Clear Up Pressure Relief

Valve Sizing Methods

electing the correct pressure relief parameter is dependent only on the fluid
Aflow-factor
technique gives
an improved
S valve size for overpressure _Qro-
tection is usually done based on
an API orifice area and inlet pipe
size. The API National Board of Boiler and
Pressure Vessel Inspeetors certified relief
properties and, where applicable, on the
valve back pressure derating factor. Using
a computer or calculator to determine valve
size would be easier and the KA parameter
calculated would be a more accurate mea-
measure ofa capacity of valves selected on that basis, sure of the valve size required.
however, can vary widely, depending on
relief valve's what nozzle coefficient and orifice area
performance. was used in calculating the valve capacity. Criteria for sizing
To make valve selection easier and to The procedure used in sizing and select-
know what relief capacity should be select- ing safety relief valves is to establish the set
ed. a new valve parameter called a "flow pressure where the valve is to open. deter-
factor" is proposed. This factor would be mine the required relieving capacity, and
equal to KA. where K is the ASME valve calculate the required orifice area of the
nozzle coefficient and A is the actual valve valve. The set pressure is usually deter-
nozzle area. Use of such a factor would mined by the applicable code or particular
make it easier to compare and select valves operating conditions. The ASME Section
and would make standards such as API VIII Code (2) for unfired pressure vessels
526 (1) more meaningful since a better requiresthe primary pressure rellef valve be
measure of a valve's relief capacity would set no higher than the maximum allowable
be indicated. working pressure (MAWP) of the vessel.
Secondary valves in multiple valve installa-
tions can be set higher than MA WP, but no
Donaid M. Papa, Need for KA flow factor more than 1.05% higher of MAWP (2).
Anderson, Greenwood & Co. The discrepancy between actual and Determination of the purified relieving
advertised K and A has caused confusion capacity is based on a worst case analysis
among users, inspectors, and manufacturers. of the system being protected. Correctly
Some users and inspeetors have a difficult determining worst case is based on an engi-
time trying to reconcile the difference neering judgment. Generally any equip-
between the advertised and National Board ment failure, operator error or extemal con-
lisred K and A of the different valve manu- dition, such as fire, that would result in an
facturer' s products. Manufacturers and their overpressure condition should be consid-
representatives are sometimes confused ered. Af ter the set pressure and required
when asked to explain these differences. relieving capacity are determined, the
Sizing errors can be made if the
advertised and the National Board list-
ed Ks and As are mixed. For example,
using an advertised K with a National
Board listed A would overstate the eer-
tified capacity about 10%. Conversely,
using the National Board listed K with
the advertised A would understate the
capacity about 10%.
A KA factor would simplify the cal-
culation of valve size since the sizing

CHEMICALENGINEERING PROGRESS • AUGUST1991 • 81

 SAFETY

(6)
relieving area of the pressure relief KJ =_Actual Flow
valve can be calculated. Theoretical Flow

(K = O.90KJ, ASME)
Sizing equations (Kd is based on a lift of seat disk
Three basic equations are used for great enough so that nozzle area con-
calculating the relieving area of a trols flow). The other variables in the
pressure relief valve. Two are for gas flow equations are dependent on the
or vapor, one is for liquid and one is gas, vapor or liquid properties and
for steam. Refer to equations I. 2. 3. that gas or vapor at the pressure and conditions. One exception is the der-
and4. temperature conditions in the nozzle. ating factor Kw for balanced direct
Sonic flow: A distinguishing characteristic of spring valves for liquid relief.
sonic tlow is that is is not dependent However this factor is dependent on
( la)
t.::': on downstteam pressure. Subsonic back pressure.
A=yj_MTZ_ gas or vapor flow and all liquid flow
6.32CKP, is dependent on upstream and down-
stream pressure. The transition from Valve selection
or
sonic to subsortie gas or vapor tlow Valve size is usually selected on
( l b) occurs when the absolute pressure at the basis of orifice area. The areas
A = - Wv'TZ the nozzle exit is approximately 50% referred to are frequently those listed
CKPrY~'l
of the absolute pressure at the nozzle in API Standard 526. Refer to Table
inlet. Equation 5 is used to more pre- l. Valve manufacturers usually list
Subsonic tlow: cisely calculate this pressure at the their valves by inlet size and the AP!
nozzle exit. letter designation for nozzle area.
(2a)
A = _yJMTZ__ (Sa) However that area can vary from
4645FKP, manufacturer to manufacturer.
Further inequities occur with the
or: valve coefficient of discharge.
where: Table 2 is a list of areas and nozzle
W\'TZ (5b) coefficients tor some cornrnonly used
A= _ k = Cp API -r orifice valves. The J orifice
735FKP1 \ M area is 1.287 sq. in. The actual arcas
Cv
available Irorn the different manutac-
where: Two terms common to all four turers vary trom 1..+27 to 1.635 sq. in.
equations are K and A. K is the noz- The ASME and advertised nozzle
zie coefficient of discharge and A is coefficients vary from 0.788 to 0.975.
the nozzle orifice area. The nozzle Table 3 is a list of KAs for [he
coefficient of discharge is [he actual same valves and how they compare to
tlow divided by the theoretical flow the API "1" orifice multiplied by an
(2c) for the sarne nozzle with no flow assumed K of 0.90. The deviation
losses. Refer to equation 6. from the API-KA varies trom 106%
(2d)
k= çp
Cv

Liquid tlow:
(3)
A = ---- (Jf_MLCi _
38KKwKV'/PI-P2
Steam:
W_~_
:A= _ ___ (4)
51.5 KP, KSH
The two equations given for gas
or vapor are for sonic flow and sub-
sonic flow. Sonic flow occurs when
the velocity at the exit of the valve
nozzle is the velocity of sound for

82 • AUGUST 1991 • CHEMICAL ENGINEERING

 this flow factor in any future revision
of API 526. KA would replace the
oritïce areas now published. The K in
the KA factor would be 0.90. Table 4
is a list of the proposed flow factors.

to 112%. A K of 0.90 was used since Proposed valve flow Standardization's purpose
this is theoretically the largest possible parameter Standardization is an attempt to
derated nozzle coefficient. The ASME To simplify valve orifice area cal- make problem solving easier. The
Code (3) requires all valve coefficients culations and selection, a new valve proposed KA tlow factor is a step in
be derated 10%. This requirement was tlow parameter called "flow factor" that direction. This factor would be a
added to the Code in 1962. The coeffi- is proposed. This flow factor would direct measure of a valves relieving
cients of all valves sold prior to that be KA where K is the ASME valve capacity, so when an API size valve is
date were not derated. coefficicnt and A is the actual valve specificd, the user woulel know what
In comparing the K and A of nozzle area listed with the Natienul its relicving capacity would he with
valves. additional confusion occurs Board. All valves would have a KA respect to a standurd. ~
because of a difference between the factor. When sizing and selecting a
udvertised values and the ones listed valve. KA would be calculated and a
in the National Board of Boiler and valve selected with a KA equal to or NOMENCLATURE
A = Vulve Orificc Area (sq. in.)
Pressure Vessel Inspeetors Pressure grcater than that calculated. The siz- V = Flow Rare rscrrru
Relief Device Certifications Book. ing equations would take the form W = Flow Rare (lh/hf)
Frequently. neither the Knor the A sown in Equations 7a-d. M = Molecular Wcight
lisred in the National Board Book T = Inkt Tcmpcraturc ('F +..--1-(0)----
Sonic: Oa) Z = Cornprcssibility Factor
agree with the advertised K and A; C = Gas Constant Bascd on Ratio Specific
however the advertised product of KA Heat al Standard Condinons
is always equal to or less than the KA = vY MTZ_ = _wtrz_ K = Nozzle Coeuicicm (:\SME)
National Board lisring. 6.32CPI CPI V M PI = Set +Overprcssurc + LOI.:al
Atrnospheric
P, = Outlet Pressure (psia)
Subsonic: Ob) M = Molecular Weight
Background for difference GPM = Flow Rare (gal/min)
in K and A G = Specific Gruvity
The reasou for the difference KA = VY_MlZ_ = wlIZ_ _ Kw = Back Pressure Correction Factor
(Bulanced Spring Operared SR Vs)
between the advertised K and A 4645CPI 735FPI V M
KSH = SteamSuperheat Correct ion Factor
und the National Board listed K Kv = Viscostry Correct ion Factor
-----;and A dates back to 1962 when the
ASME Section VIII Code was Liquid: (7c)
llTERATURE CITED
c h an g e d to derate all cert ifie d
l. "Flanged Steel Safety Relief Valves."
relieving capacitics 10%. Most StanJard 526. 3rd ed .. American Petroleum
..manufacturers elected to comply Institute, Washington, D.e. (Feb. 19X4).
iwi-th the code revision by not derat-
ing their advertised capacity or 2. "Boiler and Pressure Vessel Code.
Pressure Vessels." Sectien V Ill, Div. I,
[their nozzle coefficients, but by Steam: Od)
UG-134 (a), Addenda Dec. 31.1989,
.increasing the nozzle area of the Arnericun Society of Meenanical
'safety valve 10% so the product of KA= W Engineers. New Vork (July 19X9).
KA remained unchanged. However 51.5 PI KSH
3. "Boiler and Pressure Vessel Code.
rhe API orifice areas were still Manufacturers would publish KA
Pressure Vessels." Sectien VIII. Div. I,
advertised. Therefore, valves were where K is the ASME K published in UG-131 (d)(l). (d)(2)(aJ. (e)(2). Addenda
advertised as having orifice areas the National Board' s certification Dec. 31. 19X9. Arne r icun Society of
with nozzle coefficients greater book and A is the actual orifice area. Meenanical Engineers. New Vork (July
A turther proposal would be to use 1(89).
lhan 0.90.

CHEMICAL ENGINEERING PROGRESS • AUGUST 1991 • 83