finding GOC in a condensate reservoir from compositional gradient

[catapam]

Assuming a slanted gas condensate reservoir with few wells open at the top.
There is also gas composition from those wells.
There is a way to estimate GOC (for oil rim) from composition change with depth.

Also I would like some materialse that will help me to familiarise with compositional gradient.

thank you,

[kochichiro]

You need to create and match EoS model of your reservoir fluid in some EoS software package, like PVTSim, WinProp, etc. Than you can predict compositional gradient with this software, there is special option for it. After calculation it provides you with GOC estimated depth. See attached file below.

[catapam]

I attached an excel sheet where I put some calculations for
finding contact using pressure gradient and compositional gradient
I put also EOS data you send it in order to show that GOC can be easily seen ploting res pressure and saturation pressure vs depth
the same density vs depth can be ploted.

Still, in that example I found contact between gas and oil rim.
How can thickness of oil rim be estimated. In other words were is oil-water contact?

[catapam]

So I did it in PVTP

I have gas composition and also measured pressure at well depth when sample was taken.
Sample was taken at the surface and recombined.

Using gas gradient and calculated water gradient contact is at 2000m
Using compositional gradient in PVTP contact is shown at 1870 m

Well depth is 1400 m

So this is fitting and show a slightly over pressured reservoir.

How trust can put in this approach taking into consideration that I have only 1 well so one sample and one depth.
For compositional is better to have multiple samples as I understood.

[kochichiro]

Well, I try to answer your questions one by one:

1) To estimate oil-water contact you need to know capillary pressures and model saturation functions with some petrophysical software like IP.
2) You see, EoS is kind of tricky thing, you can get with it whatever you want and there is always the question which parameters to choose for EoS matching. For example, I use surface separation data to model separation process with EoS and regress it on field data. Because I really SURE in surface separation data, but I quite UNCERTAIN in lab PVT data from Russian PVT labs, so I try to ignore it in regression.
3) Of course, all calculation and prediction data still just calculation and prediction (otherwise speculation). The only straightforward method to prove and estimate contacts depth in your case is to DRILL A NEW WELL AND TEST IT WITH MDT. There is no other way out to prove it with fundamental data. The application of calculated and predicted data lies in slightly different field. It needed to prove investment in more detailed field study.

[catapam]

Well, I try to answer your questions one by one:
2) You see, EoS is kind of tricky thing, you can get with it whatever you want and there is always the question which parameters to choose for EoS matching. For example, I use surface separation data to model separation process with EoS and regress it on field data. Because I really SURE in surface separation data, but I quite UNCERTAIN in lab PVT data from Russian PVT labs, so I try to ignore it in regression.

Thanks for the answer. In fact a new well is already planned.
regarding what u said about EoS because I'm quite new in this.
for this well I have surface gas and liquid sampling
lab analises and recombination
no CCE, CVD runned into the lab so sat pressure is not known
for matcing I have 2 points, 1 is separator conditions with GOR meassured into the lab and second is surface conditions with GOR meassured into the field.
so i'm trying to match GOR
I'm using PVTP software

What do u think?

[kochichiro]

Well, you do everything right. Try to model separation process with recombined fluid and lab GOR and than regress on GOR (most statistical weight) and API gravity of stock tank liquid (allocate to it half of GOR statistical weight). In many cases field GOR may be quite uncertain.
You do not need to model recombination, because from the EoS point of view there is no equilibrium in fluid and you would have hard times trying to match EoS model with EoS recombined fluid.