UC-784
Recognise Non-Productive Time Patterns Across the Well Set
Reads the non-productive time record across every well OMV has drilled and surfaces the causes that repeat before the next well spuds.
TBDBuild Duration
XXXXIndicative ROI
XXXXBuild Cost
XXXXValue
The Challenge
Non-productive time is coded on the daily report and reviewed per well. Nobody reads across the set, so a cause that has now appeared on three wells is still treated as bad luck on the fourth.
How It Works
- The engine ingests the coded non-productive time record and the daily report narrative across every well.
- It clusters events by cause, formation, hole section and contractor and ranks the causes that repeat.
- It briefs the team before spud with the causes most likely on this well and the mitigation that worked before.
What It Removes
- Causes reviewed one well at a time
- Repeat events treated as one-off bad luck
- Coding done for the report rather than for analysis
Input Data RequirementsDaily drilling reports in the Sodir schema, non-productive time code registers, end of well reports, contractor and equipment records, formation prognosis.
Output FormatRanked repeat cause list across the well set, a pre-spud briefing pack for the next well, and a live alert when a known pattern starts.
“This has cost us three wells running and we are about to do it again.”