A hole in tubing for a gas lift operation is pretty difficult to detect before it happens; therefore, today’s case study is particularly interesting. Note, in a previous blog on early event detection, we mentioned that our solution can examine calculated sensors. In the demonstration table above, we have readings on a gas lift valve sensor. Using the gas lift design parameters, and using the casing pressure, we are able to see changes in an apparent operating gas lift valve.
If we view the changes in pressure for this event, we can see that the casing pressure drop is not as alarming. It is dropping from 900 psi to 800-something over a few days, which most likely would not be registered by an operator. But the shift of the apparent operating valve from the 11th valve (a shallow valve) to the 6th valve (deep valve) shows that something has changed. This could be a change in the operating valve, or that the severe drop-in pressure due to a sudden drop in production due to skin formation, is caused by leakage such as a hole in the tubing.
Our AI-based models look at normalized features and can detect a hole in tubing much earlier that operators experiencing these symptoms after the fact. When looking at the data above, you can see that there is an overlap between the time zone in blue, where the tubing failure likely model has detected a hole in tubing, and the grey zone, where the operator has detected the event. Additionally, the data demonstrates through the casing and tubing pressures that there was a tubing integrity test conducted by the operator to start the gray time zone, which was exactly 5 days after our AI-based model detected the hole in tubing.
As we mentioned, it is a pretty difficult event to detect as a hole in tubing for gas lift doesn’t necessarily correspond to a zero-liquid producing well. OspreyData’s models were able to give the operator a 5 day advanced notice which translated to a fairly modest ROI of revenue regained of just under $4000 per well. But, this event only lasted just over ten days. We’ve seen many instances where gas lift wells run for 2 or 3 months with a hole in tubing because the events went undetected, so these impacts can be much more detrimental what this one case may reflect.
If you want to maximize your oil production and increase your ROI on your digital oilfield, contact us to set up an appointment with our sales team to discuss easy implementation. Contact Us Today! Also, if you have enjoyed these use cases for increased ROI, you will want to watch our WebCast Replay entitled, “Top 3 Problems AI Solves in Artificial Lift Production” where Dr. Venkat Putcha, our Senior Data Scientist, and Tim Burke, VP of Operations and Client Services, discuss these cases in detail. Click Here to access our webcast replays.