Private Forums
Barriers at Xmas Tree (to break containment)
12 June 2019
Dear all

We are  preparing for a complicated snubbing fishing programme and we have just received a request that we'd like the members opinion on.
  • The GAS well is compromised with a plugged completion. Potentially 6000 psi surface pressure.
  • We are being requested to break the wing valve off the tree to rotate an instrument block.
  • There are no spares and no people to assist

If we test the LMV from above with water and keep it closed and then test the UMV with water from above does that constitute an acceptable double barrier?

I feel not but………I may be wrong?

Testing gate valves from the wrong side – is that considered an acceptable barrier to permit breaking off the wing valve? This is a gas well and we are testing with water.

As usual it is not as straightforward because we have fish and a plugged completion ………12 months ago. And nobody has a clue what state the well is in now.

We are concerned about breaking the tree integrity, because we can foresee potential damage to the ring groove and we have no spares!

I would be interested to hearing opinions on what can be accepted as barrier.

I need an opinion ASAP, please.
15 answer(s)
Drilling & Workover Liaisonman / Supervisor
Saudi Aramco
Total Posts: 7
Join Date: 15/11/18

The first thing is to have two isolation barriers. SITHP of 6k is not a joke. Install a tubing plug on the shallowest nipple or a nipple- less plug like Halliburton ''Mirage", thereafter set TWCV on the tubing hanger. A BPV might be set if you do not have a TWCV. 

With these, you are confident of 2 good barriers. LMV and UMV are not isolation barriers for such scenario. 

Unfortunately the only way to test the integrity of the LMV is to inflow test, testing from above will not give you the accurate outcome. It is not a good practice.

For the UMV, close the LMV (3rd barrier), bleed off pressure on the tree to zero, break-out the  flange to the flowline, line up to a pump and pressure test the UMV from btm.
Managing Director (
Relentless Pursuit Of Perfection Ltd.
Total Posts: 437
Join Date: 10/01/05

I'm pleased to post this on behalf of a regulator in Europe. Due to the need for Regulators to avoid conflict of interest ("making recommended course or action"), he would rather remain anonymous.

"With reference to the question raised the only immediate discussion point I can see is the gate valve design. Is the gate a ‘split gate’ design or a ‘solid gate’ design? 

Pressure testing a ‘solid gate’ gate valve in the wrong direction is not ideal to confirm its integrity. Pressure testing a ‘split gate’ gate valve in the wrong direction is slightly better due to the design of the gate valve but it is not ideal.

It would be useful to develop a working relationship with their local Furmanite International (aka Team Inc) representative for assistance.

It might also be useful to read through the Oil & Gas UK ‘Well Life Cycle Integrity Guidelines (Issue 3 dated Mar-2016)’ for further information on the verification of well barriers in difficult circumstances."


I hope this helps


SPREAD Associates
Total Posts: 29
Join Date: 16/08/10

Agreed with Well Examiner comments, of course:  no way can disassemble tree without replacement parts, and to wait for snubbing unit is sensible course, but :

  • Were Master Valves closed because of tree integrity concerns ?
  • This seems to be at a  remote location and with a restricted operator. It is possible that bringing in a snubbing unit with nowhere to go, while waiting on parts to rebuild the tree, might have a prohibitive cost -day rate and lost production. (I don't accept the six months from CIW for  API ring gaskets, nuts and bolts- just needs someone on world phone calls begging or  borrowing somewhere) but incentive to test tree first.
  • In the scenario :( "I wouldn't do it, but if I did do it, this is how" ,  OJ Simpson engineering ! ):
  1. I would have the viewpoint that the important goal is to establish tree integrity and that the issues on testing LMV and UMV as barriers and which direction, are a distraction.
  2. Work down- pressure up to 6000 psi or worst case if well had flipped to gas, test tree above UMV.
  3. Hold pressure and open UMV remotely.
  4. Test and hold again then open LMV. If pressure, water will drop and gas into tree which we have just tested. If no pressure, our 6000 psi very small water volume , can do no damage to a plugged well. ( although no downhole info or on tubing annulus pressure )
  5. No need for tubing plugs or H2 BPVs- time for that was 12 months previous-will get on with job on snubbing unit arrival.
  6. If pressure in well, we can now test UMV and LMV for inflow as we like- to max possible pressure( after 12 months )- with bleedoff line.
  7. If no pressure, then a dead well after 12 months constitutes a barrier ( barring Acts of God -not covered in NORSOK ), MVs as second.
  8. Wait for snubbing unit and test line. 
Managing Director (
Relentless Pursuit Of Perfection Ltd.
Total Posts: 437
Join Date: 10/01/05
I'm pleased to post this on behalf of a retired (UK) well examiner.  Due to the need for Well Examiners to avoid conflict of interest ("making recommended course or action"), he would rather remain anonymous.

"I think most of what I would have to say has already been covered during the day (12th June)

First off what is the rush,,, have a disaster? I would leave everything as is until all personnel and equipment is available at or near the well site.

As Pete Aird pointed out, hire an accredited expert. If they are planning for snubbing ops, I am sure the snubbing operator could provide such a person and also back up data from similar operations.

I do not understand why at this point all of a sudden they have to take the “the wing valve off the tree to rotate an instrument block”, its already noted that the well has been left alone for a year?

As Bill Abel noted, they can more or less do anything they want once they have the snubbing unit BOPs nippled up on the well

My recommendation would be get all the personnel and equipment available, sort out the programme, get professional help, do the what if’s until then don’t poke the bear."

Well Engineering Consultant and Instructor
Olango Consulting
Total Posts: 30
Join Date: 23/03/16

Table 21 on page 74 Norsok standard D10 rev 4 2013

SPREAD Associates
Total Posts: 8
Join Date: 08/04/14
Hi Steve, Where in NORSOK D10 rev4 is that statement?

I guess pp13 is the most suitable barrier solution in the event well pressure is insufficient to inflow test the LMV and UMV??
Documents uploaded by user:
NORSOK D10 rev 4 pp13.docx
SPREAD Associates
Total Posts: 8
Join Date: 08/04/14
Thanks for your rapid feedback Guys.

You are basically saying what I was thinking. I do not want to break the Tree now - just before we rig-up a Snubbing Unit on top of the Tree.The logistical issues alone makes this risky. We have no spares in country for the Tree and CIW are quoting 6mnths+ for new parts. And if we screw up with unskilled labour the Snubbing Job cannot continue. And the Snubbing job is essential for commercial reasons.

We are about to perform a Tree Integrity Test with water to give ourselves a warm fuzzy feeling before rigging up the snubbing stack (10 rams.........).

We have no means to bleed-off safely until the Test Kit arrives but we want to know if we are good to go when the Snubbing and Testing spread  arrives so..........I do not want to open the LMV until I know everything looks good - therefore I am keeping the LMV closed even when testing against the UMV. (Gonna try trapping pressure  below UMV, observe pbu, etc, I say it is basically re-assurance before rigging up the Snubbing BOP)

IF we were to test above LMV then test above UMV with LV OPEN we will have effectively tested (with water) 2 barriers from above.

I do not consider that adequate. Wrong direction. Agreed?

I do not want to open tree to well because if we get gas into the tree we cannot bleed off safely. (No equipment available )

The obvious way to handle this is a slickline deployable Retrievable Bridge Plug and BPV .

But why is a Retrievable Bridge Plug tested from above considered a barrier? I'm hearing by design?

And an H-1 can only be inflow-tested. Should be good enough.

Then close LMV and UMV.

Your thoughts...............Many thanks. 

(And thanks to Dave T for facilitating this discussion)

Documents uploaded by user:
Abel Engineering
Total Posts: 10
Join Date: 18/11/11
From past experience with snubbing operation and coil tubing interventions, it is routine and acceptable to rig up on a tree and re-enter the well without a test from the well upwards (as suggested, but no one is telling you how to get that done). While it is correct to say the test from the top does not test the seal face from below in some cases you do all you can and proceed.

I would question the logic on taking off the wing valve unless it is absolutely necessary to perform the workover. Perhaps you can install this "instrument" in a temp spool above above the tree to get the info that will provide?

I do not have enough info to feel comfortable with a recommendation, but I will say it is common field practice on snubbing jobs to test from above and proceed, because that is ALL that is feasible or possible (short of putting on a freeze, or setting a plug that might junk your well or damage the tubing when retrieved). 

I would NOT break off a tested barrier (assume tree was tested when installed) unless it is absolutely necessary. Having no replacement ring gasket is NOT acceptable. Fly one in, delay the job until that is in hand.

If possible put a BPV in the hanger and do a body test of the tree and install tested barriers above (and test in the proper direction) and treat the low master valve as a spool during the job. Remember you can stack up what you need for all contingencies. 

Bill Abel

Well Services Manager
Sasol Ltd
Total Posts: 7
Join Date: 11/04/16
Dear Ian,

I think this scenario of yours is a classic for generating the next blow out incident, if you would consider taking short cuts, hence I am suggesting extra precaution and would let management know. As mentioned by some, most gate valves have a pressure side and must be mounted the right way to allow them to seal as per manufacture. Hence I would also be reluctant to test them from above and accept that as a pressure test (not yet discussing using water as a test medium to test for a potential leak path of gas). By the way having one or two (or three or more) gate valves in a row (without the bleed option in-between) is anyway only considered to be ONE BARRIER in many companies.

If it would be me having to make the call and knowing our industry, I would suggest you go for a slickline intervention and run at least a deep set plug (if and as deep as the plugged completion allows) followed b y a TWCV/BPV into the tubing hanger.

Well Engineering Consultant and Instructor
Olango Consulting
Total Posts: 30
Join Date: 23/03/16

You are testing gates against the direction of flow and this does indeed not constitute a valid barrier.

Norsok D10 rev 4 states “Tree valves of floating gate design shall be tested in the direction of flow from reservoir. The test of master valves may be omitted if the shut-in pressure is insufficient to give a conclusive test. This is provided the wing valves are tested in the flow direction by pressuring up the X-mas tree cross.”

Do not fully understand why there is a need to install an instrument block on the tree. Recommendation would be not to compromise the tree integrity.

If the tree is not leaking now, leave it alone!!

Rig up on the top of the tree with a BOP and install any required instrumentation on the BOP. The BOP can be stump tested and that would only require the upper tree cap connection to be tested once the BOP is installed and that can be done against the swab valve leaving both the LMV and UMV closed. Ensure that the snubbing BOP has a flow line with chokes installed that allows bleeding off of any gas to a separator or MGS.

If there is a space issues rig up using additional risers. Once the BOP is rigged up and tested you can then open the tree valves and get started.

Drilling Specialist/Well Engineer/Training Consultant
Kingdom Drilling
Total Posts: 413
Join Date: 10/01/05
Call a well control specialist and seek their advice. They shall have likely done this and much more before.

Personally this is a tier 2 or a potentially tier 3 scenario, (think what ifs?) that ever very experience personnel are not that well versed or confident with. 

i.e. Without speaking to such a peer, first and foremost to have a well prepared and assured plan with contingencies in place.

Drilling Consultant
Total Posts: 1
Join Date: 15/02/19
Hi Ian,

Testing the valve from above will not constitute a relevant test as its not in the direction of flow. Can we see the schematic of the tree? To test the UMV, could you trap pressure between the UMV and LMV and then bleed off the pressure above the UMV to see if it builds up above? You may not have enough volume to build, so may still need to keep some pressure above it. This will only help you to get an understanding of the integrity of the UMV. 

Do you have a split gate valve or slab type?
SPREAD Associates
Total Posts: 3
Join Date: 24/02/18
Hello Ian
How confident are you of the current SITHP? 
If you can pressure test the LMV from above with water you could also perform an inflow test against the SITHP which would allow you to
validate at least 1 barrier   in the direction of flow. 
It may be possible to access the UMV to perform a similar test against well pressure.
without a tree schematic it’s difficult to provide more advice 
SPREAD Associates
Total Posts: 1
Join Date: 22/08/17
Hi Ian,

I would echo the comments from Daniel above as you are not testing in the direction of Flow of the Hydrocarbons so the test is a false one and not adequate.
CWI Engineer
SPREAD Associates
Total Posts: 11
Join Date: 11/05/19
Hi Ian,

My opinion on this, as follows:

  • having the valves either slab or split gate, testing from above will test on a different seating area than the area on which the valve would seal from the direction of flow
  • on a split gate valve, you could potentially test from one of the body grease fittings. In this particular case the test would subject both seating areas ( from the direction of flow and against)  to the test. A word of caution here-when the grease fitting cap is open, the well maybe under one barrier or none subject to the integrity of the check valve inside the fitting.
  • water and hydrocarbons have different allowable leak rates through the valves 
  • relative to the testing procedure proposed above..test the LMV first then test the UMV ...if UMV leaks and LMV holds you wouldn't know this unless you have a monitoring port between the valves. The cavity between the 2 valves is to small to use compressibility calculations. However, this tests would not give information about the seal integrity from the direction of flow
Finally, in my opinion, following the procedure proposed above, there will not be a confirmed tested barrier from the direction of flow.

Hope this helps,


Jump to top of the page