Play Segment Extrapolation

[DarrylHarris]

How far to people allow extrapolation of data for play segments?
If you’ve only got 2D seismic for example and you’re wanting to extrapolate say reservoir presence play chance from a well 40km away? Is it dependent on the quality of the 2D data, the geological model, changes in basin topography? All of these factors and therefore very situation dependent or do some use rules for how far you an extrapolate and at what confidence level?
Appreciate any comments.

[marcbond]

I think it is less about how far you can extrapolate the data for play segments and more about the Chance you associate with your CRS mapping for reservoir. Certainly quality, distance, data type, etc. will all be taken in your consideration for the chance value to assign.

I posed this question to Jeff Brown, our play specialist at R&A. Please find his response below.

Marc

JeffBrown@roseassoc.com

[marcbond]

Greetings from 37,000 feet. This is a very interesting question and as with most aspets of assessment of chance there are not any hard rules, just some guidelines.

First, playing by the Exxon ‘rules’, if you have a reservoir penetration within your play outline, Play Chance for reservoir is 100%. Having said that, Play Chance is actually specific to each play SEGMENT. Most plays are subdivided into a manageable number of pieces (we advocate 5-25) that we call Common Risk Segments (horrible term) or more simply, play segments, where the boundaries reflect both significant and abrupt changes in geology. Often these reflect changes in GDE or structural fabric. If you are in an unpenetrated segment, you cannot assign a Play Chance value of 100% (I have been known to use a lot of 95% values, depending upon the situation).

The Play Chancevalue you land upon in your assessment is often best derived using Delphi methods, making sure that the perspective is cleary stated – this is not the reservoir chance for the first well to be drilled in the area but rather of finding reservoir at any location in the play segment. Rarely do I have values of less than 0.3 for any individual chance element at play scale.

One final tip. You will likely be familiar with the nine-box chance matrix espoused by R&A – this provides reasonable values based upon confidence and good/bad news for Pg values for individual chance factors in one prospect. As such it is not applicable at play level. Having said that, I find that the values for OVERALL Play Chance often reconcile nicely with the values in the matrix. This can provide an element of consistency to your assessment process.

Hope this helps.
Jeff Brown
Play Specialist
Rose & Associates
JeffBrown@roseassoc.com

[Anonymous]

Hi Daryl,

Personally, would avoid the term “extrapolation” here big time. As Jeff explains your prediction on where any geological parameters are thought to be reasonably similar (where the risk is common (I like the term CRS – but let’s not go there)) should be guided by data & observations and models & analogues. The boundaries between CR segments should be informed by your GDE maps which are products of interpretation and geological insight.

“Extrapolation” takes you dangerously into the world of geostats and variograms. Rather the try to be over quantitative on the “CoS” that you ascribe to any one segment – the important conversation is around what is required to make a business decision and what data can be acquired or studies performed to give you more confidence in being able to inform that decision.

Cheers,

Graeme

Graeme@explorationexcellence.com

[Author]

Posts