GeoForum 97 (May 1997)

The European GeoQuest User Group meeting held in Prague last month was attended by around 300 delegates from Western Europe and the CIS. It was an opportunity for GeoQuest to outline its strategic directions, particularly a reorganisation of the Finder Data Management system, which is covered in this month's editorial, and also for users to relate their successes with GeoQuest products, and also occasionally to blast off at GeoQuest with both barrels for failing to give less that total satisfaction in some domain or another.

In the latter category, Barbara Michalek of BEB, the German Shell/Esso joint venture, described their experience installing and working with GeoFrame. In a very entertaining talk, Michalek likened GeoFrame to a fairy-tale monster, "Powerful, beautiful, they smile at you and then they have you for dinner!". BEB seems have survived this experience but have had a rough ride as an early adopter of this technology. Working primarily with petrophysical interpretation, various problems were encountered, not least those involved in migrating to and managing the new UNIX environment, and the complex inter-relationships between GeoQuest software, Oracle and the operating system. Migration from the legacy Schlumberger LOGOS product proved a painful process but strong support from GeoQuest during this process pulled BEB through. Michalek in common with other GeoQuest users was very keen on the establishment of a Bulletin Board, where users could post and share experiences. A Schlumberger source, commenting on the software monster analogy stated that GeoFrame was complex software in its youth, that users, while inevitably experiencing teething troubles were – deep down inside – confident in the size and strength of the Schlumberger organisation, whose Oilfield Services department has a $300 million R&D budget, while GeoQuest boasts 400 plus software engineers. "It will work".

The high point of the conference for many data managers was not a mainstream GeoQuest product, but a bespoke development for Sonatrach. Salah Mekmouche described how, after making the move from their legacy MWDB/Atlantis workstation to Finder, Sonatrach Production wanted a user friendly graphical interface which would hide Oracle forms from non IT users. They also wanted a front end in French, presenting users with localised, familiar terminology. Herve Ganem of GeoQuest came up with the goods in the form of a Wab Browser front end for Finder. This provides a fully functional access to the power of Finder, with a point and click intuitive interface, with the added benefits accrued through the use of Web based technology, not the least being a solid, easily managed multi-platform client server deployment. A spectacular demonstration of the technology ensued with Ganem programming from scratch a web browser capable of generating an interactive map, which allowed the user to select wells from which a cross section was generated on the fly. In a few minutes, pausing only to explain some of the intricacies of the methods involved, the fully functional applet was up and running.

What we appreciated most about this demo was the way in which the "big iron" technology of Finder was being opened up and made more useful by a simple add-on, in this case Microsoft's Front Page. A $150 program offering massive leverage in an environment like this. Intriguingly, this web front end for Finder is a completely separate development from GeoWeb – the official GeoQuest Web front end. The latter, at first glance appears a much less interesting approach to the same problem. GeoWeb utilises the client server paradigm of the web to provide world wide access to the corporate data store, but does not offer anything like as much scope for in-house customisation and openness. The demonstration of the Sonatrach Web Browser is an example of a variety of similar joint developments between Schlumberger and its clients.

Agip's GisForall (a Finder-like E&P GIS) and Total's DocXplo E&P document management system are both being deployed and maintained by GeoQuest and offer interesting comparisons with the mainstream GeoQuest product line.

David Feineman (BP exploration) described how an analysis of the workflow involved in subsurface analysis enabled BP to specify and deploy a complete software and hardware implementation for their Algerian joint venture with Sonatrach to appraise in record time seven tight reservoir gas accumulations in Algeria's District 3. The allotted time frame of 2 years and the intensive drilling activity (2-3 concurrent drilling rigs) meant that some slick planning was required to specify a suitable the IT infrastructure.

The primary requirement was for solid "one stop shop" support in a remote and hostile environment. Further analysis was performed using heavy duty Business Process Re-engineering techniques. This involved the following analysis:

An inventory and assessment of the tasks in hand

Process modelling to establish task sequence

Data flow between tasks

Vendor mapping of applications onto workflow

decision

Critical path analysis of the workflow revealed many potential bottlenecks in both structural interpretation and reservoir modelling. Portfolio selection came up with a best of breed selection of applications suited to the task, and adapted to the speedy solution of the critical path bottlenecks. Conclusions of the study were

More than 100 tasks were required in the appraisal workflow

Structural framework definition and 3D modelling both fell on critical paths – so had to be performed quickly

No single vendor offering provided everything

Too many products were required in the initial specification to be supportable

Support of openness is a critical issue in the multi-vendor environment.

Two defects in existing solutions were identified, Geoshare which was a weak link because of the limited number of data types supported in the different applications and the lack of support for distributed databases.

This major BPR project took around 6 man months to achieve and revealed that the combination of hardware and software best suited for an asset team was highly dependent on the job in hand. Another interesting aspect of the solution was that, contrary to many presentations of the asset team's workflow, this was a highly linear process. The results of the structural interpretation ware passed on to the modellers who in turn pass their results to the simulators. No fooling around with multi-disciplinary iterations in this New BP go-faster high pressure environment!