An Engineer’s Perspective of Delivering on the Promise of Digitization
On the way back from a park that we frequent, I was pleasantly surprised to see my 2-year-old daughter spot squirrels. “Appa, squirrel!” she exclaimed, chasing the closest one until it ran up a tree. I later found out that she had been read to about squirrels at her daycare and was excited to be able to see one. As I joined in her moment of triumph, I reflected on my recent meeting with a client.
I was presenting new insights from our product platform by applying algorithms over their operations data. I, too, was excited to share this with my clients. My daughter and I, indeed, were in a similar space, each sharing our joy of learning.
Data has always been a source of pride in our industry. Data-driven decision-making has been a hallmark of our subsurface work flows. A cynical view of the current data hype is that we always have used data as a means to improve decision-making, given all the subsurface uncertainties. Purists also point out that this has helped us to evolve to our current subsurface work flows. In my opinion, we have been the stranded sailor in the sea (of data) for too long.
Access to cheaper computing resources and algorithms presents an opportunity to rethink our work flows that were optimized for a different time and age. The many learning resources presented through open initiatives in software (Python) and the mathematical libraries, in particular, have reduced the barrier to entry for all engineers willing to embrace new tools and techniques in their analysis.
Is there anything the Google generation cannot learn? What is stopping us from giving a whole-hearted embrace to these exciting tools and techniques in our models? It turns out, we have a two-fold problem with subsurface data: access and structure.
Our current subsurface work flows have led to the creation of silos within organizations. Each group within those silos ends up being the custodian of their own critical data—think about a wells team with completions and well-intervention data, geophysicists with their well logs, reservoir simulation engineers with history-matched data, and so on. Each group may use separate software and viewers that are further complicated by specific license requirements restricting the ability to access the data.
The industry’s collective experience has helped us realize that every decision pertaining to field operations—such as choosing the type of completion for infill drilling or identifying effective injector wells during waterflood operations—requires analyzing data in an integrated fashion. The creation of these silos and the dependency on myriad software to access and visualize them restricts access and, in some cases, creates artificial barriers that delay critical decision-making.
The lack of structure further compounds the challenge of using data in a useful way. As an example, there is no single format to report routine experimental data that are required as part of our subsurface work flows. The standard phase behavior experiments or even the relative permeability data obtained from core flooding experiments vary across different vendors or even across different business units in large organizations. Having been an engineer myself, nothing is more frustrating than looking through many folders of Excel sheets containing experimental data hoping to magically stumble upon the data set you need to use in your new simulation. In this case, while access is not a problem, the lack of structure makes it impossible to run analytics, making this a laborious and futile exercise. An urgent need exists to digitize and, most importantly, standardize reporting formats of such data.
The decision-making in our industry primarily emanates from our subsurface work flows. The availability of advanced analytical tools and their easy deployment enables the creation of new work flows as well as the improvement of existing ones. It is for this reason that we need to fix data issues around access and structure.
We also need to address the more important question of data accessibility in a broader context. Less costly acquisition and storage implies that we will continue to store data—more than ever before. A concerted effort is needed to ensure structure and standards around data, to make it accessible beyond the confines of a single organization. A sharing mindset will enable a thriving ecosystem that will spur innovations from startups as well as academic research groups.
It is heartening to see some great initiatives under way towards this end—the UK Oil and Gas Authority making its data public, Equinor releasing the Volve field data set to spur innovation in analytics, as well as the recent launch of a new forum on Open Subsurface Data Universe. In addition, many companies are teaming up with startups to realize their goal of digital transformation.
We have always sought adventure in our industry. Our collective and childlike enthusiasm for learning and improving work flows will help break organizational silos. It is only by working together that we can deliver on the promise of digitization.
Don't miss out on the latest technology delivered to your email monthly. Sign up for the Data Science and Digital Engineering newsletter. If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.
19 May 2020
15 May 2020