I’ve got my second in-person conference of the year under my belt – this time at the Energy in Data conference. Now I’ll admit, the audience for this conference was a bit “data heavy” in that most folks there were either data scientists or geologists that were, dare I say, a bit data obsessed. That being said, a true overarching theme for the conference really could have been that “data is the new oil”. Peeling back the layers on that, here were some of the specific themes that resonated with me.
Who owns the data?
Is it the leaseholder? The company that put the sensors in the ground? The company warehousing the data? This industry most definitely has a culture where people hold their data very close to their vest. There is a fear of giving away the “secret sauce” of how they got to an answer. We saw this first-hand starting with our very first oil & gas data extraction project which focused on pulling directional survey data out of documentation that was filed with the Railroad Commission of Texas per regulatory requirements. It seemed as though operators did everything they could to make the data as inaccessible as possible. Following the letter of the law, while not quite the spirit.
At Energy in Data, I began to hear a shift in this thinking as folks began to seriously toy with the ideas of “what if” we shared this subsurface data with one another. Driven by the Open Subsurface Data Universe (OSDU) and its efforts to create an open data platform and ecosystem with a goal of sharing non-differentiating technology to accelerate innovation and transformation in the industry, what if you could:
- Accelerate your organizations own design and development efforts by taking advantage of work that had already been proven?
- Reduce the cost burden of creating your own subsurface data platform?
Now we aren’t there yet; we haven’t even gotten everyone to agree on standard yet, no less actually sharing data. But with some of the “big boys” in the space driving this initiative, I’m excited to see how the industry will continue to evolve.
Can you trust your data?
I heard loud and clear that data integrity is an issue. These are huge multi-million dollar decisions that folks are making and faith in data is just really low. Wells span back 50-100 years. You can’t just “go back in time” and sample, you have to do the best you can with what information was gathered at the time. We know that the most reliable data is going to be that which is closest to the source. And so in all honesty, those actual source documents – the unstructured content – is often a much better source of data than any database that an organization might have manually created over the years.
Today, we are blessed with technologies that can help with this (and if you have a problem with lots of unstructured content sitting around, either in physical boxes or on some sad file share or neglected repository somewhere, please reach out!) There’s technology that can take those forms and reports and automatically pull out the relevant information with high accuracy and then feed it into whatever systems you need to drive your business.
The downside? There is a LOT of data. I heard one presenter say that for every 1GB of data (documents), there is 10GB of metadata; that metadata provides the context. Even with technology, it could be cost prohibitive to go back through (cleanse, classify, etc.) all legacy data, so you need to be smart about what you are actually extracting. Which leads me to the last theme…
What problems do you want to solve with the data?
Gathering data for data’s sake isn’t adding value, it just makes you a hoarder. So organizations need to really think about what questions are you trying to answer.
- Should I lease this land?
- Where should I drill?
- How do I optimize production?
In order to invest in pulling the data out of your unstructured content, you need to tie it to a business problem – to something you are trying to do and achieve. We know there is pain there. I heard folks say things like “it takes me 5-6 weeks to gather all of the information on an asset”. But as a “housekeeping” project, adding the infrastructure to extract and analyze data from your legacy content is likely to be rejected. During one of the Energy in Data breakout sessions, we lamented on this issue specifically and our conclusion was that we need to “hide” this work in other projects because without that more strategic initiative driving, it just won’t happen. When tied to specific business initiatives, it can become a no-brainer.
