Page 204 - Intelligent Digital Oil And Gas Fields

Workflow Automation and Intelligent Control 159

though each well behaves differently than others, the troubleshooting and
remediation process could be unique or slightly different for each situation.
Well issues, mechanical malfunctions, electronic equipment failures, or any
well anomalies have a recommended action plan. Engineers use supporting
tools—such as process models, statistical analysis, data mining technology,
expert systems, pattern recognition, neural network tools, physical models,
as well as knowledge bases of best practices and lessons learned. The auto-
mated workflow should be integrated with these process models to capture
knowledge and create rule expert system.
Learn and improve. The final step in an optimal DOF decision-making pro-
cess is to measure, analyze, and improve the action plan. We believe that all
aspects of a decision-making process—monitoring, analysis, actions, and most
importantly,thepositiveornegativeresults—shouldberecorded.Withcurrent
technology, recording the action plan is easy. The challenge is how to process
the recorded analysis and results, and then incorporate the learnings to update
the automatic decision-making process. Even today no consistent technology
exists to capitalize on lessons learned without human intervention.

5.2.5 Automated Workflow Levels of Complexity or Maturity

Brule et al. (2008) describe five levels of workflow automation maturity and
where the E&P industry is for each of those levels:
• Level 1: Reporting what has happened: reporting systems, common-
place in E&P.
• Level 2: Analyzing why something happened: ad hoc queries, KPIs,
gaining popularity in E&P.
• Level 3: Predicting what will happen or why something might happen:
analytical modeling, full-physics models, and integrated asset manage-
ment (IAM) in E&P (see Chapter 6).
• Level 4: Operationalizing what is happening: continuous update,
time-sensitive queries, and in-database analytics on “billions of rows
of data—Big Data.”
• Level 5: Real-time decision-making to make things happen: actionable
data-driven real-time optimization. Early success with event-driven
closed-loop IAM in E&P.
Today, many technology companies have reached levels 4 and 5 of maturity;
for example, Google, Yahoo, Intel, Facebook, Boeing, Intuit, Amazon,
eBay, T-Mobile, ATT, and others. For many large oil and gas companies,
levels 4 and 5 are the goals for their DOF automation; however, barriers to