WBHP not equal to WBP when a well is shut in in Eclipse

[halim78]

Hello friends,

I am running a forecast case for a black oil model with many existing wells in history.

I have a problem with these producing wells, when the producer will shut in in the forecast the WBHP will not equal to WBP.

The well datum defined in WELSPECS is the same of reservoir datum defined in EQUIL

For the new wells defined in forecast , WBHP=0

Thank you very much

[vinomarky]

Yes - saw your post on the SPE TIG as well

If I read your question correctly, your issue is that the WBHP is zero when shutin in Eclipse. If this is the case, then yes I see this as well, and a solution is simply leave the well on WCONPROD control, but set to RATE control, specifying a very very low rate (ie 0.001 BOPD ) - this way Eclipse does not shutoff the connection and you get a consistent WBHP trace in your reports

[halim78]

Thank you for your quick answer.
I think you misunderstand my question.
The problem is that WBHP is not equal to WBP when the well is shut in in forecast but it is the case in history(See the attached document).
I didn't understand why?
The well has no cross flow.

The document is here :

[link Point to another website Only the registered members can access]

[vinomarky]

It does indeed look odd. I'd probably go to SLB support and ask them to take a look

Just a thought - it seems to behave 'normally' until the THP goes to zero. Why dont you go ahead and try setting WCONPROD to 0.001 BOPD at this point and see if it makes a difference.

[halim78]

I contacted SLB support and they didn't give me any valuable answer that resolve my case:
These are the answers given by SLB:
------------------------------------------SLB answers---------------------------------------------
Answer 1 :
Hello,

It should be OK, as the WBP's is calculating your average reservoir pressure in the drainage area which should more
then your bottom hole pressure (WBHP).

Please let me know if you need other details.

Best Regards

Answer 2 :
Hello,



There is different reasons why BHP differs from the WBP when a well is not producing:


* BHP is not calculated at all when the well is SHUT. Shut wells are treated as non-existing by the simulator during the dynamic flow calculations (although their potential is calculated and they can be re-opened by WTEST and several other options in ECLIPSE ...)

* If the well status is STOP then the well calculations are performed by the simulator. In this case, the well BHP may be different from the block pressure when there are cross-flows between different layers.

* If the well is not SHUT (and not producing) , it's BHP may be different from BP because BHP and BP have different pressure datum. BHP is referenced to the center of top perforation by default if it is not explicitly defined in item 5 of WELSPECS. BP is expressed with reference to the reservoir datum in EQUIL or DATUM keyword (if defined). To find the pressure at the connection of the well use mnemonic CPR.



I hope this will help.



Best Regards

Answer 3:
Hello,

For the period (history) where the well was closed, the reason of this is may be the fact of having zero rates for all
phases with the status of the well being OPEN in WCONHIST. could you please try to change status of the wells to SHUT
(in WCONHIST keyword) whenever the well is not producing (zero rates) and see the effect.

Please let me know if it helps, if not could you please send me the data file I will try to reproduce the issue.

Best Regards

Answer 4:
Hello,



1- Sometimes when you don’t set your wells status as closed in WCONHIST and they have a rate of zero, it may cause convergence problems in ECLIPSE.



2- From the well inflow performance relationship Q=T*M*(Pavg-Pw-Hw). Where T is the connection transmissibility factor, M is the phase mobility, Pavg is the pressure in the grid block (WBP, WBP5 or WBP), Pw is the bottom hole pressure, and Hw is the well bore pressure head between the connection and the well’s BHP datum depth. If the well is shut at the surface, Q=0, then Pavg=Pw+Hw. So the difference is the well bore pressure head, mostly liquid hydrostatic pressure.





I will check for the history phase why it is showing the same value.





Best Regards

---------------------------------------End of SLB answers------------------------------------------------------

[vinomarky]

I've suggested it twice, and I saw another person suggested the same thing on the SPE TIG to put in a low rate rather than shutting the well. Have you tried this, and what was the result?

[Shakespear]

Vino is correct

Try this with Chap.DATA

Very small rate " 'P1', 'ORAT' 0.010000 / "

-- SPECIFY REPORT AT 10 DAYS
TSTEP
10.00000
/

-- CUT OIL RATE TO 100 BPD
WELTARG
'P1', 'ORAT' 0.010000 /
/

-- SPECIFY UPPER LIMIT OF 1 DAY FOR NEXT TIME STEP
TUNING
1 /
/
12 1 50 /

Then using SHUT

-- SPECIFY UPPER LIMIT OF 1 DAY FOR NEXT TIME STEP
TUNING
1 /
/
12 1 50 /

-- SPECIFY REPORT AT 10 DAYS
TSTEP
10.00000
/

-- CUT OIL RATE TO 100 BPD
-- WELTARG
--'P1', 'ORAT' 0.010000 /
--/
WELOPEN
--'*' / opens all shut or stopped wells in the field
'P1' SHUT / shuts well P1, if it is open
/
-- SPECIFY UPPER LIMIT OF 1 DAY FOR NEXT TIME STEP
TUNING
1 /
/
12 1 50 /

-- ADVANCE SIMULATION TO 50 DAYS
TSTEP
100.00000
/
WELOPEN
'*' / opens all shut or stopped wells in the field
--'P1' SHUT / shuts well P1, if it is open
/

-- PUT OIL RATE BACK TO 1000 BPD
WELTARG
'P1', 'ORAT' 1000.00000 /
/

-- SPECIFY UPPER LIMIT OF 1 DAY FOR NEXT TIME STEP
TUNING
1 /
/
12 1 50 /

In the first case you will see the pressure build up. In the second case over the same time frame it will be ZERO, just as SLB said. The well is not connected to the simulation blocks. Hence zero pressure.


You will need this in SUMMARY to see the right results

-- BOTTOM HOLE PRESSURE FOR WELL
WBP
'P1'
/
WBHP
'P1'
/

[halim78]

My friends,
I think you are misunderstanding my case,
I didn't shut in the well in the forecast, the well is shut in automatically due to low FTHP.
Even I put small rate, the well will not produce.

[vinomarky]

Change to rate control with no vlp's at the point it is shutin

C'mon, we are trying to help you troubleshoot this - it's frustrating asking the same first step over and over again - I'm not going to offer anything more until it's done

I understand switching to rate control may not be what you desire as far as simulation prediction, but the first step is to identify what is causing the issue, once that is nailed down you can start thinking about how to overcome it

[halim78]

Change to rate control with no vlp's at the point it is shutin

C'mon, we are trying to help you troubleshoot this - it's frustrating asking the same first step over and over again - I'm not going to offer anything more until it's done

I understand switching to rate control may not be what you desire as far as simulation prediction, but the first step is to identify what is causing the issue, once that is nailed down you can start thinking about how to overcome it

I tested reducing the rate to 2 Bopd, and the WBHP ~=WBP9.But in the shut in period they are different.
Here is the file

[link Point to another website Only the registered members can access]

[Shakespear]

How large is the difference % wise ?

[vinomarky]

Had a look at the doc - so it proves the theory that having a small but finite rate yields expected behaviour

I suppose it boils down to what you need to achieve. With this test as 'proof of concept' you could simply disregard WBHP values when shutin and go with WBP9 instead. If you are history matching, you could change all zero rate events to 0.1 bpd rates. If you are predicting, expect a well to be shutin due to liquid loading AND require WBHP to be correct, then you could use ACTION statements in the SCHEDULE section, triggered on you well rates dropping below 0.1 bpd after which it would be 'recompleted' as BHP pressure control with 0.01 BPD rate target....

I've seen WBHP going to zero before with shutin wells, but not this sort of offset... by chance are you using multisegmented wells option? I'm just wondering whether there is some strange changing liquid levels being modeled in the wellbore... but I guess if that were the case we'd expect to see same behaviour in 2 bpd case as well.... nope, out of ideas for the time being, unless you want to post the deck for us to take a proper look at....