Optimizing logistics for oil & gas drilling operations
Optimizing Logistics for Oil & Gas Drilling Operations
The Problem:
Land drilling operations consist of several high-level operations performed sequentially with the occasional need for 3rd party services and products to be delivered to the wellsite. Given most drilling operations occur in remote areas, the time to mobilize and deliver these products and services can be many hours or days. There is an imbalanced operational cost to the timing of these 3rd party services. If they are scheduled too early there, is increased cost of demurrage and overtime for these services. If they arrive too late, the drilling rig and crew will incur non-productive time. The prior state was for a human decision maker to manually estimate when a drilling rig will need a 3rd party service, and schedule it accordingly.
The Data:
Our client, an oil & gas service provider was developing a software platform for planning and executing well construction operations. This platform provided equipment sensor data for major drilling equipment from over 50 rigs, including well telemetry, top drive and draw-works sensors, as well as drilling fluid properties and flow rates.
The Solution:
From the given sensor data, we were able to identify contiguous high-level activities, characterize them with intelligent features engineering, and predict the duration of future activities given current well conditions and planned telemetry. The models generating these activity predictions were optimized using a custom cost-function derived from real world financial data, in order to minimize the overall cost of prediction error. The models were then deployed as a part of a schedule planner module in the client's software platform, automatically predicting the duration of activities and alerting the user of when to best schedule 3rd party services.
The Impact:
Due to the accuracy of our models this scheduling tool succeeded in reducing demurrage, overtime, and non-productive time. Drilling rigs where it was deployed realized a 90% decrease in associated costs.