Hi Pavan @pxyarala ,
First I calculated the backpressure at the PSV outlet flange with rated flow using my program HEM:
Estimate of backpressure with 3 data states (Model F).
Lengths, diameters and losses for pipes.
Mass flow rate = 124525 lb/hr
Surrounding pressure = 14.70 psia
Pipe length = 45.92 ft
Inside diameter = 4.026 in
Total turb. fitting VH losses = 1.20
Total laminar fitting losses = 0
Elevation difference (up is +) = -34.50 ft
Pipe sand roughness = 0.00180 in
========================== PROPERTIES ==========================
DATA STATE A B C
========== ======= ======= =======
Pressure, psia 94.45 28.80 14.70
Wt. fraction gas 0.00000 0.00001 0.03757
Gas density, lb/ft3 0.22690 0.07003 0.03731
Liquid density, lb/ft3 58.890 58.875 59.830
================================================================
PIPING RESULTS:
Upstream press. = 27.35 psia
PIPE EXIT:
Pressure = 28.81 psia
Thrust / area = 14.7 psi
Velocity = 6.6 ft/s
Wt. frac. gas = 0.00000
Friction f = 0.01702
Next, I evaluated the “installation” using my program SafSiz:
CASE:
Liquid expands and vents through a bellows Spec. K pressure relief valve.
Relief device flowrate was specified.
DESIGN CONDITIONS:
Set pressure = 72.5 psig; Back pressure = 12.6 psig
Overpressure = 10 %; Input sizing temp. = 248.0 F
INSTALLATION RESULTS:
Nozzle area = 1.287 in2; Rated coeff. = 0.650
Inlet pressure loss is 0.6 psi, based on the actual flowrate.
Flowrate (lb/hr) = 125987 (rated) 139986 (actual)
Installation-rated flowrate for bellows liquid PRV = 125562 lb/hr
Maximum replacement valve capacity = 272.3 gpm water at 70 F.
PHYSICAL PROPERTIES:
Properties based on WATER at 248.0 F and 94.4 psia.
Liquid density = 58.89 lb/ft3;
Use HEM to obtain a Stored Energy estimate, if required.
WARNING:
Flashing may occur in the discharge piping. If so, evaluate with HEM.
ADDITIONAL DATA:
=================== PIPE & FITTING SUMMARY ===================
Inlet
=========
Diameter, in 3.068
Length, ft 2.8
Reynolds number 2.585E+05
Fitting fric. factor 0.0189
# Contractions 1
# 45 degree elbows 0
# 90 degree elbows 0
# Gate Valves 0
# Rupture disks 0
Other Vel. Hd. losses 0.00
Total fit. VH losses 0.33
================================================================
Lastly, I modelled the PSV in my program HEM to double check there is essentially no flashing at the exit of the PSV’s flow nozzle:
PSV Sizing for Two-Phase Flashing Flow
Estimate of flow rate with 3 data states (Model F).
Areas and coefficients for nozzles.
Reservoir pressure = 94.45 psia
Surrounding pressure = 14.70 psia
Nozzle area = 1.287 in2
Nozzle coefficient = 0.6500
========================== PROPERTIES ==========================
DATA STATE A B C
========== ======= ======= =======
Pressure, psia 94.45 28.80 14.70
Wt. fraction gas 0.00000 0.00001 0.03757
Gas density, lb/ft3 0.22690 0.07003 0.03731
Liquid density, lb/ft3 58.890 58.880 59.830
================================================================
REAL NOZZLE RESULTS:
Mass velocity = 97042.0 lb/hr/in2
Mass flow rate = 124893 lb/hr
Stored energy = 219.16 Btu/ft3
NOZZLE THROAT:
Pressure = 28.80 psia
Thrust / area = 70.2 psi
Velocity = 67.0 ft/s
Wt. frac. gas = 0.00002
I call 0.00002 wt. fraction gas, essentially no flashing. So, everything looks good to me. Any questions?
Hi Latexman,
Thank you very much for taking time in performing the HEM calculations for me in your software. I am really very grateful to you.
I have some questions regarding the results. Kindly help me understand the results when you have some time.
-
Under piping results: I see the upstream pressure as 27.35 psia, and the pipe exit pressure as 28.81 psia. That means P1 at pipe entrance is 27.35 psia. Then what is pressure at the pipe exit. Under the pipe exit I see pressure as 28.81 psia. Is this pipe exit pressure?. If so how can this be higher than pipe upstream pressure. What is the PSV outlet piping DP?. Also I forgot to mention there are 90 degree LR elbow - 4 Nos in the pipe. Can you please add these into the outlet piping and re-run the program. My calculation per API 521 5.5.10 showed upstream pipe pressure as 37.22 psia while the downstream pressure as 26.37 psia, which seems to wrong compared to your results.
-
Under PSV Sizing : I see that the nozzle/throat pressure as 28.80 psia. Is this pressure at PSV nozzle exit. And you say that the pipe entrance pressure is 27.35 psia. How come both these pressures are different?. Shouldn’t they be same?.
I am sorry I could not understand how to interpret the results. My questions may sound silly but it would be really helpful if you explain the results to me.
Thanks and Regards,
Pavan Kumar
Hi Latexman,
I forgot to ask you if Safsiz estimated the required orifice area using the required relief rate of 119362 lb/hr at 79.75 psig relieving pressure and 248 Deg F relief temperature. Did it estimate that “J” orifice with API orifice area = 1.287 in2 is required?.
Also it will be really helpful if you can share a copy of the paper “Easily size relief devices and piping for two-phase flow” by J.C Leung.
Thanks and Regards,
Pavan Kumar
No, the “Sizing” section of SafSiz uses API areas and defaults for flow coefficients, not PSV specific/certified numbers, and it selected a “K” orifice. I did not copy the sizing section to the website, as I did not want to confuse. The “Installation” section uses your PSV specific/certified numbers that I manually entered, and the installation is adequate, 125987 > 119362.
I’ll look for that article when back in the office Monday.
Hi Latexman,
Can you please help me the paper “Easily size relief devices and piping for two-phase flow” by J.C Leung, CEP, December 1996, pages 28-50. The copy that I have the pages from 36-40 missing. I want to use this paper to calculate the PSV outlet line pressure drop.
Easily Size Relief Devices and Piping for Two Phase Flow - JC Leung.pdf (2.0 MB)
Sure, I asked our Information Resources Manager to obtain it for me today. I want it for my library too.
Did you correct for elevation?
If not, then a comparison of our numbers with elevation correction is:
|
Backpressure |
Exit Pressure |
|
psia |
psia |
| pxyarala |
37.22-14.1 = 23.12 |
26.37 |
| Latexman |
27.35 |
28.81 |
Not too bad, IMO. Both say the PSV is adequate.
Hi Latexman,
I am not sure we are comparing apples to apples here. The method in API 521 5.5.10 calculates P2 based on the choked flow condition. The greater of choked pressure ( Pc) and pressure outside of the pipe (P3) is taken as P2. Once this is done the calculation back calculates P1 based on the pipe resistance and accounting for flashing. Also to be noted the method does not taken in account for the elevation changes. It is good for horizontal line.
Considering my line is horizontal I got P2 as 26.37 psia which happens to the be the choke pressure and I back calculated to get P1 = 37.22 psia with frictional DP=10.85 psi. Now if you say P2=26.37 psia includes the static head=14.1 psia then P2 due to friction, I call it P2f=26.37-14.1=12.27 psia and then adding the 10.85 psi DP, I will get P1=23.12 ( =12.27+10.85).
This still does not make sense to me because P2 calculated by API 521 was based on choked flow condition. I am not sure adding the static head pressure to get the choked flow pressure is the right way or not.
I am stuck in this question and want to check the how Leung accounts for pipe elevation change in his paper “Easily Size Relief Devices and Piping for Two Phase Flow”. I see him mentioning the pipe inclination factor Fi i the paper. It would have been very helpful if there was a worked out example.
If I use Aspen with a orifice for relief valve and pipe segment for the outlet pipe can I get the same answer?.
Thanks and Regards,
Pavan Kumar
I think I see what you mean. As a parallel thought to that, which may emphasize the effect, the absolute pressure in the pipe at any point will impact the vapor/liquid water equilibrium at that point. I can see how with downflow the static head gain is almost cancelled by the frictional loss as the water goes down the pipe, so the dynamic pressure stays more constant. If it were upflow, the dynamic pressure would be reduced by the frictional loss and the static head change, and will change much more rapidly.
As such, I think the static pressure correction is a must, so the correct absolute pressure at a point in the pipe can be used to predict the vapor/liquid equilibrium. Especially on a system like this, where the conditions are right on the edge of being sub-cooled, saturated, or two-phase.
This reminds me of a real life example, when a cooling water return line to the cooling towers goes up high into the pipe racks, it is quite common to have cavitation problems at that point/level. Warm water with decreasing pressure (friction) raised to a level where it boils.
I’m only an occaisional Aspen user; not an expert. I’d use a flow nozzle; not an orifice. It should work.
The article is on order. I wish I had an ETA, but I don’t.
Hi Latexman,
I fully agree with you that the effect of static head loss of the vapor/ liquid equilibrium must be taken. But also with reduced liquid fraction due to flashing the static head is determined by the liquid fraction that remains at the end of the pipe. I will read Leung’s paper of the pages that I have once again and work it out. I will also model the system in Aspen as you said and try.
Thanks and Regards,
Pavan Kumar
Hi Latexman,
I am very grateful to you for sending the paper. I will read the missing pages and try and get an answer. I can pay the cost for acquiring this paper. Please let me know how I can compensate you?.
Thanks and Regards,
Pavan Kumar
No worries!
I sent a different, but highly pertinent .PDF to you that was too big for the website. I hope my and your e-mail systems don’t scrape it off for being too big.
Hi Latexman,
I saw your email on my company email address. Thank you so much for sending the presentation. It will be very useful for me in the coming days.
Thanks and Regards,
Pavan Kumar
