PSV installation distance from PCV

[txplpro]

My company is reviewing a client's application for adding a new 36" 760kbpd ASME Class 600 crude pipeline to an existing Class 150 B31.4 tank terminal. The client, the terminal owner/operator wants our engineering company to accept their design as-is with a PSV installed 1600 linear feet away downstream of the new PCV. The specification break is at the first manual block valve 20 ft downstream of the PCV. There are additional manual isolation valves between the PCV and PSV, all locked open, according to the client. If the PCV fails open or at too high a pressure above setpoint, the terminal piping can be overpressured.
I called this into question as there is too much distance between the PCV and PSV. I would rather see the PSV installed as close as possible to the first blockvalve / specification break.
Does anyone agree that the PSV is close enough at 1600' away? Why?

Facts of interest-
Client's design is at 60% review stage. A HAZOP was done by others, without my co. participating.
The PSV sizing was done before this project began. The client 'feels' it is adequate as to location and relieving capacity. It is located in Class 300 piping, but the piping between the PCV and PSV is Class 150. I don't know what the setpoint is.

The client is an international major like (Exxon, Chevron, Enbridge, or one like them).
The clients own standards prohibit doing what they are planning to do.
-locate PSV as close as possible to the spec. break (not 1600' away)
-overpressure protection must be automatic and continuous (not dependent on operator intervention)
-Terminal piping downstream of the pipeline discharge must be Class 300 minimum (not Class 150)

Many thanks and regards. I use this forum frequently.

 

[EmmanuelTop]

A sketch would help tremendously. I am not sure if I figured the whole story.

Dejan IVANOVIC
Process Engineer, MSChE

 

[LittleInch]

In theory it could be any distance and have any number of LO valves once you take into account the pressure drop at max flow and hence set the PSV at a pressure where the pressure doesn't exceed the rated pressure due to pressure drop at your max flow rate.

In practice and due to the imbalance between consequence (uncontrolled full bore rupture) and cost of moving the PSV, most operations would move the PSV and eliminate any valves between the pressure break and the PSV.

what sort of safety studies have been done? HAZOP, HAZID, LOPA, SIL?? These could provide the backup seemingly needed for such an issue. what other shutdown systems exist to protect the d/s pipework? Are they SIL rated?

My basic philosophy has always been that if there is a valve, then someone can open or close it. LO and procedures, even removing the handwheel to my mind count for nought compared to the potential consequence. no valve - can't do anything - simple.

At the 600 / 150 break, the 150 pipe / valve/ flange could easily rupture if exposed to the sorts of pressures you get.

This might sound odd but how "sophisticated" is the client and his team. Do they have safety standards?, are they part of a bigger group / organisation which will have some safety standards to maintain?

I would stick to your thoughts which without any further information look good to me.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TD2K]

Simply locating the PSV 1600 ft downstream from the pressure reducing valve in itself wouldn't automatically be a no go in my opinion. I would want to look at the hydraulics and satisfy myself that the built-up backpressure at the spec change when the PSV is at relieving pressure is within the capacity of the class 150 piping.

Over and above that, Littleinch has nicely summarized some of the points to think about. If you did have a leak/fire/explosion, is this near people, out in the middle of nowhere, adjacent to water? All could be factors to consider.

 

[bimr]

Have you asked why the PSV was located in that spot? Among the reasons why safety valves are installed is to protect environmentally sensitive areas and high populated areas.

 

[txplpro]

Thank you for the replies, Emmanueltop, LittleInch, TD2K, and BIMR.

To Littleinch, check the edited first post.
In reply to the latest response-the client did not share the HAZOP report with us.

To TD2K, the question to you is, how can the piping between the PCV and PSV be protected from overpressure if the PSV will not see overpressure until overpressure crude has travelled 1600'? I claim it is impossible for the PSV to protect upstream piping close to the PCV under this scenario. By the time the PSV reacts and unloads to 110% of design pressure, the piping adjacent to the PCV is above the class limit. Assume the PCV fail mode will be a quick fail 100% open.
Also assume the client will be able to lower the PSV setpoint to Class 150 limits from the current Class 300 system it is currently installed in, or change to Class 300 piping after the PCV.

BIMR-I have no transient repsonse modeling experience, however, I do know that the PSV is not a fast opening Hi-flo Dan-Flo type valve. To prevent overpressure 1600 ft. upstream of the valve the contents at overpressure must be relieved sufficiently fast at the PSV location to prevent an overpressure at the PCV, while the pipeline is operating at 760kbpd. Also bear in mind, the existing PSVs are (2) 24" connected on a 24" manifold downstream of 208 lf of header pipe in the tank terminal-a flow restriction that adds even more time to relieving the overpressure. Translation-"Ain't gonna happen."

 

[bimr]

Regarding the "By the time the PSV reacts", have you calculated this? It will be almost instantaneous.

The pressure wave travels down the pipe at the speed of sound in the liquid. If the liquid is water, rigid pipe and ambient temperature, the wave velocity is 4,720 ft/sec

 

[LittleInch]

Sounds like a mess to me. You're in an awkward place from what you've said so far. I would use the company standards and say it doesn't comply. What did the hazop say about this issue our did it ignore it?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

It appears you have done the pressure drop checks to confirm that the built up backpressure d/s the PCV is in excess of the piping class used at that coicident temp for the max relief flowrate, which is 760kbd? If the relief flow you used for this calc is >760kbd, how did you derive this higher flowrate? Is this feed from some upstream pump?

This is a tie in to a tank farm - to clarify, pls explain why this is a PCV and not an LCV? This line feeds into a storage tank? As suggested, pls post a sketch of source feeds and destination with process controls.

It is not advisable / standard operating practice to lock open (frequently operated) operational valves that may be in the path from the PCV to the PSV - ok for (lower frequency of operation) maintenance isolation valves. This is because the permit to work restrictions on operational valves that are locked open become so inconvenient that Operations will ultimately undo the lock open feature on the valve.

If the lock open valves are all maintenance isolation valves, then one solution would be to upgrade some upstream portion of the 1600ft pipe length to 300lb.

The client, on the other hand, may perhaps be reluctant to shift the PSV further upstream because of the increased backpressure that would be developed on the PSV exit piping that may compromise the PSV capacity and / or its setpoint?? If this were to be the case, changing the type of PSV (to tolerate higher exit built up backpressure) may help to persuade them to shift the PSV upstream to make things safer?

Knowing details of the client's perpective on things may help to bring about a solution that is accepted by both parties?

All this is only to do with the ultimate overpressure protection integrity - what about the instrumented PSHH trip - is there one? Is liquid hammer possible on this line?

 

[TD2K]

< I claim it is impossible for the PSV to protect upstream piping close to the PCV under this scenario. >

So 1600 ft is out. OK. What about 800 ft? What about 100 ft? What about 10 ft? What about 6 inches? Where's the magical cut-off that means the upstream section of piping is now protected?

If something downstream of the PSV closes or starts to close, the flow will be restricted. Pressure will start to rise. At some point, the PSV will start to relieve. Now you have flow going through the PSV that you've calculated is sufficient at some set pressure and overpressure to protect that section of the pipeline (or someone has, hopefully). To maintain that flow to the PSV, you have hydraulic losses through the pipe which sets the upstream pressure which you have to evaluate against the maximum allowable pressure for the lower pressure rated section of piping. Now, maybe it won't work but you've provided very little specifics about the system, the cause(s) of the overpressure, allowable overpressures, etc. On the other hand, maybe the event you are concerned with is water hammer. That's another issue and typically, PSVs don't react fast enough to address water hammer. Now, if it's a thermal relief PSV, 1600 ft is not an issue.

I'm not saying 1600 ft is acceptable but I wouldn't rule it out without some numbers behind me.

< I don't know what the setpoint is >

That's sort of key.

Your velocity is in the area of 7 ft/sec. In a 36" pipeline, I wouldn't expect to see a huge dP over 1600 ft at the design flow of 760 kBPD. What's the relief flow?

I'm not saying your concern is misplaced. A class 600/150 spec break should always raise the question of overpressure.

 

[txplpro]

Thanks for the continued or first replies TD2K, georgeverghese, LittleInch, and BIMR.

TD2K and BIMR-I have not calculated anything, nor has the client provided the PSV setting or relief capacity. We are at the client's 60% design stage, still dealing with stuff at high level.

There is no 'magic' distance that will make the PSV operate fast enough. But you can say with certainty that the pressure upstream of the PSV will be higher than the pressure downstream of it. If the entire Class 150 system is downstream of the PSV and open to tankage, you have placed it at the most beneficial location.

The issue I am discussing is how the Class 150 piping is adequately protected from overpressure if the PCV fails wide open or just "too open" and allows the pressure to increase beyond Class 150 limits.

My claim is that even if the PSV is set at the correct pressure of 275 psig or less by the co. standards, there is too much distance to 'catch' the overpressure before it is too late and damage is done to the Class 150 piping system. That is assuming the one or more isolation valves are not closed and block off the PSVs from the Class 150 piping.

George-I raised the concern about locked open valves with the client already. To them, it is not an issue. It seems penny wise and pound foolish to me to rely on the PSVs downstream of three isolation valves, but it is not my plant, and they can claim the valves will always be locked open.

The controls system does monitor pressure downstream of the PCV. If an overpressure due to a wide open PCV is detected, there is still a lag time of closing the 36" ESD valve upstream of the PCV to deal with. There is no way the ESD can react quickly enough to prevent overpressure. Pipeline valves are intentionally slow moving to prevent surge on the pipeline.

I added a simple sketch with many missing details that shows the points of concern for overpressure.

 

[LittleInch]

txplpro,

Not taking back anything I've already said, but I think you're too worried about the extra pressure on the #150 pipework. I assume this is B 31.3 pipe which allows quite a large margin of overpressure for short term events ( I also assume these PSVs rarely operate).

The key point is what these PSVs are set at and their capacity. They need to be able to relieve at the maximum incoming flow to avoid pressure accumulation which could be quite severe if the ESD valve takes a while to close.

Yes the PSVs are not in an optimum place and you should continue to state that and recommend new ones or moving these ones, but in reality they will probably do the job. I personally think the risk they won't is much higher than the cost of moving them, but if you're up against a brick wall with the client all you can do is say move them or walk away.

I'm not sure of your position in the project / company, but this sounds like a good time to put it all down in writing noting where the design doesn't comply with good practice and the company procedures and hand it on to someone higher up the food chain to sort out. Often putting things in writing can be the only way to resolve things as then there is a smoking e-mail which needs to be resolved.

you might need the assistance of some safety engineers either in your company or the clients to point out the error of their ways.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

The process control concept here appears to be that the PIC - PCV restricts flow through this line to be within the permissible backpressure possible on this limiting 150lb piping class.

There is a pressure surge valve upstream of the PCV, so there must be some relief disposal vessel at the PCV end to allow for the PSV to be relocated with minimal exit piping length leading to this disposal tank?

A 36inch trip SDV closing time would be too long to enable quick isolation for a PCV failure case, thus making the PSHH trip loop somewhat useless. Presume this PCV is a fail close design.

Perhaps the PSV ( in its current location) has been set low enough to cater to the max permissible pressure on this 150lb segment at max flow - this needs to be checked. This may be the only reason to justify the PSV in this location, assuming the plant owner is confident of their operating skills in managing permit to work access to these 3 intervening valves. This pressure drop calc should obviously be done with crude at max normal viscosity.