[OSGeo-Discuss] GIS applications to off-shore E&P (oil & gas)

[Paulo Marcondes]

2008/2/21, Bruce.Bannerman at dpi.vic.gov.au <Bruce.Bannerman at dpi.vic.gov.au>:
> By E&P, I'm assuming that you're referring to Exploration and Petroleum?

Exploration and Production, as is usual in the industry, sometimes
also referred as upstream

> I don't claim to be an expert, but I personally think that this is just
> another application of spatial data, albeit with its own unique challenges,
> e.g.:

What I am more concerned is how to build a bussiness case to sell the
idea of using GIS in the reservoir management role, in part as a
decision support system. Asides may be well the storage and management
of spatial and non-spatial data.

> - As I see it the problem space is essentially 4D in nature, where true 3D
> non-uniform data needs to be modelled, while taking into account changes

Also, the time dimension is usually non-continuous, although that may
not be a problem if you treat time as a continuum.
I dislike the term 4D, at least when applied to to time-lapse seismic data.

> - The amount of data to be managed is challenging. We're looking at in the
> vicinity of 60TB and increasing rapidly.

I don't even have a ballpark as to the size of information managed by
the company.

> - There are a number of 'true 3D' modelling tools (NB: for others, I'm not
> talking DEMs here) that allow you to model sub-surface data at a mine scale
> (e.g. Datamine, Vulcan etc),

I happen to know them (at least by name), because I have also worked
in the mining biz.

> and several others for regional scale work (we're using GoCad and GeoModeller).

I also happened to have used GoCAD here, and have a mixed feeling regarding it.
Too many important features are lacking, uninportant features galore,
the fact that it is unstable with large amounts of data.

> I don't know of any serious 3D data management in a database yet, though I
> understand that Oracle has announced something with v.11. As always, a good
> Spatial Metadata solution is essential.

We had a discussion sometime ago regarding methods and processes to
store all these products and subproducts. One key point that was
brought repeatedly was about the importance of the metadata and
softdata associated with those [geo,fluid,flow,etc] models.
However I got out that loop, so I don't know how it progressed.

> - 2D spatial data and tools such as GIS and Remote Sensing are very useful
> in getting an understanding of your area of interest. This includes
> traditional data

That is quite well known, but not widely used outside "expert" areas.
What I feel is that some gains are in spreading the use of these tools
around the E&P bussiness.

> Cross Sections through the earth
> surface may also be considered as 2D data, though they also have
> considerable value in the 3D modelling process.

Well, I would consider them do be 3D, mostly because they represent the Z axis.
Ah, then talking about the Z axis, we have a big issue.
Geologists talk about Z in an "altimetry" sense. Sealevel is 0, above
sealevel is positive.
Engineers usually talk about Z in a depth sense. Land surface or
seabottom is zero, below is positive.
Geophysicists further confuse things, because they talk use travel
time in the vertical direction and sometimes Z also represents depth.
So, we have two different domains for the vertical axis and two
different sets of views for one of them.

> - Sonar, sesimic etc data is also 'spatial' data.

That was easy =] Seismic is also 4D (non-continuous time). Sonar and
other seabottom surveys can also be time-lapse.

> - There are dedicated Petroleum systems that you may be interested in e.g
> Petrosys, dbmap etc.

I am not aware of any of these two, let alone the "et cetera" part of it.

> - When it comes to modelling Geoscientific data, the variety and depth of
> inter-relationships of aspatial requirements is probably beyond the ability
> of most (if not all) GIS to manage effectively.

What you mean with aspatial? What kinds of data?

> You may want to have a look at the work happening with GeoSciML and the 'One
> Geology' project for more background on this. See:

Interesting on a quick glance. Will delve into it.
-- 
Paulo Marcondes = PU1/PU2PIX
-22.915 -42.224 = GG86jc