Pressure Relief for High Pressure Separator 1

[liecdre97]

I just started with a rupture disc company and we were approached to come up with a pressure relief solution for a High Pressure Separator. This is a relatively new application for my company, so I was hoping to get input on what solutions others have used for this application and their experiences with those solutions (good or bad).

 

[EmmanuelTop]

Separator is just a separator. The "HP" term is common in Oil and Gas Production and gas compression systems.

Dejan IVANOVIC
Process Engineer, MSChE

 

[don1980]

liecdre97 - I know this doesn't make money for your company, but your clients are poorly served when they install disks in applications which are suited for a re-closing device (PSV). Disks cause a lot of avoidable problems (plant reliability, safety, environmental) due to premature failures. Unless the application has a specific need for a disk, it's best to avoid them.

Nothing personal - Good luck in your new job.

 

[liecdre97]

Thank you for your input so far. Have you used a PSV in this specific application? Thank you again for your responses.

 

[don1980]

Yes, I've used PSVs in separator (high and low pressure) services. The only cases in which I've used disks are applications in which the gas (e.g. ethylene) can experience a decomposition reaction. That's an example of a case that requires a disk (PSV won't respond fast enough). A lot of care and attention when into spec'ing those disks because of the risks posed by pre-mature failures. Every specification detail (disk type, burst temp, zero burst tolerance, disk metallurgy,) was scrutinized with the aim to get the premature failure risk as low as possible.

 

[MortenA]

As a matter of fact we dont know exactly what kind of "high pressure" you are talking about, if its less than say 3000 psig then its nothing out of the ordinary. There _may_ be some situations where a bursting disc could be better, but they are rare. If its connected to a flare system your should consider installing two PSV's that each can be isolated (for service).

Best regards, Morten

 

[liecdre97]

Thank you very much for your responses. They are very helpful. The general requirements are as follows (280 bar – 4,000 psig, up to 250 degrees C and 6-inch diameter outlet). Besides speed of response, are there any other limitations that cause a PSV to not be the best solution? For example, is there a diameter, pressure or temperature that make a PSV impractical? Does leakage and/or chattering become an issue? Do you ever use rupture discs in series with PSV’s?

Don1980, with the rupture disc solution you used, did it work well for your application? Any features/attributes you would have changed?

Thank you again for your help.

 

[georgeverghese]

On the other hand, another option to an RD would be a buckling pin operated relief valve, which is quicker to reinstate, has a lower risk of accidental rupture - dont know if these go up to 4000psig/250degF though.
Presume you are aware of the risks associated with installing an RD in series with PSV.

 

[don1980]

liecdre97 - The description "HP separator" is a loose phrase, potentially describing a broad range of specific applications. However, I now think you're probably referring to an application that needs a fast-responding relief device to mitigate a rapid increase in pressure (deflagration, decomposition reaction, etc.). If so, an RD is appropriate. My comment earlier was simply to say that an RD should rarely be the first choice for a relief device. RD's have a broad range of failure modes as compared to PSV's - thus they are an avoidable liability for most general relief applications.

Did the RD work well in the application I described? The answer is that it's tolerable. Given a choice, we'd prefer to not have the RD for the reason I mentioned, but it's necessary due to the explosive rate of pressure rise in this application.

 

[MortenA]

To OP, with regards to size of PSV they mst commenly comes in standard sizes that you can find in API 526. The largest has the "T" designation and has a relief area (theoretical) of 26 sq inch, the smallest is "D" with a relief area of 0.11 sq inch (i dont know what happened with A-C).

280 barg and 250 deg C is a bit out of the ordinary but not "first time". I guess that a #2500 would be able to handle this.

The most common reasn for using a RD in series with a PSV is to save money for the materials of the PSV. If you have an agressive medium you can make the RD relatively inexpensive in the right material and use cheap CS for the PSV.

Best regards, Morten

 

[liecdre97]

Unfortunately, it sounds like the overall experience with rupture discs hasn't been the greatest. My company recently started a dedicated new product development team, of which I am a part, and this is exactly the type of situation we were put together for. Our target is to create the most value for our customers. Our team has a unique way of asking and verifying what is important, what the gap is with the current accepted solutions, and how much of an impact a measured improvement of each attribute can make to the process and organization. We would like to know who, from the production plant side, would need to be present to gain a holistic view of the process/production/safety/potential roadblocks so we can come up with the valuable solutions wanted/needed. We are open to solutions outside of rupture discs. Are you a good place to start? If not, who would you recommend? We simply want to gain a better understanding of the industry and safety relief devices as a whole, so we can create solutions that will take care of customer needs. Thank you very much for all of your help and time.

 

[LittleInch]

liecdre97,

As I think you've probably gathered, the issue with rupture discs / bursting discs is concern over premature / incorrect operation and the lack of re-sealing and time taken to replace a disc after bursting.

RDs are inherently susceptible to this and the issue of production is not a key issue or concern.

RDs are indicated in some applications as noted either high or low pressure, but the key point will always be about risk, consequence and probability.

If there is a very low probability of an event with a high consequence (rupture / over pressure) then you can accept the limitations of a RD. These include the inability to absorb any over pressure, the perception that these discs weaken over time / fatigue or become corroded and also weaken and the time taken to re-seal after operation. That event would normally need to be something that resulted in a very high pressure compared to the MAWP and / or much higher relief flow rate than would otherwise be required. Some examples are listed above but also low pressure shell side of higher pressure tube heat exchangers often utilise RDs for the unlikely event of a tube failure / rupture.

The more "normal" over pressurisation is dealt with by re-sealing devices.

The addition of a RD usually implies greater margins required between Normal OP, PRV relief set points, RD set point and MAWP / design pressure.

I would think it highly unlikely that a RD would be your only form of pressure relief due to the one time action and consequential down time of the system.

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

 

[liecdre97]

LittleInch,

Thank you very much for the detailed information. I would like to ask a few more questions.

1. Are there a good amount of applications where an overpressurization is expected on a semi-regular basis? I know that a reasonable amount of RD's are used to protect valves from the process media. Can I assume that those applications are not expected to overpressure regularly?
2. Many modern RD's actually perform quite well over many years with no real fatigue issues. Also, as stated above, many are used to protect valves from the process media because they can relatively inexpensively be made from corrosion resistant material. Assuming those two concerns are satisfied, would a resealable rupture disc have any advantages over a traditional PSV for most "normal" over pressurization applications? In other words, are there issues with using PSV's, even for "normal" applications, where the attributes of RD's would help?
3. Your last statement regarding the addition of a RD implies a lower operating ratio for RD's than PSV's. What is the normal operating ratio assumed for a PSV? What is the operating ratio you assume when using a RD?

Thank you again for your time.

 

[georgeverghese]

If you are in product development, may be you could try to make a rupture disc to work in a similar way to how a buckling pin operated RV works? That would more or less remove one significant disadvantage to inline RDs'?

 

[LittleInch]

You should never design a vessel to regularly exceed it's MOP to the extent that relief systems are activated. If that happens you've designed it wrong or you're operating it wrong.

what I meant was that the probability of such an event needed to be borne in mind when deciding what type of relief device to use. Very rare, but high consequence / high flow events could indicate use of a RD, but otherwise some sort of re-sealable device is usually indicated. Any design where a RD was to break on anything less than a 5- 10 year basis would not be thought of as a good thing by operators due to the down time to replace it and re-start the system.

I've heard of but not come across the use of an RD in front of a relief valve or other valve, I suspect their use is relatively rare.

By operating ratio I assume you mean set point accuracy.
A "normal" spring relief is something like - 8 to -10% below set point, with overpressure possible up to 5%
A pilot operated one more like 3%

RD something like 3-5%

All vendor dependant and how much you want to pay.

There is no such thing as a re sealable rupture disc - that's a conflict of terms.

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