Pressure Release Valve Spec?

Go look at a water heater somewhere; see who manufactured the relief valve; call them.

There are likely codes to cover the conditions you describe.

There are also code enforcer enforcement officers to cover the work performed by untrained, unqualified personnel who illegally install such equipment.

That's you.

Consult a QUALIFIED, LICENSED plumbing contractor.

Or, hire a plumber.

My thoughts, not knowing if this relief action is routine or only in emergency, are towards inline rupture disks which are manufactured in a variety of materials.

It's for an emergency situation, not a routine pop-off. and good idea with the rupture disks, I'll see if I can make that applicable for this

I'm also not the one installing it...just doing the engineering legwork of sourcing multiple options to present to the staff and then move forward.

It is good to know that you are concerned for safety. The operator of a reactor at a chemical plant took it upon himself to replace a rupture disk with a steel plate when he could not get the proper disks in time. The explosion had a shockwave that ripped a chain-link fence out of the ground and broke windows in a 5 mile radius.

You said, "just doing the engineering legwork of sourcing multiple options".

If you are presenting, "multiple options" "to the staff" under the guise of an engineering study, you are a fraud!

This is a personal safety (as in life-or-death) situation, and you should be fired immediately.

You are taking the advice of total strangers and presenting it as a valid "engineering" study.

Get a life, before you kill innocent people who may trust you.

Loser!

Most reactors in chemical plants and other installations use BOTH a pressure relief valve (prv) and a rupture disk. The valve is rated and sized according to the normal relief pressure. This relief valve is used as a safety device and not a normal operating device. In conjunction with this is a rupture disk. It is mounted on a separated pipe as close to the reactor as possible and vented to an appropriate location. The pressure rupture point for it is slightly higher than the pressure relief valve. This way if the pressure relief valve fails closed or its piping becomes blocked the rupture disk will break prior to an explosive condition existing. Also in cases where there is a rapid increase in pressure it will relieve the pressure faster in parallel with the prv.

For example: reactor rated at test point of 200psi; reactor rated at maximum operating pressure of 120psi; normal working pressure of 80psi; relief valve rated at "critical point" of reaction at 100psi; rupture disk rated at 105psi. If the reaction reaches 100psi for any reason the safety opens. If safety doesn't relieve then rupture disk ruptures at 105psi. reactor internal pressure now 0 psi. These are all theoretical points used only for an example. Sometimes if the reaction pressure is not critical the rupture disc is rated at 90-100% of maximum operating pressure rating of the reactor.

Good Luck, Old Salt

Just a talking point....everything you've written is correct, but I wanted the OP to do some research of his own rather than just give answers. I know us grizzled vets like to do that (me too). I'd prefer we give him some reference points and let him complete his on-the-job homework himself. Just like we'd do with any student.

I agree with you that the first hand, hands on method of research is the best way to learn. No one was every kind enough to give me the answers, except in sand-box in Kindergarten grade (I "A"ced it!).

My reasons were two fold: 1) it seemed that the terminology for the different devices and parts of the system were getting confusing. Some posts seemed to call the same devices different names. (It is the same whether you call it pop or soda). 2) Unless I am mistaken many submitters mentioned PRV's (is it a pressure relief valve or a pressure reducing valve?) few submissions indicated that what he is looking for is a system and not a single device.

Perhaps I "over provided" the OP with information about a system but-- it was intended to get him started developing a system and not get stuck on one device. (Can't see the forest for the trees). In my other posts I did tell him to Google several times.

If I compromised his development in engineering I apologize.

Good Luck, Old Salt

I am an engineer therefore I do not do dental work. I am not a heart surgeon therefore I do not do open heart surgery. I am certainly not qualified to do these things.

You are not an engineer therefore you should not attempt to do engineer work. You are certainly not qualified to do these things you are attempting to do.

Remember the Titanic? Remember the Hindenburg? Remember the Challenger? Remember the Columbia Shuttle? Remember the Sultana? Remember the Piper Alpha? Remember the Exxon Valdez? Remember the Deepwater Horizon? Remember TWA Flight 800? Don't add to the list by attempting to do something, potentially fatal to others and yourself, that you don't know anything about.

Besides how are you going to tell if some of these answers might be "bogus" answers?

Good Luck, Old Salt

I did work similar to this several times in the last few years. Here is the process I followed:

1. Remember these 4 words: Google Is Your Friend. Seriously, the Internet is free and best research tool you could ask for. Learn to use it well.

2. There are a number of highly technical papers written by knowledgeable authors which describe the process of specifying a PRV. I found half a dozen in less than an hour's worth of searching. Mostly they cover sizing and placement based on design conditions, but there are plenty of chemical/material compatibility charts available too.

3. Once you have done your homework, contact several valve manufacturers for recommendations. When you understand the basics of the topic, you will be able to understand what the manufacturers reps say, and you will be able to make informed decisions.

4. You mentioned a relief pressure of 100 psig, so I assume the working pressure of the tanks and piping exceeds 15 psig. Therefore, they fall under the auspices of ASME Section VIII Pressure Vessel Code and ANSI B31 Pressure Piping. You and/or the "engineering staff" should have access to these documents. Yes, they are expensive; yes, they are a necessary business expense. And yes, they are much much cheaper than the legal liability consequences of poor design. Even with recommendations from manufacturers, you are not absolved of legal liability.

5. Since ASME Section VIII applies, you will need your system to be inspected and code-stamped. Have you made provisions for this?

6. Lastly, and this is not meant to be snarky, I assume you are not an engineer and have no expertise in this area. If so, why are you conducting this "engineering study" rather than the engineering staff? Pressure vessel safety is a serious issue and if you or the engineering staff lack expertise in this area, you should be consulting with people who do have expertise.

everyone, this was not meant to get blown out of proportion, nor am I in any way shape or form intending on just slapping some valves on stuff and winging it

I am a junior engineer at a company and a mentor gave me this research project to just see what I could come up with to basically test my resourcefulness. anyhow, any final decisions will be fully vetted and made sure to comply with code by senior PE's.

*deep breath*

ok...after the OP, I did find the ASME VIII info and also found some hastelloy-C alloys that seem to stand up decently to exposure to both chemicals. I am in touch with several valve manufacturers about sourcing something properly rated/sized

the only part I am having difficulty in finding is if there is a min/max pipe length allowed for the section between the tank and the PRV itself

thanks for the comments everyone, sorry if the original wording was ambiguous or seemed shady

I still suggest you google those technical papers I talked about in re: piping length. I don't remember any specific requirement but it is common practice to put a relief valve in piping somewhat removed from the vessel. I would expect it has to do with the response time of the valve and the margin of safety between the valve set pressure and the vessel burst pressure. A pressure wave would take some time to travel down the pipe to the valve in order to trip it, and it matters how long it would take to relieve the pressure to a safe level.

We didn't mean to be snarky with you, but all too often we get posts from unqualified people with questions that are very basic. The fact that they can't figure out engineering basics concerns us and sometimes we get carried away. Sometimes I get scared that unqualified people are doing work they are not suited for, and potentially dangerous.

Google "knock out tank". That is how many chem plants get the pressure away from the press. valve but have almost immediate response. Google more, life is easier that way. Good Luck, Old Salt

If it's an "ordinary" relieving situation i.e. gradual build-up to relief pressure, length of pipe up and downstream isn't a problem. You need to establish design maximum relief flow, and calculate pipe losses. Make sure losses are allowed for in the valve sizing. If the pipes are long you might need to try a couple of options to get optimum combination - small dia pipes could mean bigger valve and vice versa.

Also as the vessels are drained down periodically for regen it's worth looking at pipe routing. Upstream of valve - do you want the pipe to stay full of liquid or drain back into the vessel? If the latter, in service it's likely to be part full of pressurised air. Is this a problem? Downstream of the PRV it's usually best for the pipe to drain to the discharge point.

For Relief on Process Plant API 520 / 521 are the generally accepted standards.

The inlet losses in these standards are restricted to 3% of set pressure.

ie You do you PRV sizing calculation; find the next largest standard size PRV and calculate the rated flow (the maximum flow that the PRV can pass); then do a pressure drop calculation between the vessel and the PRV for the line size selected.

Remember to be aware of any static head losses as the PRV is usually (>99%) above the vessel.

The line size can NOT EVER be smaller than the inlet flange on the PRV

If your question is how far away can / should the PRV be from the tank, then my answer is:

• The PRV should be mounted directly (within 24") off the tank with a flanged coupling.
• Any length of piping should be after the PRV and lead to a capture tank.
• Remember that there is friction in the piping and the PRV is for an emergency unload of pressure.
• You want the tank to depressurize as fast as possible so size the PRV and piping to be of equal size as the inlet.

This is just my humble opinion and I hope it helps.

I think there are a number of conflicting comments here which I think can cause confusion as I sometimes agree with the statements and sometimes not.

My experience is as a chemical engineer with process design contractors mainly in oil and gas facilities - which does include water systems.

You describe the equipment in your original post as both a vessel and a tank.

In common usage the two are interchangeable but typically I would use vessel for an ASME VIII item designed for a pressure greater than 50 psig (3.5barg) and tank for an item designed for inches water gauge to API 650 / 2000. So your water filters would be vessels.

These distinct codes will have different (although in many cases overlapping) requirements for relief protection.

You have not specified if the vessels are water filled or if there is a vapour space in them as this has an impact on the relief cases and maybe on the allowable connections.

Generally my experience is that ASME VIII vessels requiring API 520/521 relief valves end up with flanged connection valves as per API 526. In looking up a vendor you can get ASME VIII boiler code screwed relief valves but the most common use of screwed valves that I have used is for thermal expansion cases ONLY.

My experience is that bursting disks are not as frequently used as others here have seen. The limitations that they have is that they do not close again after the relief event and so the plant has to be shut down purged etc for their replacement. Note that the cost of a plant not operating is usually huge. For me bursting disks are used in there is a risk of an explosive reaction or on heat exchangers where there is a risk of high pressure gas from a burst tube flowing into a liquid filled shell side. Conventional logic is that relief valves do not open quickly enough for this high pressure spike. For me a filter is not a reactor and I would be surprised to see a bursting disk on a filter.

We also seem to have the old mixture of thoughts on PRV - which you specify as Pressure Relief Valve (also known as PSV - Pressure Safety Valve) as opposed to the also used Pressure Reducing Valve.

A Pressure Relief Valve will NOT reduce the pressure in an emergency situation. They are designed to open at the design pressure of the vessel (usually, sometimes a lower pressure) and vent the fluid build up that has caused the pressure excursion. When the pressure has fallen the spring in the valve causes the valve to reseat and seal and it is then possible for operations to continue uninterrupted although if you didn't carry out investigations as why the PRV lifted you would be in dereliction of your duty

In the areas in which I am familiar a pressure reducing valve is a control device that holds a fixed pressure upstream and a lower pressure downstream so it could be considered that they do not reduce the pressure in the vessel either.

Blow down valves are provided on most vessels to prevent their failure in a fire case. These DO reduce the pressure by failing open and venting the contents of the vessel. These have an air bottle holding a sprung loaded valve closed. The air is vented the spring opens the valve which then cannot close.

It is CRITICAL to properly size the lines into and out of the relief valve. The 3% rule mentioned above is designed to stop valve chattering. In my understanding of the code you cannot have a smaller line and a bigger valve. At the instant before the PRV opens the system is at the design pressure and there is no flow so the vessel and the PRV see the same pressure. As the PRV opens there is flow down the line. The pressure in the vessel will stay constant and so for flow to occur the pressure at the PRV will fall - if it falls too far due to high pressure drop because of too small piping then the valve will close. This process then cycles sometimes rapidly possibly causing the fatigue failure of the PRV.

Downstream of the PRV the line must be properly sized so that the pressure drop from the PRV to the discharge point does not cause a back pressure on the PRV that reduces its capacity.

If you are specifying a relief valve in line with API 520 / 521 then the piping configuration has no impact on the required valve size.

For atmospheric tanks due to the very low set pressures PRVs are located on top of the tank and a very short distance away. However for pressure vessels this is not always the case and it depends on the disposal method for the relieving fluids. If they are to go to a flare header (unlikely for liquids but if the relief case causes a vapour release then possible) the flare header is usually on the high point of the pipe rack and the PRV has to be above the flare header. In this case the filter could be at grade (ground level) and the PRV on a platform at 7m.

Discharge piping is almost always larger than inlet piping. This is usually because there is a larger allowable pressure drop for the inlet side than on the outlet and it is quite common for both sizes to be larger than the flanges on the PRV. This is more so with gases then liquids as the change in pressure causes a change in density for gases that you do not get with liquids (unless the liquid flashes in which case you need to carry out HEN Method calculations). This is why I asked earlier if the filter had a vapour space.

"6 tanks see 6%HCL"

"11 tanks see 4%NAOH."