Safety Valve Inlet Pipe Pressure Loss Criteria For High Viscous Service

[jokosmg01]

Dear Experts,



Kindy please advise the inlet pressure loss criteria for high viscous service?

We have viscous fluid viscosity around 15 Pa.s.



As we are aware that API states the inlet line pressure loss shall be less than 3% of set pressure.



Thanks and best regards,

Joko

 

[georgeverghese]

Correct, the inlet dp loss upper limit recommended in the API and ASME of 3% of set pressure applies to all operating viscosities.

 

[jokosmg01]

Hi George, thank you for your feedback
Do you have any reference? Then I can follow as basis of our calculation. thanks

 

[jokosmg01]

Due to based on our calculation it will be difficult if we shall use 3% of set pressure as criteria.

 

[georgeverghese]

This limit may be found in API 521 and maybe API520 also. Also in ASME in one of the appendices. Maybe you can tell us how this is difficult to meet in your case?

 

[jokosmg01]

George,
PSV Set pressure 3.5 barg
Criteria of pressure loss for inlet line of PSV is 3% = 0.105 bar
We have inlet line length 1 meter and 1 gate valve.
Viscosity 15000 cP.
Flow 128250 kg/h
Density 726 kg/m3
The resulf of pressure drop as follows :
size 10" = pressure loss is 0.205 bar,
size 12" = pressure loss is 0.170 bar,
it means higher than criteria.

So that's why Iam asking is there any other standard? or for high viscous service is excluded from API/ASME criteria?

Please advise,

 

[georgeverghese]

Going up to a bigger line size would be inevitable. Also note that a reduced bore gate valve produces about 5x more press drop than a full bore gate valve. Also account for the piping reducer at the inlet of the PSV. Obviously, you should use the rated flow of the PSV and not the calculated process design flow.

 

[jokosmg01]

"INLET PRESSURE LOSS
(BAR)"
10" 12" 14" 16"
0.161 0.17 0.184 0.21

Eventhough we increase the line size, still we can't meet criteria 3% of 3.5 barg = 0.105 barg
How is it?
thanks

 

[georgeverghese]

On a 10inch line, Nre = 12 - am afraid the fitting loss estimation program I have is probably not valid at these conditions ( its more for turbulent flow, and I havent updated it for the latest Crane formulae which have some adjustments for low Nre), so I would not attempt to give you some feedback on the accuracy of these values you have indicated for total dp, since it does look like the fitting losses dominate the total dp.

Maybe there are others reading this discussion who can help with these fitting equivalent lengths for the gate valve and reducers at these low values for Nre.

Also check that you've accounted for exit loss for the RV nozzle on the protected equipment.

 

[georgeverghese]

I get Fanning friction factor to be 1.33, 1.59 and 1.85 for 10, 12 and 14 inch lines, and the general trend is for dp to get less and less as line size goes up in this range, which is different from your results. So pls check friction factors.

Since the exit loss is not included in my approx calcs, it does look like you need more than 14inch.

 

[georgeverghese]

Are your head loss for fittings calcs done with the Hooper 2-K method or the Darby 3-K method ? - for these low values of Nre, would be safer to do these calcs with the 3-K method.

 

[Jprocess01]

The solution is to use a pilot PSV with remote sensing line, and size the valve using upstream pressure of allowable accumulation - inlet line losses.

Also the 3% rule is part on the non-mandatory Appendix M of ASME. You could perform an engineering analysis outlind in API to see if the valve will chatter or not, and whether the pressure accumulation is acceptable.

 

[The Obturator]

Jprocess01 A pilot op PSV may solve the pressure loss issue, but it will not be good for high viscosity fluid due to the small passages in the pilot operator. You may get some special designs offered with chemical barriers.

With the information so far, I would be looking at a spring operated PSV with a top mounted actuator with lift assist. Depending where the installation is located, this set up should be allowed by ASME VIII since the opening is not restricted. The controller to the actuator may also be a triple type as per ISO/PED EU rules (1 working, 2 redundant.

You also need to consider the viscosity correction Kv factor to be applied to the PSV sizing by factoring the Reynolds number per API 520 against initial sizing (as well as the inlet pressure loss). Maybe also consider a heated inlet design by lagging, jacketing or other. These factors of course affect the rated flow.

Per ISO, only the term Safety Valve is used regardless of application or design.

 

[The Obturator]

I neglected to add...The actuated spring operated design allows controlled lift at set pressure (or before if desired - but check with vendor), thereby eliminating chatter or unstable valve operation. If closing (blowdown) is an issue, consider the actuator to be double acting which would include assisted closing (if permitted by code).

The controller/sensor for the actuator would need a chemical barrier.

Per ISO, only the term Safety Valve is used regardless of application or design.

 

[georgeverghese]

Suspect there is something amiss with the pressure drop calcs here.

 

[vladrath]

I know from personal experience how difficult it can be to pipe up these highly viscous fluids.

One thing I didn't notice anyone else commenting on. If the gate valve is on a line you are designating as your 'relief path' you should only have a full bore valve.

API 520 part 2 includes provisions on engineering a larger inlet pressure drop. You might want to proceed down this path. I would point out that you should ensure you carefully document your thought process for this project. So that if other folks in your plant who don't readily understand the process will at least understand that you do.

The pilot operated valve would be my second choice if you cannot engineer out the inlet pressure drop. Of course you will have some challenges with the sensing line. You might need to trace it, or something to ensure you don't plug it up.