7 replies to this topic

[JoeWong]

Dear All,

I have a stabilised condensate transfer system which consist of a pump, 0.9m (ID) x 1.2m (T/T) filter and 100 m long 10" piping. Entire system is fully rated for pump shut-in pressure, aboveground and expose to solar heating. The filter is designed according to ASME code and a PSV is provided to protect against fire (as well as hydraulic expansion due to solar heating). A DN20xDN25 TSV is provided on piping to protect piping from hydraulic expansion due to solar heating.

Are there any issue with above system from over-protection perspective ? Personally i didn't see any. Your comments are welcome.


I further analyse the system, i found that the filter wetted area and trapped volume is much lower than the 10" piping.

i) Why the filter is protected from fire but not the piping ? Isn't the piping with higher inventory and higher wetted surface area is more hazardous than the filter ?

ii) Should we calculate the fire relief area for piping and ensure total TSV area is more than the calculate fire relief area ?

iii) Is single TSV is sufficient for any piping length ?



Appreciate you advice.

regards,

JoeWong

[pleckner]

Overprotection? Well that depends. If the filter can be blocked-in then it would need its own relief valve. If not, having a single relief valve on the piping would be acceptable if sized properly.

i) Why the filter is protected from fire but not the piping ? Isn't the piping with higher inventory and higher wetted surface area is more hazardous than the filter ?

Good question. Pipelines fall under ASME B31.x, not ASME Section VIII divisions but the filter falls under ASME Section VIII divisions. Within processing units, pipes should most definitely be included in the fire calculation but because their wetted area is typically significantly smaller than the process equipment they are attached to, they are often ignored. I'm not a fan of this and feel we should take them into account. You can always bump up the wetted surface area of the process equipment by some factor to try to account for the piping.

For long pipelines, I would think their wetted surface area contributes the bulk of the relief area so yes, I think you need to perform the calculation as if the piping were the process vessel. But, you also must consider the fire zone. Can I leak at a flange? Yes so the fire is within the flange area unless pooling allows the fire to travel a large distance. Obviously this is not cut and dry. Why this particular pipeline is only protected by a basic DIN 20 x DIN 25 can only be answered by the designer(s). Maybe they know something I don't.

One last thing, let's not try to categorize one thing from being more hazardous than another. They are all bad.

ii) Should we calculate the fire relief area for piping and ensure total TSV area is more than the calculate fire relief area ?

I don't understand this question. We always make sure the actual relief valve area is equal to or greater than the worst case relieving scenario required area.

By the way, you happen to hit a nerve with the term "TSV". I will argue that there is no such thing as a temperature safety or relief valve. These devices work on pressure, not temperature. They do not work with thermal couples or bimetallic strips or the like.

iii) Is single TSV is sufficient for any piping length ?

Ouch, that term again! Any ways, I will have to yield this to someone more knowlegable in pipeline design than myself. I've read where multiple PSVs are used when dealing with long distances but I can't give you any more detail.

[JoeWong]

Phil,
Thanks for your prompt response. You are always most responsive Che Jedi in this particular forum...

Overprotection? Well that depends. If the filter can be blocked-in then it would need its own relief valve. If not, having a single relief valve on the piping would be acceptable if sized properly.

In my opinion, filter will always can be blocked-in UNLESS stringent operating / safety procedure is followed or LO the block valve (which normally exist at inlet & outlet) is provided. Somehow, some may approach it conservatively where it may be blocked-in and followed by fire. It very much subject to safety and operating philosophy.

Good question. Pipelines fall under ASME B31.x, not ASME Section VIII divisions but the filter falls under ASME Section VIII divisions. Within processing units, pipes should most definitely be included in the fire calculation but because their wetted area is typically significantly smaller than the process equipment they are attached to, they are often ignored. I'm not a fan of this and feel we should take them into account. You can always bump up the wetted surface area of the process equipment by some factor to try to account for the piping.

Yeah. I am having similar appraoch then you. I have experienced many times inclusion of piping wetted surface area causing PSV one or two size bigger during detailed design, eventhough i have considered additional 30% margin during Basic design.

For long pipelines, I would think their wetted surface area contributes the bulk of the relief area so yes, I think you need to perform the calculation as if the piping were the process vessel. But, you also must consider the fire zone. Can I leak at a flange? Yes so the fire is within the flange area unless pooling allows the fire to travel a large distance. Obviously this is not cut and dry. Why this particular pipeline is only protected by a basic DIN 20 x DIN 25 can only be answered by the designer(s). Maybe they know something I don't.

I am in agreement with you where piping should be included in estimating the relief area for safety reason, good engineering reason, etc eventhough the code doesn't clearly called for it e.g. ASME B31.X. This is the main idea that i would like to bring-up for discussion.

One last thing, let's not try to categorize one thing from being more hazardous than another. They are all bad.

OK. The intention is only to have sense / feeling of the hazard level.

I don't understand this question. We always make sure the actual relief valve area is equal to or greater than the worst case relieving scenario required area.

This is separate topic in my mind which shouldn't be here. Appologize. Please ignore.

Last, thanks and appreciate.

regards,

JoeWong

[vicini]

My favorite topic, piping and thermal relief (fire/sunshine). I've seen a pipe yard with pumps and filters and headers to these items that looked like a forest with 50 3/4" relief valves sticking up all over the place. The yard was classified as a DOT regulated area and any overpressure system must be inspected and calabrated every 6 months. They had buckets of 3/4" valve to repair and replace the ones in use. Here was my solution:

about 25% on the lines came into the yard from underground. The underground pipe is a heat sink and the sunshine or fire heat transfered will just thermal syphon the aboveground and below ground pipe. Removed all those 3/4" valve. Where we had valves that could isolate filters and other ASME vessel from piping, we put a 3/8" check valve around the block valves so that the piping could relieve pressure toward either a vessel or a piece of pipe that had a relief valve on it for either thermal or process upset. If the line was heated up, the the pressure in the line would exceed the pressure on the other side of the shut valve and eneough fluid would leave to equalize the pressure. If the pressure in the line or vessel exceed MAOP, the 3/4" relief or its large process relief valve would take care of the situation.

In one case, we had Butadiene (bad stuff) and I removed all the 3/4" thermal reliefs on the meter piping at a remote station and just used check valves back to the 3 mile long buried pipeline. At the other end there was a flare that could incineate the stuff and we had no way of venting it at the remote site.

[pleckner]

vicini,

I'm having a difficult time picturing what you are doing with check valves. Can you sketch out something and post?

I think it would be educational for all interested.

Thanks.

[vicini]

In the top drawing PSV 4,9,10,&11 are thermal relief valves on above ground 8" lines, each section is 50 feet. If v 1,2,6,7 or 8 are shut and thermal heat causes the liquid to expand, then one of those PVS's will vent to the atmosphere or a very large relief header system.

Below v 25, 26, 27 are added. These are 3/8" tubing check valves and 3/8" tubing lines in and out. If the thermal heat cause the pressure in a section of pipe to rise, the pressure will increase and the check valve will open and relieve the pressure into either the pipe lines that are underground and connected to a huge heat sink, the earth. Check valve v26 will put liquid into the vessel and v5 or 16 will vent to atmosphere, protecting both the one piece of pipe and the vessel.

There are other ways to set it up. All the check valves could relief backwards toward the line coming in or forward. The descision is based on meter security (you could end up with a flow path unmetered).

If you could imagine this set up reated 8 times at a pipeline terminal with each pump having 8 suction lines and 8 discharge lines.

Attached Files

•  psv_vs_check_v.xls   149.5KB   156 downloads

[JoeWong]

vicini,

Looks a good idea to minimise PSVs. However, have you or your team ever addressed the following ?


i) Check valve P28 continuosly in stagnant position, can it be rusted or corrodant / solid causing check valve stucked ?

Infact there are still argueable if the check valve internals form part of obstrucle in the line which potentially block the line e.g. can we put a check valve on the PSV inlet ?

ii) V15 is provided for E-5 isolation during maintenance. Check valve P-30 may defeat the isolation purpose. Similarly for V-19 and V22.



regards,

JoeWong

[pleckner]

Vicini,

Joe is being too kind! You can't do what you are doing (I hope this is only a proposal and not in service)?

Check valves do fail and leak. I agree with Joe, the check valve can actually be a source of trouble by itself.

Joe pointed out the isolation of E-5 and what about E-4? You have no positive isolation with this installation! Actually, you don't need P-30, or is it V-36, can't tell?

I would need to reread some of B31.1 but I think you are violating the letter, if not the spirit of the Code with this installation.

Please rethink your position for your plant's sake, for your maintenance people's sake. I wouldn't want to be in your shoes if something happens.

Has this been HAZOPed?