Safety Relief Valve Lifting

[JPiii]

I am at a plant working on a few product transfer lines. I transfer from building to building crossing out side containment areas. As part of an environmental project, I had automatic valves installed at each building. Before the operator transfers, he must test the line for leaks by charging it for 5 mins. He shuts one valve, pumps until it is full, then shuts the second valve. During our PHA it was determined that we needed Safety Relief Valves on each line. I had these installed which are set to discharge at 165 psi. There are five lines total, of which one relief valve keeps lifting slightly only discharging about a quart of product. I thought it was a bad relief valve so i swapped it for another. That one again lifted. It seems to only be happening in the very beginning of the charge. I removed the failing relief valve and installed a tell-tale pressure indicator on the same flange. The highest pressure shown on the gauge was 70 PSIG. I do not understand, both relief valves were sized for 165 PSIG. The product temperature is about 62 deg C.

I was thinking I needed an expansion tank to absorb the air that could be trapped in the product transfer line when the operator began charging.

But then why is it not shown on the pressure indicator?

Any Ideas?

 

[BigInch]

May be a rapid transient pressure response, but that is difficult to discern without much more info.
What product is this?
Is there air in the line, or product at vapor pressure?
Does a pump start in order to pressure these lines?
Does a valve open quickly to pressure these lines?
Is this the first relief valve after the pump?

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[Artisi]

It's those pesky little transients, and can they be a nuisance. They are usually found in pump systems, electrical switch boards, pressure regulating and control systems etc. - nearly always in the most critical applications but never found in applications where they wouldn't be a problem .

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[JPiii]

This is an acid transfer line which transfer from building to building. Since they are transferring outside during the winter, they blow the lines out with air. This particular acid begins to thicken at 50 deg F and fully freezes at 32 deg F. Even though the line is steam traced, the operators have still had issues with the line plugging so they blow it out.

The pump is used to pressurize the line. The valves that are being opened and closed at the terminal points are air operated and are not that quick but I would say have typical operating speeds.

This relief valve is the one and only one on the line and it just so happens to be the longest line. (Initially I thought the steam tracing was causing the expansion during the 5 min pressurization test, but I no longer think that)

If it is rapid transient pressure response, I am assuming an expansion tank would compensate for this?

 

[Latexman]

What type of pump is it?

Good luck,
Latexman

 

[BigInch]

Try this sequence.
Open the valves, start the pump, clear the line of air.
Close the downstream valve, or if the relief goes off, stop the pump, then close the downstream valve.
If the pump is off, restart the pump.
Bring to pressure. Close the upstream valve. Stop the pump, if you need to (it gets hot, cavitates, or somthing).
Hold for 5 minutes.
Open both valves and start the pump.

Let us know how that went.

Independent events are seldomly independent.

 

[Latexman]

Is the containment of the lines being compromised between transfers, like switching hoses at a hose station or something? Are they 100% hard piped? How frequently are transfers made - 1/day, 1/quarter, etc.?

Good luck,
Latexman

 

[JPiii]

Thanks for the help.

I checked the sequence of operation. It is in fact operating as you suggested. Both valves are open and the pump begins transferring to clear the air in the line. Then the downstream valve shuts followed by the upstream valve.

I do have another hypothesis. On the relief valve that is acting up, we installed a longer leg off of the transfer line and then put the flange for the relief valve. This was done to clear piping in the way. I am assuming this is trapping air that is in the line. As the line fills the pressure increases enough to push the air past the seat of the valve. (I was told that safety valves are rated for air or hydraulics but not both) Since the air is thinner it is able to escape followed by some product until the PRV seats. This seemed logical, but when I called the PRV vendor they told me that if the air pressure was within 10% of the set pressure, then possibly that is what is happening. But they don't believe that it the case here as we only see 70 psig.

What do you guys think? Are there any general rules of thumb on this?

I can send a picture of the install.

 

[chicopee]

Make sure that when the second valve is shutoff that it is done slowlyyyyy, then see if the relief valves discharge.

 

[JPiii]

I thought about that, I will have to check that on Monday. These are all air operated jamesbury valves. I may be able to exercise each one to see how fast they are closing.

 

[BigInch]

When exactly is the overpressure event occuring within that sequence?

Air in the line will dramatically complicate the issues, as volumes expand and contract quickly enough to accelerate liquid in the line and the resulting pressure spikes can become large and fast as they come to a halt, 10% or not.

Independent events are seldomly independent.

 

[Latexman]

A sketch would be useful. With the verbal descriptions, I can visualize multiple piping arrangements.

Good luck,
Latexman

 

[JPiii]

The overpressure is happening just after the first valve shuts. It makes sense that air is trapped in the leg just before the relief valve. Do you believe this is the cause of the problem? Are there any rules of thumb that you know regarding the installation of PRVs? Attached is a quick sketch.

 

[BigInch]

I think it might be the cause. That configuration is nothing but a big air trap.

First valve. (?) I don't know which step, or which valve you mean, in the sequence.

Can you turn the pump off before closing the valve.

Can the 3 foot leg be reduced. That's not helping.

Why does this system need that relief valve? Does the pump have the capability to exceed pipe allowable pressure. Were high transient pressures, requiring a relief valve, expected? I'm having a hard time seeing why it must be there, and located there in the rack.

Independent events are seldomly independent.

 

[Latexman]

"The overpressure is happening just after the first valve shuts." THAT valve is probably closing too fast.

Good luck,
Latexman

 

[BigInch]

I don't know if he's closing valve 1 or 2 first.
Anyway, if he's closing one, or any valve, he could probably turn the pump off first, if it's not off already. I imagine that could fix the problem right there.

Independent events are seldomly independent.

 

[Latexman]

If he has to immediately turn the pump back on after the valve is shut for the "pressure test", why not leave the pump running and close the valve s l o w l y .

Good luck,
Latexman

 

[BigInch]

OK.
It would help me a lot, if we had an actual numbered stepped sequence list that matches a diagram of valves #1 and #2 with the relief event inserted between the proper numbered steps.

Independent events are seldomly independent.

 

[Latexman]

Me too, I don't know which valve "THAT valve" is. I'm going by cause and effect.

Good luck,
Latexman

 

[JPiii]

Sorry for the lack of information. Lets call Air Valve #1 the upsteam valve and Air Valve #2 the downstream valve. Here is the sequence of operation, Both Air Valves #1 & #2 are opened. Pump is turned on to fill the line. After a mass flow meter (after air valve #2) detects flow into the tank, the operator shuts Air Valve #2. When he does this a line test program is initiated. The pump continues to run until a inline pressure indicator reaches a set pressure. Once set pressure is reached, Air Valve #1 is closed. Both valves now remain closed for 5 minutes. If pressure is held, Air Valves #1 and 2 are opened and full transfer occurs.

I already scheduled maintenance to make the changes on the inlet leg of the relief valve for the end of the week. I do not see/hear any hammering but I figured it is worth a shot. I had access to a lift today and was able to throttle the air supply to Air Valve #2 to slow down the closing speed. Since all the equipment is tied into a PLC, the program only allows for 2 seconds for the valve to close before it throws a Valve Failure code. I slowed it down as much as I could without generating a failure. If in fact it did help, I'd have to have the program modified. I should know tomorrow night when the next transfer occurs.

Thanks for your help guys.