Automated drilling operational plan

Challenge

Historically, well design and drilling operation planning have been hindered by organizational silos and fragmented software systems. Although recent efforts attempt to streamline the process via digitalization, it is still challenging to achieve agile well planning and construction because of the difficulty of revising existing plans and programs based on operational observations or unexpected events.

Innovation

The Automated Drilling Operational Plan (DrillOpPlan) introduces a probabilistic framework to automatically evaluate the design/plan candidates capturing uncertainty, performance and potential incidents using Monte Carlo simulations. The simulation platform that is being developed is a hybrid approach where physics-based models are enriched with learning techniques and historical data. This is crucial for effective uncertainty management and reduction of computational time.  

The framework includes automated agents that simulate decision-making during operations. To represent the different levels of decision making, we consider a hierarchical structure of nested Decision Agents, each having access to a set of Mission Agents, responsible for the execution of a specific mission. .

Value

DrillOpPlan, as part of the workflow for planning, evaluates the most promising candidate wells from Automated Well Engineering, automatically select the most promising drilling program and generate the associated drilling operational plan. This reduces both time and resource demands required to maintain an up-to-date plan during execution. Additional value can be obtained by creating a smooth transition into the execution phase and supporting safer, more efficient well delivery.

Furthermore, DrillOpPlan can also be used as a stand-alone analysis tool in some situations, for example for validation purposes of the framework or as a post analysis tool of already drilled wells.

Status

The framework has been used as a post analysis tool on a North Sea field case of a horizontal multi-lateral well. 

Next steps

• Further develop Automation and Mission Agents
• The ultimate goal is to include operational plan optimization and have an iterative automated workflow from the planning phase to the execution phase.

This work is part of the centres workpackage 1: Agile well-construction workflow: