ESDV Closing Speed

[maddocks]

This has come up a few times in the past but I think it's worth throwing it back out for consideration. Does anyone know of a criteria or spec for Emergency Shutdown Valve (ESDV) closing speed? I'm aware of the concerns with respect to hydraulic surge if the valve is located on a liquid line.

However, for the most part, our work involves ESD valves in gas service. Any ideas on closing speed criteria?

 

[ione]

The ESD valve should be coupled with a surge relief valve.
The surge relief valve could be operated in parallel with the ESD valve in order to redirect the stream flow in a closed loop circuit until a pressure decay is detected. It is understood that a transient regime analysis have to be conducted in order to properly size a surge relief valve capable to handle the usteady flow conditions.

 

[dcasto]

Its all up to you and your risk assesment. Some piepline companies insist that compressor stations must be 100% isolated and 100% blowndown to under 15 psig in under 60 seconds. When pressed why? They all respond, its what our design manual says to do.

Think about a 10" line filled with liquid propane and you shut the valve slow enough not to cause hammer (or you have a relief valve around the ESD). How much energy escapes downstream in BTU's? Then look at your 10" line with nat gas closeing just as slow, how many BTUs of fuel would be added to the fire?

 

[maddocks]

That's an interesting point. We're going to retrofit most of the critical inlet and outlet ESD valves with quick exhausts so they dump the air fast, but a 16" ESDV doesn't close very fast no matter what you do.

After that, depending on shutdown mode, we either block and hold, or dump the plant to 100 Psig in less than 15 minutes. I guess I could do a quick calc for the volume of gas moved through the ESD valve as it starts to close and assume a linear Cv reduction over a 20-30 second period. Given the volume of gas and propane contained in the plant, I'm not sure it's going to amount to very much.

 

[dcasto]

Here are some interesting facts about that 60 second closing. Have you ever seen a 8" ballvalve open up to atmos with 900 psig behind it? Every year they would test their system and clock the blowdown.

At one site, they did the test. It only cost them about $50,000. The next year, the cost was even more. Here is what happened. All the lube oil from the compressors was laying in the headers. When they blew the gas down, all that lube oil sprayed the landowners property, cost to cleanup, $50,000. The next year they did the same thing. This time the landowner was going to sue and cancle the surface lease. I was called in and we installed 1" RO under all the 8" ball valves and piped them to a KO drum and a large single vent.

The next year it took closer to 15 minutes, but there was no oil slick.

A couple of years later I was at a different station on the day they were doing the same test. My personal car was covered with oil and the hyway outside the fence was so slick, I spun out. They had to have the road shutdown until they steam cleaned it. Needless to say, in went the RO's.

 

[MortenA]

since its a closing speed you are requesting, and i assume its a ball valve on a liquid line: I very often use 1 sec/inch valve size. So an 8" valve would be 8 seconds closing - if i know nothing else.

Best regards

Morten

 

[maddocks]

This seems to be a commonly accepted closing rate. I just can't find any standard or spec that outlines this. Not knowing any more, I'll likely defer to this rate. At least it sounds pseudo-technical enough and I've heard this verbally from an ESDV supplier as his criteria.

 

[JLSeagull]

For offshore platforms in the United States, API RP 14C C.2.1.4, Pneumatic Supply and Response, requires pneumatic devices to accomplish a platform shutdown within 45-seconds. The emergency shutdown system safety tests require recording the time after operating a manual remote station for a flowline survace valve to close. I believe that 45-second shutdown requirement may exist somewhere within the BOEMRE 30 CFR 250 documents.

 

[MortenA]

14C now also has an ISO equivalent ISO 10418:2003(E) - i assume c2.1.4 refers to annex C - however this does not contain a section 2.1.4 - but i cant find your reference anyway. Could you help me with a section title?

Best regards

Morten

PS: most valves are smaller than 45 - so a 24" closing in 24 secs would comply - but could have a larger surge than a 45 sec closing time

Best regards

Morten

 

[JLSeagull]

My copy of API RP 14C is the Seventh Edition, March 2001.
The paragraph that mentions the 45 second requirement is in Appendix C - Support Systems. The paragraph begins on page 65, C.2.1.4 Supply and Response. The last sentence on page 65 continues on page 66 with the 45 second requirement.

“The time it takes for any safety device (e.g. PSH, BSL, ESD station, etc.) to effect component or platform shutdown should not exceed4 5 seconds. To achieve this response, consideration should be given to pneumatic line sizes, safety device bleed port size and the use of auxiliary quick bleed devices. Pneumatic lines that supply and bleed should be sized for optimum bleed conditions. Because of volume and flow characteristics a line that is either too large or too small will require excessive time to bleed.”

The next sentence applies an obsolete API Recommended Practice 550 to be used as a guide when designing instrument and control systems. Instead the paragraph should reference 551-554, etc.

 

[MortenA]

I think they 10418 and 14C now follows each other. I only have access to 10418 - and it dosnt mention the 45 sec. At least i think so - tried to search the PFD version for "45" althrough the document and only some references came up in the "bypass" came up.

But latest 14C is form 2003?

Anyway - its a fine number Too fast closing time may also sometimes give you problems when considering surge etc.

Best regards

 

[maddocks]

Thanks to all who posted - I'm going to stick with the 1 second per inch of valve diameter - for most of the lines, this is 8 seconds up to 16 seconds.

Most valves will require quick exhausts to dump this much air quickly, but we'll test and record closing times as part of the commissioning process and I'll put a note in the manual requiring yearly confirmation of closing times.