Modeling tilted OWC in Eclipse

[hkazem]

Does anyone know or have an experience in modeling model tilted OWC with Eclipse. I appreciate if you could share your experience, knowledge, documents, etc. in this case with me. You can also reply me on hkazem@petromehras.com.

Regards,

Mehrass

[vinomarky]

90% of tilted OWC cases I have seen have in fact been due to (a) compartmentalization or (b) capillary pressure changes due to (normally) changing perm across the field

If you actually DO have a tilted contact (requiring good perm AND a strong aquifer - you really cant have hydrodynamic OWC change without both), then the most robust way to initialize is to set highly productive constant head aquifers (honoring the change in head seen across the field) on either ends of your model and allow it to simulate for 500 years or so (however long is required to stabilize the contact) prior to putting any wells in there. I've only had cause to do this once - in a field in the PNG Highlands where a lot of data points to this sceanrio

[dipak_m]

Dear Vinomarky,
I am unable to understand "to set highly productive constant head aquifers". Please elaborate.

Thanks

[vinomarky]

'Head' is another word for pressure, but instead of being expressed in traditional units of pressure, it is expressed in height of a fluid (usually water). So a pump that can produce a head of 100ft of water, can produce a maximum output pressure of 100 x 0.433 psi/ft = 43.3 psi. Another common usage is expressing atmospheric pressure as mm of Hg

So, constant head = constant pressure

Highly productive simply means you set the effective kh such that it does not become a rate limiting part of the process. In this case, with highly productive constant head aquifers either side of the model, the rate at which the system stabilizes on for water throughput will simply be a function of your model kh between the aquifers

So - you measure (for example) 1000 psi at 2,250ft SS at one side of your field, and note that on the other edge of your field at the same depth the pressure is 1013 psi (against which I'd expect a 30ft change in contact height), you could initialize your model with two aquifers as follows;

AQUCHWAT
1 2250 HEAD 2250 1E8 /
2 2250 HEAD 2280 1E8 /
/

In this case, aquifer 2 will be producing water, and aquifer 1 will be taking water, and there will be a 30ft (pressure equivalent) pressure gradient imposed across your model

Note, you can achieve the same with a row of water producers and injectors instead if you like - as usual there are many ways to achieve the outcome desired, which is a stable tilted contact created by flowing fluid.

[dipak_m]

Dear Vinomarky,

Thanks for good explanation. We have a reservoir where I found that in same fault block there are different OWC as observed by log. How I incorporate this phenomena in my model where in only one side there is aquifer.
Thanks.

[vinomarky]

First step is to have a clear theory (which does not violate physics) for what is going on in your reservoir.

You are telling me that (a) this block is bounded, and so no dynamic water flow is possible and (b) within this block there are different OWC's

How?

Perched water? Is there a chance you actually do have intra-block compartmentalization (below seismic resolution)?

My question to you is tell me what is physically causing the water levels in an otherwise 'connected' system to be non-uniform

We can then discuss further how to initialize this in a model

[dipak_m]

Dear Vinomarky,
This is fault block but faults are not sealing.

[vinomarky]

So if they are not sealing, what is causing the change in observed contact levels?

You can not simply model observed 'symptoms' and hope to get out of it a predictive model, you must have some idea of what is physically going on.

Were the logs taken after some production, so could the observed water levels be from a high perm stringer sucking water from the flanks?
Are faults blocks actually sealing?
Capillary pressure changes with perm/facies change?

Come back with your idea of what is actually going on, and I'll help suggest an appropriate way to initialize a model to honour it..... otherwise I'm sorry but can be of no further assistance

[Coca-Cola]

Vinomarky,

You make my head spin... I am so confused...

[barbod]

thanks vinomarky,
You explained well from the point of view of a reservoir engineer and I'd like to add another field to this discussion.

I just want to ask about the structure and the trap form, is it a structural trap or stratigraphic? if structural, is the structure continuous or it has saddles every here and there so different traps and culminations exist. in this case, different OWC levels would be natural because of local culminations but not tilted. for regarding the tilted OWC assumption man has to be careful. Assumption of tilting without prior information of regional information and charge and discharge points is not easy...

And a way, how to define tilted OWC, you can define pseudo-injection well like a charging point in your model. Upon achieving best model, you may start to define other wells into your model and go ahead...

[vinomarky]

Good points barbod - again underscores the need to establish what scenario (which does not violate physics or reasonable geology!) is the most likely cause of the observations before attempting to model it.

I've seen a scenario where the two legs over a saddle had different OWC's and different water lines, yet were in pressure connection... the reason? One leg was sealed, and the other had aquifer support, which with a gas cap meant that the water pushed some oil over the saddle from the open to the sealed leg, giving rise to different contacts (and through different oil columns above the water, different water pressures).... a bit confusing in abstract terms without drawing a picture I grant, but I relate the story as it is another example of an observation that without a lot of thought would otherwise be modeled simplistically and incorrectly

[temr]

First of all is there a tilted contact ?
The problem is that in eclipse what you get at the end having 2 different eqlnum with different contact is transition zone and big increase of time calculation due to eclipse needs calculates not only sat changes as results of your well activity but also due to cross flows between cells.
So from my point of view first analyst the data you using to define the OWC , what is reliable and what is not
the 1 priority is well test data and only at initial stage of development of the field , as with time OWC tend to go up - keep this in mind
From technical point of view tilted contact you need to do 3 steps
1 step - define zones with different contact of you layer and making according eqlnum
2 step - make eclipse run with no wells let say for 50 years
at this step you create equilibrium
3 step - upload you initial water saturation (SWAT) after run and use it for you calculations as initial