Temperature/Pressure drop due to wafer check valve

[engr2GW]

Hello,
I have never used a wafer check valve before. My Operators used it for the first time on a line feeding the suction of a compressor for natural gas (not dehydrated yet).

The suction control valve is set at 50 psig (or the suction pressure is 50 psig), this keeps the suction header at 50 psig or a little less.

But the pressure upstream of the check valve is 180 psig, usually, the entire suction header is fairly consistent with the suction control pressure.

upon removing the wafer check valve, I realized that they are usually not full port, a 6" wafer check has a 4" opening, and also, the flapper does not open all the way. While I had doubts as to whether that is sufficient to cause pressure that inconsistence, I observed that it did. Almost like the small opening of (what appears to be less than) about 1.5" - 3" in the 6 inch header was acting a bit like a RO. Granted, the downstream of the check valve is kept at 50psig because of the compressor, but the upstream side(I thought) should be fairly close as long as the check is open, albeit only 1.5 - 3 inches. My question is:

1. Have anyone experienced this before, or is it possible for this to happen or could there be something else (luckily, I have a pressure gage immediately up and downstream of the check valve)?
2. Additional question is: Is there a rule of thumb to estimate temperature drop if you know P1 and P2 across a RO...especially, when, as in this case, the pressure drop is not naturally/entirely because of the RO, but the P2 is forced mechanical to be 50psig by a compressor.

Thanks for your patience and help.

As much as possible, do it right the first time...

 

[danschwind]

Could you post a diagram?

Daniel
Rio de Janeiro - Brazil

 

[engr2GW]

THANKS, HERE'S A DIAGRAM

As much as possible, do it right the first time...

 

[engr2GW]

SORRY, HERE

As much as possible, do it right the first time...

 

[LittleInch]

All check valve can be a lot smaller than the main pipe size.

The are many types of wafer checks. What is yours?

Butterfly checks can be a lot better.

But 180 psi to 50 ?? WOW.

Time for a bigger pipe.

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

 

[danschwind]

Wait...are you measuring the pressure both upstream and downstream the check valve? Or are you measuring this 50 psig downstream of the pressure control valve?

If you are measuring 50 psig downstream of the control valve, why are you associating this pressure drop to the check valve itself? If I understood it right, you have a pipeline (at what pressure?) supplying this receiver (is this a buffer vessel? is there any pressure control here?) that is in turn connected to a pressure control valve that drops the pressure from whatever the pressure is to 50 psig. To know what is the contribution of the check valve itself to the overall system pressure drop, you would need to measure the pressure upstream of the control valve - this is not clear from the diagram.

Daniel
Rio de Janeiro - Brazil

 

[georgeverghese]

Wafer type dual plate check valves are commonly used in the oil gas business. What you describe seems odd - it is not possible to have 130psi dp across such a check valve unless it is grossly undersized or the check valve has jammed nearly closed.

 

[1503-44]

Come on. Its a pressure drop problem.
Why do you not tell us
DIAMETER OF PIPE
FLOW RATE OF GAS

There are typical Cv curves for wafer check valves.
If you know the FLOW RATE, you can calculate the expected pressure drop at the valve.

If you try to flow 100MMCFD through that valve, then yes, maybe it is a high pressure drop, but you do not post sufficient information. Your drawing is nice, but also has no info.

What do you think the pressure drop in this suction line should be.
Post your calculations.
"Do it right the first time."



A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[LittleInch]

"My operators used it for the first time"...

What did they use before?

Why not a proper flanged check valve?

Did no one look at this before it went in the line and think "gosh this looks a bit small"??

But "normally" loosing 20 psi over only 2 miles tells me your line is whistling along.

Temp drop won't be a lot at that sort of pressure. maybe 3-4 C?

Did someone put it in backwards?

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

 

[engr2GW]

Thanks for all the response:

@danschwind: yes, upstream and downstream of the check valve, we just happen to have gauges at those points. the gauge downstream of the check valve (which is upstream of the control valve) is the same as the one downstream of the control valve, i.e., 50 psig. But the one upstream of the check valve is 180psig. I removed the check valve (as stated in the question) and realized that wafer checks are always reduced (pipe is 6", opening is 4"), and does not open all the way (so, kind of smaller than 4" opening).

@LittleInch: I thought about the same thing, but didn't think it will contribute that much, Pipeline and receiver are both 8", compressor suction header is 12", but the line where the check valve is installed between the 8" receiver and 12" header is a 6" line and only about 100 feet. Gas rate is about 4MMCFD

As much as possible, do it right the first time...

 

[danschwind]

engr2GW,

Then you are not measuring the pressure drop caused by the check valve itself, but by the check valve plus the control valve.

I would guess that most of the drop is caused by the control valve. Do you have feedback on the opening% of this valve right now?

Daniel
Rio de Janeiro - Brazil

 

[danschwind]

Or have I again misunderstood and you have 3 gauges total? Could you mark in your diagram exactly where the gauges are?

Daniel
Rio de Janeiro - Brazil

 

[LittleInch]

Well in a 6" line at about 3 barg, that equates to about 20m/sec I think.

So fast but not crazy.

I think something else is going on here.

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

 

[1503-44]

Put in a 6" swing check.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[engr2GW]

@Danschind. I agree, the 50 psig is the responsibility of the control valve to the compressor. The dilemma is, that 50 psi should be able to keep the line upstream of the check valve relatively low too. Hence the suspicion that the check valve and its small opening and flapper is causing that side to stay that high...like a choke nipple or an RO. I have instructed for a second parallel line to be installed to increase flow path (would have simply upsized the check valve, but that will result in a shut down). hopefully, that will increase the interaction between suction control valve to compressor and the segment upstream of the check valve...

Thanks again.

As much as possible, do it right the first time...

 

[1503-44]

Install a 6" full swing check in the parallel line and, at next shutdown, a block valve, locked closed, in place of the malfunctioning wafer.

I suggest a proper Tom Wheatly pipeline check valve, like this one...
And you will also be able to run pigs, if you do the bypass right the first time.

https://www.slb.com/-/media/files/v...r?la=en&hash=5DA0FA932DD832D2C58D909D47B0C220

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[GD2]

engr2GW,
First, Why will you use a check valve in the suction/intel line of the compressor? Second, how can you get 180 psig pressure at the check valve if the source pressure is 60-70 psig?

GDD
Canada

 

[georgeverghese]

"yes, upstream and downstream of the check valve, we just happen to have gauges at those points. the gauge downstream of the check valve (which is upstream of the control valve) is the same as the one downstream of the control valve, i.e., 50 psig. But the one upstream of the check valve is 180psig."

So there are only 2 guages, one upstream of the check , and one downstream of the control valve. And the check valve is bolted onto the flanges of the control valve. I dont see anything amiss here, the control valve is doing what it should be, and the check valve is not jammed.

The PIC here is most likely full open normally when suction flow is within compressor capacity, and the control valve closes when feed flow exceeds compressor capacity, backing up pressure upstream of the control valve to maintain compressor suction pressure at no more than 50psig. While pressure upstream of the compressor could go up to as high as 180psig, which may be the setpoint of some upstream PIC not shown on your sketch.

 

[1503-44]

Oh. So that's what,

"The gauge downstream of the check valve (which is upstream of the control valve) is the same as the one downstream of the control valve, i.e., 50 psig. But the one upstream of the check valve is 180psig. I removed the check valve ... "

means. That confused me. Thanks George.

engr2GW, If you can't write better descriptions, then please make essential info and diagrams. I should have known to avoid trying to answer questions that do not give essential info. Shows lack of

So I do agree with George. Forget the check valve stuff.



A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[engr2GW]

Thanks for all the post.
The suspicion was correct. we put another line of just 4" parallel to the existing one with a regular full check valve and the 50psig can not be seen upstream of both the new and existing problematic check valve. Therefore, the check valve had too small opening and flapper wasn't opening well, hence the restriction and resulting DP.

As much as possible, do it right the first time...