Regulator reports on BP Atlantis documentation

Following last year’s whistleblower charges, US Bureau of Ocean Energy Management, Regulation and Enforcement report finds ‘significant problems’ with BP’s engineering document management but no safety risk. Editor Neil McNaughton reflects on what this means for the ‘digital oilfield.’

As you will no doubt remember, before Macondo, BP was in the firing line following a complaint by a whistleblower, aided by an activist group, Food and Water Watch (FWW), that BP failed to properly manage the engineering documentation on the Atlantis production platform in the Gulf of Mexico. The whistleblower alleged that BP did not properly maintain the engineer-approved as built drawings of Atlantis systems and structures and that this increased safety risks for the facility and its personnel.

We wrote about this at the time (Oil ITJ July 2009) but we were practically alone. Despite FWW’s best efforts at calling press conferences (which were rather well attended by the news services and mainstream media) the story got little coverage. Not until Macondo when the mainstream media and the blogosphere all piled-in with ‘evidence’ that Atlantis was another Macondo waiting to happen and that production should be shut in forthwith.

Back in 2009 we commented as follows, ‘Speakers at document and data conferences across the upstream, from geoscience to construction, have bemoaned the parlous state of their data and the difficulty of getting adequate resources for its management. Some have forecast exposure to regulatory risks and non compliance ‘issues.’

Many companies are struggling to address the problem of maintaining up-to-date engineering documentation across the complex design, build and commission life cycle of the modern offshore facility. Whatever the outcome, the FWW case will be music to the ears of engineering document management software vendors’!

Early this month the Bureau of Ocean Energy Management, Regulation and Enforcement (Boemre) released the findings of its investigation into Atlantis’ engineering documentation. Their report1 should be compulsory reading for all involved in IT, data and/or engineering. The report tracks how BP coordinates engineering documentation from a large number of design and build contracting organizations and how it handled a significant modification to the subsea system, when inspection identified a metallurgical anomaly in a pipeline termination. Ironically, it was the decision, by BP, to rebuild the manifold in the interests of ‘preventing an inadvertent discharge into the Gulf of Mexico’ that led to the charge that its engineering documents were not up to date!

Changes in engineering specification and rebuilding part of a facility—retrofits, revamps and the like are, on the one hand, well understood processes that happen all the time. On the other hand, they are rather tricky to track with IT systems that may or may not have been designed for such. We are in Matthew West country here (see my January 2011 editorial) with the issue of ‘four dimensional’ data modeling that captures all successive states of a model. If you don’t believe me, check out the offerings from Sword CTSpace and Aveva on page 12 of this issue. As the Aveva release states, ‘The reality of the situation for most Owners Operators is that information quality and accessibility is still a significant challenge to manage...’

The headline news is that Boemre cleared Atlantis as fit for service. But this conclusion was reached in a rather tortuous way. The executive summary of the report states that, ‘The electronic document database that BP used to store documents developed during the design, construction and installation of Atlantis was disorganized and inadequate to handle the large volume of documents generated by BP and its contractors. In addition, BP used a confusing labeling system for engineering drawings in the project files. Those drawings had other defects and deficiencies, including undated and missing stamps and signatures, and inconsistent titles for types of drawings.

However Atlantis was judged as safe because ‘the documents in the electronic database were not the materials relied upon by Atlantis operations personnel.’ This sound familiar, rather than relying on some half-baked ‘database’ BP’s engineers were rolling their own—or as Boemre put it, ‘the process for transferring the drawings and documents used by operations personnel on the Atlantis facility involved the use of systems handover packages, [...] engineering documents and drawings compiled by a team [...] including engineers with knowledge of each component being handed over [...] prior to the start of production.’

In other words, the document handover process is more ‘standard operating procedure’ than all-singing-and-dancing ‘digital oilfield’. While the digital oilfield aficionados like to ‘dis’ Excel as a data repository, what is important is that the documents are correct, whatever the format. Of course handing over a bunch of Excel files and CAD documents is not a perfect solution, hence Boemre’s qualifying remarks above. But as Boemre director Mike Bromwich concluded, ‘Although we found significant problems with the way BP labeled and maintained its engineering drawings and related documents, we found the most serious allegations to be without merit, including the suggestion that a lack of adequate documentation created a serious safety risk on the Atlantic facility. We found no credible evidence to support that claim.’

However, the legal action drags on. Late this month a federal judge declined to dismiss a lawsuit from the whistleblower claiming that BP ‘took $10 billion worth of oil and gas belonging to the U.S. government by submitting false documents about operations at its Atlantis rig.’

The Boemre report is interesting in that it shows how attractive the ‘digital oilfield’ approach is in that much of the naming issues could be fixed by a database. But an intellectually unchallenging task like document workflow becomes nightmarishly complex as the real world maps to the IT system. And the shift from clunky but tried and tested manual processes to a shiny new database is tricky to say the least.


Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.