity would occur since the AP is greater than 2 : 1 (315/136 = 2.31). However, the AP is along a length of pipe, so we will use Mak's method. For commercial steel pipe: f = 0.023 id = 1.049 in = 0.0874 ft P2/Pl = 136/135 = 0.43 fL/D = 0.023(200)/0.0874 = 52.6 M2 = 0.28 (from Figure 1) Note that if the flow were critical, M2 would be 1. M2 = 1.702 x 10-5[W/(P2D2)][ZT/M~]o'5 P2 = 136psia D2 = 0.08742 = 0.00764 T = 460 +60 = 520 O R Mw=16 Z = 1.O (given) 0.28 = 1.702 x 10-5[W/(136 x 0.00764)][1.0(520)/16]o~5

Some calculations require knowing the critical pressure at which sonic velocity occurs. This is calculated with Equation 5 . The applicable equations are Based on M1 MI = 1.702 x 10-5[W/(PlD2)](ZT/M,)0'5 (1)

Based on M2 M2 = 1.702 x 10-5[W/(P2D2)](ZT/Mw)0'5 (3)

W = 0.28 x 1.039 x 105/(5.70 x 1.702) = 30001b/hr

Fluid Flow

15

Pipe Diameter, in Feet - D

Pipe Diameter, in Inches - d LIVE GRAPH


Click here to view

Figure 4. Friction factor chart.

16

Rules of Thumb for Chemical Engineers

LIVE GRAPH
Click here to view

Figure 5. Z factor for natural gas.

Figure 6. Z factor at low reduced pressures.


LIVE GRAPH
Click here to view

Figure 7. Z factor for near atmospheric pressure.


LIVE GRAPH
Click here to view

a c. P

18

Rules of Thumb for Chemical Engineers

Example sketch

M, = Gas molecular weight P1,P2= Inlet and outlet line pressures, psia Pcrit= Critical pressure for sonic velocity to occur, psia T = Absolute temperature, O R W = Gas flow rate, lb/hr Z = Gas compressibility factor
Sources

NC

X
1in
200 equiv. ft

60 O F 100 psig

L 4
1. Mak, Henry Y., New Method Speeds Pressure-Relief Manifold Design, Oil and Gas Journal, Nov. 20, 1978, p. 166. 2. API Recommended Practice 520, Sizing, Selection, and Installation of Pressure Relieving Devices I Refineries, 1993. 3. Flow of Fluids through Valves, Fittings, and Pipe, Crane Co. Technical paper 410, 1981. 4. Crocker, Sabin, Piping Handbook, McGraw-Hill, Inc., 1945. 5. Standing, M.B. and D. L. Katz, Trans. AIME, 146, 159 (1942).

C,
300 psig = 315 psia 60 O F

Nomenclature D = Pipe diameter, ft d = Pipe diameter, in f = Moody friction factor L = Line equivalent length, ft M1,M2 = Mach number at the line inlet, outlet

Compressible Flow-long

Pipelines
Weymouth.

Equations Commonly Used for Calculating Hydraulic Data for Gas Pipe lines Panhandle A.

Qb = 435.87 X (T / pb),07*

~2.6182

xEx

I
L

P 1 -P 2 -

0.

Tavg x Zavg ~0.8539 x L x Tavg x Zavg

1
A

0.5394

Pavg is used to calculate gas compressibility factor Z

Nomenclature for Panhandle Equations Qb = flow rate, SCFD Pb = base pressure, psia R Tb = base temperature, O = average gas temperature, O R Tavg P1 = inlet pressure, psia P2 = outlet pressure, psia G = gas specific gravity (air = 1.0) L = line length, miles Z = average gas compressibility D = pipe inside diameter, in. h2 = elevation at terminus of line, ft

Panhandle B.
Qb

= 737x (Tb/pb)
2 2

l.020

~ 2 . 5 3

xE

0.0375 x G x (h2 - h1) x Pavg P 1 -P2 Tavg x Zavg ~0.961 x L x Tavg x Zavg
L

1
1

0.51

Fluid Flow

19

hl = elevation at origin of line, ft Pavg = average line pressure, psia E = efficiency factor E = 1 for new pipe with no bends, fittings, or pipe diameter changes E = 0.95 for very good operating conditions, typically through first 12-1 8 months E = 0.92 for average operating conditions E = 0.85 for unfavorable operating conditions Nomenclature for Weymouth Equation Q = flow rate, MCFD Tb = base temperature, "R Pb = base pressure, psia G = gas specific gravity (air = 1) L = line length, miles T = gas temperature, "R Z = gas compressibility factor D = pipe inside diameter, in. E = efficiency factor. (See Panhandle nomenclature for suggested efficiency factors) Sample Calculations Q=? G = 0.6 T = 100F L = 20 miles P1 = 2,000psia P2 = 1,500psia Elev diff. = 100 ft D = 4.026-in. Tb = 60F Pb = 14.7psia E = 1.0 Pa", = 2/3(2,000 + 1,500)) = 1,762psia

Panhandle A. Qb = 435.87 X (520/14.7)1'07BB X (4.026)2'6182 X 1X


0.5394

0.0375 x 0.6 x 100 x (1,762)2 [(2,000)' -(1,500)' 560 x 0.835 I (0.6)'8539 x 20 x 560 x 335 Qb = 16,577 MCFD Panhandle B. Qb = 737 X (520/14.7)1'020 X (4.026)2.53 X 1X

(2,000)' -(1,500)' -

0.0375 x 0.6 x 100 x (1,762)2 560 x 0.835 (0.6)'961 x 20 x 560 x 335

Qb = 17,498 MCFD Weymouth. Q = 0.433 x (520/14.7) x [(2,000)' -(1,500)'/ (0.6 x 20 x 560 x 0.835)]1'2 x (4.026)2'667 Q = 11,101 MCFD Source Pipecalc 2.0, Gulf Publishing Company, Houston, Texas. Note: Pipecalc 2.0 will calculate the compressibility factor, minimum pipe ID, upstream pressure, downstream pressure, and flow rate for Panhandle A, Panhandle B, Weymouth, AGA, and Colebrook-White equations. The flow rates calculated in the above sample calculations will differ slightly from those calculated with Pipecalc 2.0 since the viscosity used in the examples was extracted from Reference 2. Pipecalc uses the Dranchuk et al. method for calculating gas compressibility.

1,500 - (2,000 x 1,500/2,000

Z at 1,762psia and 100F = 0.835.

20

Rules of Thumb for Chemical Engineers

Equivalent lengths for Multiple lines Based on Panhandle A Condition 1. A single pipe line which consists of two or more different diameter lines.
Let L E = equivalent length L1, L2, . . . L, = length of each diameter D1, D2, . . . D, = internal diameter of each separate line corresponding to L1, L2,. . . Ln D E = equivalent internal diameter
4.8539 4.8539

dl, d2, d3 & d, = internal diameter of individual line corresponding to lengths L1, L2, L3 & Ln

r
LE = L1Ld12.6182

,
aE

2.6182

,1.8539

+ d22.6182 + d32'6182 + ... d, 2.6182

+...
2.6182

Le =L1[%]

+L2[%]

+***L,[2T39 when L1 = length of unlooped section L2 = length of single looped section L3 = length of double looped section d E = dl = d2 then:

1.8539

Example. A single pipe line, 100 miles in length consists of 10 miles 10Y4-in. OD; 40 miles 123/4-in. OD and 50 miles of 22-in. OD lines. Find equivalent length (LE) in terms of 22411. OD pipe.
4.8539 4.8539

LE =50+40[-]

+lo[%]

= 50 614 364 = 1,028 miles equivalent length of 22-in. OD

when d E = dl = d2 = d3 then L E = L1 + 0.27664 L2 + 0.1305 L3

Condition II. A multiple pipe line system consisting of two or more parallel lines of different diameters and different lengths.
Let L E = equivalent length L1, L2, L3, . . . L, = length of various looped sections dl, d2, d3, . . . d,= internal diameter of the individual line corresponding to length LI, L2, L3 & Ln d E 2.6182 2.6182 +d3 +... d, 2.6182

Example. A multiple system consisting of a 15 mile section of 3-8Y8-in. OD lines and 1-10Y4-in. OD line, and a 30 mile section Of 2-8Y8-h. lines and l-103/4-in. OD line. Find the equivalent length in terms of single 12411. ID line.
122.6182 LE =15 [3(7.981):ili? 122.6182

+ 10.022.6182

1.8539

LE = 'I[

d;.6182

+d2

2.6182

+...

',[
Let

d12.6182

dE2.6182 2.6182 2.6182 +d2 +d3 +... d, 2.6182

1 1

1.8539

+ 30[

= 5.9 + 18.1 = 24.0 miles equivalent of 12-in. ID pipe

+ 10.022.6182 2(7.98 1)2'6182

1.8539

1.8539

L E = equivalent length L,. L,. L, & L, = length of various lootled sections

Example. A multiple system consisting of a single 12-in. ID line 5 miles in length and a 30 mile section of 3-12-in. ID lines. Find equivalent length in terms of a single 12-in. ID

Fluid Flow

21

LE = 5 + 0.1305 x 30 = 8.92 miles equivalent of single 12-in. ID line

2. McAllister, E. W., Pipe Line Rules of Thumb Handbook, 3rd Ed., Gulf Publishing Co., pp. 247-248, 1993. 3. Branan, C. R., The Process Engineer5 Pocket Handbook, Vol. 1, Gulf Publishing Co., p. 4, 1976.

References 1. Maxwell, J. B., Data Book on Hydrocarbons, Van Nostrand, 1965.

Sonic Velocity
To determine sonic velocity, use V, =4 where V, = Sonic velocity, ft/sec K = Cp/C, the ratio of specific heats at constant pressure to constant volume. This ratio is 1.4 for most diatomic gases. g = 32.2ft/sec2 R = 1,544/mol. wt. T = Absolute temperature in O R If pressure drop is high enough to exceed the critical ratio, sonic velocity will be reached. When K = 1.4, ratio = 0.53. Source Branan, C. R., The. Process Engineer 5 Pocket Handbook, Vol. 1, Gulf Publishing Co., 1976. To determine the critical pressure ratio for gas sonic velocity across a nozzle or orifice use critical pressure ratio = [2/(K + l)]V(k-1)

Metering
Orifice 2g = 64.4ft/sec2 Ah = Orifice pressure drop, ft of fluid D = Diameter C, = Coefficient. (Use 0.60 for typical application where DJDp is between 0.2 and 0.8 and Re at vena contracts is above 15,000.) Venturi Same equation as for orifice: C, = 0.98 Permanent head loss approximately 3 4 % Ah.

(u,'

= C,(2gAh)'12

Permanent head loss % of Ah Permanent D /D Loss u 0.2 95 0.4 82 0.6 63 0.8 40 One designer uses perman nt loss = Ah (1 - C,) where U, = Velocity through orifice, ft/sec Up= Velocity through pipe, ft/sec

22

Rules of Thumb for Chemical Engineers

Rectangular Weir

Pitot Tube Ah = u2/2g Source Branan, C. R., The Process Engineers Pocket Handbook Vol. 1, Gulf Publishing Co., 1976.

F, = 3.33(L - 0.2H)H3/*
where Fv = Flow in ft3/sec L = Width of weir, ft H = Height of liquid over weir, ft

Control Valves
Notes: 1. References 1 and 2 were used extensively for this section. The sizing procedure is generally that of Fisher Controls Company. 2. Use manufacturers data where available. This handbook will provide approximate parameters applicable to a wide range of manufacturers. 3. For any control valve design be sure to use one of the modern methods, such as that given here, that takes into account such things as control valve pressure recovery factors and gas transition to incompressible flow at critical pressure drop. where
= Maximum allowable differential pressure for APallow sizing purposes, psi K, = Valve recovery coefficient (see Table 3) rc = Critical pressure ratio (see Figures 1 and 2) PI = Body inlet pressure, psia P, = Vapor pressure of liquid at body inlet temperature, psia

This gives the maximum AP that is effective in producing flow. Above this AP no additional flow will be produced since flow will be restricted by flashing. Do not use a number higher than APallow in the liquid sizing formula.

liquid Flow Across a control valve the fluid is accelerated to some maximum velocity. At this point the pressure reduces to its lowest value. If this pressure is lower than the liquids vapor pressure, flashing will produce bubbles or cavities of vapor. The pressure will rise or recover downstream of the lowest pressure point. If the pressure rises to above the vapor pressure, the bubbles or cavities collapse. This causes noise, vibration, and physical damage. When there is a choice, design for no flashing. When there is no choice, locate the valve to flash into a vessel if possible. If flashing or cavitation cannot be avoided, select hardware that can withstand these severe conditions. The downstream line will have to be sized for two phase flow. It is suggested to use a long conical adaptor from the control valve to the downstream line. When sizing liquid control valves first use Apallow = Km(P1 - rcpv)
Critical Pressure Ratios For Water

VAPOR PRESSURE-PSIA

Figure 1. Enter on the abscissa at the water vapor pressure at the valve inlet. Proceed vertically to intersect the , on the ordicurve. Move horizontally to the left to read r nate (Reference 1). LIVE GRAPH
Click here to view

Fluid Flow

23

Critical Pressure Ratios For liquids Other Than Water


2

Table 1 Critical Pressure of Various Fluids, Psia*


Ammonia 1636 Argon 705.6 Butane 550.4 Carbon Dioxide.................1071.6 Carbon Monoxide 507.5 Chlorine 11 18.7 Dowtherm A 465 Ethane 708 Ethylene 735 Fluorine 808.5 Helium 33.2 Hydrogen 188.2 Hydrogen Chloride 1 1 98

I 1.0

2 0.9 w 5 0.8
t
v)

8 0.7 a
Q

0.6

0
0

.I0 .20 .30 .40 S O .60 70 .80 .90 1.00 VAPOR PRESSURE- PSlA CRITICAL PRESSURE- PSlA

........................... ................................... ................................. ............... ............................. ........................ ................................. .............................. ............................... ................................... ............................. ............

lsobutane lsobutylene Methane Nitrogen Nitrous Oxide Oxygen Phosgene Propane Propylene Refrigerant 1 1 Refrigerant 12 Refrigerant 22 Water

.............. 529.2 ............. 580 ............................. ,673.3 .............................. 492.4 ................... 1047.6 ............................... 736.5 ........................... 823.2 .............................. 617.4 ........................... 670.3 ........... 635 ........... 596.9 ........... 716 ................................ 3206.2

*For values not listed, consult an appropriate reference book.

Figure 2. Determine the vapor pressurelcritical pressure ratio by dividing the liquid vapor pressure at the valve inlet by the critical pressure of the liquid. Enter on the abscissa at the ratio just calculated and proceed vertically to intersect the curve. Move horizontally to the left and read rc on the ordinate (Reference 1). LIVE GRAPH
Click here to view

Gas and Steam Flow

The gas and steam sizing formulas are Gas


I

Some designers use as the minimum pressure for flash check the upstream absolute pressure minus two times control valve pressure drop. Table 1 gives critical pressures for miscellaneous fluids. Table 2 gives relative flow capacities of various types of control valves. This is a rough guide to use in lieu of manufacturers data. The liquid sizing formula is C,=Q where C, = Liquid sizing coefficient Q = Flow rate in GPM AP = Body differential pressure, psi G = Specific gravity (water at 60F = 1.0) Two liquid control valve sizing rules of thumb are 1. No viscosity correction necessary if viscosity S20 centistokes. 2. For sizing a flashing control valve add the C,s of the liquid and the vapor.

Table 2 Relative Flow Capacities of Control Valves (Reference 2)

d,,

Double-seat globe Single-seat top-guided globe Single-seat split body Sliding gate Single-seat top-entry cage Eccentric rotating plug (Camflex) 60open butterfly Single-seat Y valve (300 8 600 Ib) Saunders type (unlined) Saunders type (lined) Throttling (characterized) ball Single-seat streamlined angle (flow-to-close) 90open butterfly (average)

12 11.5 12 6 12 13.5 14 18 19 20 15 25 26 32

11 10.8 11.3 6 11 12.5 13 15.5 16.5 17 13.5 20

11 10 10
na

11.5 12 12 14
na na

15 13 18

20 21.5

Note: This table may serve as a rough guide only since actual flow capacities differ between manufacturers products and individual valve sizes. (Source: ISA Handbook of Control Valves Page 17). *Valve flow coefficient C, = C, x d 2 (d = valve dia., in.). +C,ld2of valve when installed between pipe reducers (pipe dia. 2 x valve dia.). **C,ld2 of valve when undergoing critical (choked) flow conditions.

24

Rules of Thumb for Chemical Engineers

Steam (under 1,OOOpsig)

Table 4 Correlations of Control Valve Coefficients (Reference 2)


CI = 36.59 Cf Cl = 36.59 c, = CIC, K, = C$ = F L ~ C, = 1.83 CfC, c, = 19.99 C&,

c, =

Qs(l+O.OOO65TSh) PI sin - :: 3

c, = 19.99 c,

Steam and Vapors (all vapors, including steam under any pressure conditions)

Values of K, calculated from C, agree within 10% of published data of K,. Values of C1calculated from K, are within 21% of published data of C1.

c, = 1.06&&in[

Qs

T&]
3417 AP
deg.

When the bracketed quantity in the equations equals or exceeds 90 degrees, critical flow is indicated. The quantity must be limited to 90 degrees. This then becomes unity since sin 90" = 1. Explanation of terms: C, = Cg/C, (some sizing methods use Cf or Y in place of Cl) Cg= Gas sizing coefficient C, = Steam sizing coefficient C, = Liquid sizing coefficient d, = Density of steam or vapor at inlet, lbs/ft3
Table 3 Average Valve-Recovery Coefficients, K, and C1* (Reference 2)
Type of Valve

20

40

60

80

100

Percent of rated travel

Figure 3. These are characteristic curves of common valves (Reference 2). LIVE GRAPH
Click here to view

Cage-trim globes: Unbalanced Balanced Butterfly: Fishtail Conventional Ball: Vee-ball, modified-ball, etc. Full-area ball Conventional globe: Single and double port (full port) Single and double port (reduced port) Three way Angle: Flow tends to open (standard body) Flow tends to close (standard body) Flow tends to close (venturi outlet) Camflex: Flow tends to close Flow tends to open Split body

0.8 0.70 0.43 0.55 0.40 0.30 0.75 0.65 0.75 0.85 0.50 0.20 0.72 0.46 0.80

33 33 16 24.7 22

G = Gas specific gravity = mol. wt./29 PI = Valve inlet pressure, psia AP = Pressure drop across valve, psi Q = Gas flow rate, SCFH Q, = Steam or vapor flow rate, lb/hr T = Absolute temperature of gas at inlet, OR Tsh= Degrees of superheat, O F The control valve coefficients in Table 4 are for hll open conditions. The control valve must be designed to operate at partial open conditions for good control. Figure 3 shows partial open performance for a number of trim types. General Control Valve Rules of Thumb

35 35

24.9 31.1 35

*For use only if not available from manufacturer.

1. Design tolerance. Many use the greater of the following: Qsizing = 1-3 Qnomal Qsizing = 1.1 Qmaximum

Fluid Flow

25

2. Type of trim. Use equal percentage whenever there is a large design uncertainty or wide rangeability is desired. Use linear for small uncertainty cases. Limit max/min flow to about 10 for equal percentage trim and 5 for linear. Equal percentage trim usually requires one larger nominal body size than linear. 3. For good control where possible, make the control valve take 50%-60% ofthe system flowing head loss. 4. For saturated steam keep control valve outlet velocity below 0.25 mach.

5. Keep valve inlet velocity below 300 ft/sec for 2 and smaller, and 200ft/sec for larger sizes.

References
1. Fisher Controls Company, Sizing and Selection Data, Catalog 10. 2. Chalfin, Fluor cove, Specifying Control Valves, Chemical Engineering, October 14, 1974.

Safety Relief Valves


The ASME code provides the basic requirements for over-pressure protection. Section I, Power Boilers, covers fired and unfired steam boilers. All other vessels including exchanger shells and similar pressure containing equipment fall under Section VIII, Pressure Vessels. API RP 520 and lesser API documents supplement the ASME code. These codes specify allowable accumulation, which is the difference between relieving pressure at which the valve reaches h l l rated flow and set pressure at which the valve starts to open. Accumulation is expressed as percentage of set pressure in Table 1. The articles by Rearick2 and Isqacs3 are used throughout this section. Full liquid containers require protection from thermal expansion. Such relief valves are generally quite small. Two examples are 1. Cooling water that can be blocked in with hot fluid still flowing on the other side of an exchanger.
Table 1 Accumulation Expressed as Percentage of Set Pressure
ASME Section I Power Boilers LIQUIDS thermal expansion fire STEAM over-pressure fire GAS OR VAPOR over-pressure fire
ASME Section Vlll Pressure Vessels
10

2. Long lines to tank farms that can lie stagnant and exposed to the sun.
Sizing

Use manufacturers sizing charts and data where available. In lieu of manufacturers data use the formula u = 0.442gAh where Ah = Head loss in feet of flowing fluid u = Velocity in ft/sec g = 32.2ft/sec2 This will give a conservative relief valve area. For compressible fluids use Ah corresponding to Y2P1if head difference is greater than that corresponding to Y2P1(since sonic velocity occurs). If head difference is below that corresponding to Y2P1use actual Ah. For vessels filled with only gas or vapor and exposed to fire use 0.042AS A= (API RP 520, Reference 4)

vplcal Design for Compressors Pumps and Piping 25

20
10

20
10

20
10

20
10

20

20

A = Calculated nozzle area, in.2 PI = Set pressure (psig) x (1 + fraction accumulation) + atmospheric pressure, psia. For example, if accumulation = lo%, then (1 + fraction accumulation) = 1.10 A6 = Exposed surface of vessel, ft2

26

Rules of Thumb for Chemical Engineers

Table 2 Relief Valve Designations

0.1 10 0.196 0.307 0.503 0.785 1.287 1.838 2.853 3.60 4.34 6.38 11.05 16.0 26.0

0
0 0 0 0 0
0

0 0 0
0

1 x2

1.5~3

2x3

2.5 x

4x6

6x8

Valve Bodv Size (Inlet Diameter x Outlet Diameter), in.

This will also give conservative results. For heat input from fire to liquid containing vessels see Determination of Rates of Discharge. The set pressure of a conventional valve is affected by back pressure. The spring setting can be adjusted to compensate for constant back pressure. For a variable back pressure of greater than 10% of the set pressure, it is customary to go to the balanced bellows type which can generally tolerate variable back pressure of up to 40% of set pressure. Table 2 gives standard orifice sizes.

Plants, situations, and causes of overpressure tend to be dissimilar enough to discourage preparation of generalized calculation procedures for the rate of discharge. In lieu of a set procedure most of these problems can be solved satisfactorily by conservative simplification and analysis. It should be noted also that, by general assumption, two unrelated emergency conditions will not occur simultaneously. The first three causes of overpressure on our list are more amenable to generalization than the others and will be discussed.
Fire

Determination of Rates of Discharge

The more common causes of overpressure are 1. External fire 2 . Heat Exchanger Tube Failure 3. Liquid Expansion 4. Cooling Water Failure 5 . Electricity Failure 6. Blocked Outlet 7 . Failure of Automatic Controls 8. Loss of Reflux 9. Chemical Reaction (this heat can sometimes exceed the heat of an external fire). Consider bottom venting for reactive liquid^.^

The heat input from fire is discussed in API RP 520 (Reference 4 ) . One form of their equation for liquid containing vessels is
Q = 2 1,000FAwo.82

where

Q = Heat absorption, Btu/hr Aw = Total wetted surface, ft2 F = Environment factor

Fluid Flow

27

The environmental factors represented by F are Bare vessel = 1.0 Insulated = 0.3/insulation thickness, in. Underground storage = 0.0 Earth covered above grade = 0.03 The height above grade for calculating wetted surface should be 1. For vertical vessels-at least 25 feet above grade or other level at which a fire could be sustained. 2. For horizontal vessels-at least equal to the maximum diameter. 3. For spheres or spheroids-whichever is greater, the equator or 25 feet. Three cases exist for vessels exposed to fire as pointed out by Wonge6A gas filled vessel, below 25ft (flame heights usually stay below this), cannot be protected by a PSV alone. The metal wall will overheat long before the pressure reaches the PSV set point. Wong discusses a number of protective measures. A vessel containing a high boiling point liquid is similar because very little vapor is formed at the relieving pressure, so there is very little heat of vaporization to soak up the fires heat input. A low-boiling-point liquid, in boiling off, has a good heat transfer coefficient to help cool the wall and buy time. Calculate the time required to heat up the liquid and vaporize the inventory. If the time is less than 15 minutes treat the vessel as being gas filled. If the time is more than 15-20 minutes treat it as a safe condition. However, in this event, be sure to check the final pressure of the vessel with the last drop of liquid for PSV sizing. Heat Exchanger Tube Failure 1. Use the fluid entering from twice the cross section of one tube as stated in API RP 5204 (one tube cut in half exposes two cross sections at the 2. Use Ah = u2/2g to calculate leakage. Since this acts similar to an orifice, we need a coefficient; use 0.7.

liquid Expansion The following equation can be used for sizing relief valves for liquid expansion.

Q=,
where

BH

(API RP 520, Reference 4)

Q = Required capacity, gpm H = Heat input, Btu/hr B = Coefficient of volumetric expansion per OF: = 0.0001 for water = 0.0010 for light hydrocarbons = 0.0008 for gasoline = 0.0006 for distillates = 0.0004 for residual fuel oil G = Specific gravity C = Specific heat, Btu/lb O F
Rules of Thumb for Safety Relief Valves 1. Check metallurgy for light hydrocarbons flashing during relief. Very low temperatures can be produced. 2. Always check for reaction force from the tailpipe. 3. Hand jacks are a big help on large relief valves for several reasons. One is to give the operator a chance to reseat a leaking relief valve. 4.Flat seated valves have an advantage over bevel seated valves if the plant forces have to reface the surfaces (usually happens at midnight). 5. The maximum pressure from an explosion of a hydrocarbon and air is 7 x initial pressure, unless it occurs in a long pipe where a standing wave can be set up. It may be cheaper to design some small vessels to withstand an explosion than to provide a safety relief system. It is typical to specify llqn as minimum plate thickness (for carbon steel only). Sources 1. ASME Boiler and Pressure Vessel Code, Sections I and VIII. 2. Rearick, How to Design Pressure Relief Systems, Parts I and 11, Hydrocarbon Processing, August/ September 1969.

so,
u =0 . 7 4 m For compressible fluids, if the downstream head is less than Y2 the upstream head, use Y2 the upstream head as Ah. Otherwise use the actual Ah.

28

Rules of Thumb for Chemical Engineers

3. Isaacs, Marx, Pressure Relief Systems, Chemical Engineering, February 22, 1971. 4. Recommended Practice for the Design and Installation of Pressure Relieving Systems in Refineries, Part I Design, latest edition, Part 11-Installation, latest edition RP 520 American Petroleum Institute.

5. Walter, Y. L. and V. H. Edwards, Consider Bottom Venting for Reactive Liquids, Chemical Engineering Progress, June 2000, p. 34. 6. Wong, W. Y., Improve the Fire Protection of Pressure Vessels, Chemical Engineering, October, 1999, p. 193.