5 Data Science Tips for Drilling Challenges

I had a sit down conversation with a prominent operator in the #Permian this week. He was looking for us to solve a particular challenge he was having with some #BHA components that I must admit I haven’t seen before. We started to talk through the challenge and I asked him what his #data looked like. After seeing what we had to work with I was inspired to write this brief outline of some key best practices to put in place when seeking #data forward solutions.

?Seeking a #data solution to #drilling challenges requires features from quality management in addition to technological implementation. A robust process should include the following five pieces for the data science to effective help:

1.?????Data collection and analysis needs to include good and bad data points.?Because the implementation of #data collection has a cost associated with it, many #operators choose to only capture information when there is a failure. Leaving out the good #data eliminates the ability to perform trend analysis and create baseline performance metrics.

2.?????Data collection of BHA components needs to include information about the tool before and after it is used.?Capturing information about one half of the equation leaves an enormous amount of opportunity on the table when it’s time to examine how #downhole conditions caused changes to the tool.?

3.?????A cloud driven solution is better than a non-cloud solution.?Due to the complex nature of the multiple tool supplier locations and the #data flowing from each one, having a cloud based system makes it easiest to collect the #data as close to the source as possible. Furthermore, #rig activities are becoming more remote driven from a command center. By building #data sets in the cloud, you open many channels to easily pass this information between systems as needed.?

4.?????The process should possess a feedback loop.?Collecting #data alone is not going to solve the problem. Dashboards and BI software need to be used to explore and review the #data on regular intervals. As users begin to explore #data they will begin to come up with new and innovative ways to capture new key data points and track trends. This #data visualization step is also key in holding the #data collection process accountable and ensures clean, reliable information.

5.?????The process needs to be continued.?The idea of implementing a #data solution shouldn’t be looked at as a way to solve a problem, but rather adopted as part of the process that is here to stay. Many times, when economics get tight there is a desire to eliminate costs that don’t directly tie back to physical #drilling. Data solutions sometimes have murky, unclear ways of providing value that may only be measurable over long business cycles.??

These fundamental building blocks will greatly increase the opportunity of finding a #data solution to #BHA components. They shift the process to a pro-active approach enabling the #data to develop into a large source to build solutions off of in the future. When the time comes to find a solution, you will have a tremendous asset to pull from.?