01The challenge

Functional groups worked separately. Production engineers, reservoir engineers, facilities and operations teams each held part of the picture and none held all of it.

The consequence was slow collective decision-making, inefficiency, and limited trust between onshore and offshore teams. There was also a safety dimension: every avoidable trip to the field carried exposure.

The early misconception was that this was a technology problem. It was not.

02My role

I led the process implementation team, covering Shell's African assets, Kazakhstan and ONEgas, with remote support into New Orleans.

That meant workflow mapping, asset engagement, process coaching, embedding and training. The same operating model had to work in very different environments, which is where most standardisation attempts fail.

03What I found

The documented process and the real process had drifted apart, differently in each location. People had built workarounds to get the job done, and those workarounds were invisible to anyone reading the procedure.

Decisions were being taken several steps away from the people holding the information. Meetings existed but were not designed around the decisions that actually needed making.

04What changed

We built a collaboration model that defined what would be done, who had to be involved, and how the interactions between key individuals and standard meetings would work. Roles and responsibilities were made explicit.

Physical collaboration spaces were designed around those workflows rather than the other way round. Analytical and productivity tools supported decisions rather than replacing them.

Crucially, the new ways of working were embedded through coaching and compliance measurement, not announced and abandoned.

05Results

Estimated annual benefit of over US$75m in production efficiency.
Well restart times cut from six days to under three for wells that had quit production.
Field travel reduced by 15% or more, lowering HSSE exposure.
Contributed to a field-wide arrest of production decline, including a circa 3% increase in one year against a prior decline.
Reliability above 97%, overall availability of 93%, and near-zero maintenance backlog.
None of this required new equipment. The value came from how people worked, who decided what, and when.