Well Integrity Technical Section

Good afternoon team,

Hope everyone is fine. During pressure test of annulus gate valves with positive pressure at 1500 psi and it 15 minutes it shows 100 psi drop. I was searching for the acceptable passing criteria as per API both for positive pressure test with liquid and negative pressure tests for gate valves.

Appreciate your continued support.

Thanks and Regards

Ali Ahmed

What you mean by negative pressure for gate valves

Hi guys!

I'm assuming your annulus valve is at the Xmas Tree/WellHead Equipment.
In this case, if the referred gate valve is an USV or SSV, please refer to
API 6AV2, for negative (inflow) test criteria.

For positive test, refer to API 6A and API 17D.

Rgds,
Rafael

Hello Ali,

I'll echo Rafael's comment and also add that you have to provide more details re: your exact test procedure (pics, a schematic, type of valve, valve arrangement, etc)... the type of fluid used for the test, as well as the inflow (and outflow) volumes will influence your test and how to interpret results. This forum is ideally suited to help frame our discussions and get feedback, but the quality of that feedback (output) is related to the quality of the input ;)
Rgds

Pierre-edouard,

I had similar questions when I first saw the original question.  I think of a "negative test", like on a packer or bridge plug, where you run a drill string, displace with seawater, and sting into the packer and see if everything is holding below the packer/retainer.  

I suppose on a gate valve, the same principle could happen:  Positive test against the gate from the downstream side, or a negative test by bleeding the pressure from the upstream side (production manifold for example).  

I would need to see a schematic to be sure.

Doug White

Sugar Land

Original Message:
Sent: 08-14-2023 08:32 AM
From: Pierre-edouard Vincent
Subject: Gate valves Pressure test passing criteria (Positive + Negative)

Hello Ali,

I'll echo Rafael's comment and also add that you have to provide more details re: your exact test procedure (pics, a schematic, type of valve, valve arrangement, etc)... the type of fluid used for the test, as well as the inflow (and outflow) volumes will influence your test and how to interpret results. This forum is ideally suited to help frame our discussions and get feedback, but the quality of that feedback (output) is related to the quality of the input ;)
Rgds

Ali,

A little late with a reply and I'm afraid I can't provide you with a definitive answer to your specific query, but possibly raise more questions/debate on the interpretation of test results against acceptance criteria such as what you detail in your question.

Is a 100psi (6.7%) pressure drop over 15mins for a 1500psi pressure test for an annulus gate valve acceptable?

For me the simple answer is 'it depends', as there are multiple factors to take into consideration such as the function of the well, the exposed source, is the valve connected to a flowline or open to atmosphere, the volume exposed to the test pressure, the test fluid.

Depending on the conditions of the specific well, the criteria above may or may not be deemed acceptable, but without any clear guidelines/requirements, the acceptability of a test is more open to interpretation.

Factors such as the fluid composition in the test volume, the size of the test volume, air in the test lines will affect the initial response, so a consistent approach to what is considered the start of the test period is one criterion that needs to be established. API RP 6A provides an example of this in defining the start of a pressure test hold period against a specific rate of change of pressure.

However, from a practical standpoint a critical factor in determining the acceptability of a test is the pressure trend over the hold period. Using the example of an acceptability criteria for a BOP test of a 10% drop over 30mins against the casing volume, for a 3000psi test a 290psi drop over the test period without any further guidelines/criteria would be considered acceptable. However, if the trend has exhibited a consistent drop of 29psi/min over the duration of the test then should it still be considered acceptable?

Unfortunately, it is not uncommon for latter to be considered acceptable for whatever reason - be it a lack of experience, expediency, a lack of QA/QC and/or a lack of clear guidelines/criteria against which to assess the results of the specific test in relation to what it is trying to validate.

Kind regards

Justin Parker CEng FIMechE
Well Integrity
Occidental
Email: justin_parker@oxy.com<mailto:justin_parker@oxy.com>


Original Message:
Sent: 8/15/2023 9:15:00 AM
From: Douglas White
Subject: RE: Gate valves Pressure test passing criteria (Positive + Negative)

Pierre-edouard,

I had similar questions when I first saw the original question.  I think of a "negative test", like on a packer or bridge plug, where you run a drill string, displace with seawater, and sting into the packer and see if everything is holding below the packer/retainer.  

I suppose on a gate valve, the same principle could happen:  Positive test against the gate from the downstream side, or a negative test by bleeding the pressure from the upstream side (production manifold for example).  

I would need to see a schematic to be sure.

Doug White

Sugar Land

Ali,

Two Points

First Point. Every pressure test has some acceptable leak rate.  A 3% or 5% fall in pressure implies that this is an acceptable leak rate.  The normal pressure test equation is Delta Pressure = Delta Volume / (Total Volume x Compressibility Factor) .  But if you rearrange the equation it is Delta Volume = Delta Pressure x (Total Volume x Compressibility Factor).  You can actually measure how much fluid is lost during the test from this equation.  I teach how in some of my classes. Let me know if you want more info.

For a small volume test like a valve with low volume a tiny leak can lead to a large drop in pressure during the test.  This may be acceptable in a subsea wing valve or a chemical injection valve in which the leak is not a safety issue. If people will be working on the equipment downstream of the valve and cannot be exposed to significant pressure build up then it may or may not be acceptable. This is where knowing leak rate is important to understand the importance of the activity, leak rate, and consequences. 

Second Point.  Because of Point 1 the purpose of most pressure tests are not to get perfect leak tight systems but to get mechanically strong systems that won't fail under pressure.  We perform different kinds of tests to the same equipment at various times in the equipment life.  The valve for example may be tested to well above it's working pressure rating to prove it will not fail at working pressure.  But once installed we rarely pressurize the system above design pressure and often only test to a lower operating pressure. We often test frac iron to maximum expected pressure but will only test for a duration of a few minutes to see if there is a catastrophic leak.  But there are expected to be numerous small leaks all through the system and various valves and connections.  If the pipe stays together and the leak rate is consistent from test to test we move on. We do have mitigations to a pipe failure with exclusion zones and pipe restraints. 

If your valve is being tested to prove that it will hold pressure from a pump like a normal tree and wellhead test a small leak would be allowed in most cases. I can explain why if you need. A large leak could lead to a significant problem.  A 10% drop in pressure over a few minutes on a small volume field used valve is too much. A 5% drop after an hour is near perfection for a field used valve. 

Check out the leak rate of these two scenarios and discuss with your team what is an acceptable leak rate for that test. 

Let me know if you have any questions,

------------------------------
Dan Gibson
aka The Well Doctor
Completion & Well Integrity Advisor
Houston, Texas
------------------------------

Hello Ali,

It sounds high, but a few factors to consider - any temperature drop of the time, fluid compressibility, the size of the valve.

There are a few references to API6AV2 standard, and I will second them for the gas filled cavity. The calculations will not work for the liquid filled cavity.

out of curiosity: did you try to flush and reseat the valve and shift the volume as fast as possible to create a sufficient dP for the valve to sit?

And apologies for the late response.

Hi Ali,

There have been a few very good and relevant replies to your initial post... and I apologize (for myself and on behalf of others as well) if I (we) could not give you a simple answer!
The reason is that... there are none, I'm afraid ;)

If you test with water and the test volume is minuscule, even a very steep pressure variation may indicate only a very small leak. If you test in a very (very) large volume (like a subsea flowline... we have to do this sometimes) the test duration will have to be extended greatly to allow for any meaningful pressure derivative analysis! (also you'd have to wait long enough for thermal equilibrium). So even a quasi stable pressure might show you have an unacceptably high leak!

Confused? You should be.
Once the equipment is in service, there really are no solid figures for what's an acceptable leak rate (though there are performance requirements out of the plant)... in my experience what's acceptable would depend on whether a leak would be for breaking containment, from conduit-to-conduit, to another fully-rated pressure-containing envelope, or to the environment!
So there are more variables, which in real-life you will often have to consider for a meaningful risk-assessment (fire-propagation, inventory in your lines, presence of operators nearby, etc).