Page 197 - Intelligent Digital Oil And Gas Fields

Workflow Automation and Intelligent Control 153

Process to be
Motivation automated Select the Define
(Define technology Decion making business
business goals) (Identify (toolkits & CPU) metric
repetitive tasks)
Fig. 5.3 The five main steps to build an automated workflow.

for different engineering applications. Automated workflows not only
reduce the valuable time that an engineer must spend doing repetitive,
data-preparation tasks, but they also ensure consistency in methods, reduce
the probability of input errors, and create a repository of lessons learned and
best practices. By definition an automated workflow is a synchronized inte-
gration of people, processes, and technology. Fig. 5.3 shows the main steps
to build an automated workflow.

5.2.1 Motivation for Automating E&P Workflows
Manual workflows require multiple manual interactions with varied data
sources, analytical calculations, and process models. Engineers often work
from static electronic forms (e.g., reports, pdf, word processing files and
spreadsheets, etc.) that require human data entry and reentry and have mul-
tiple related (but offline of the actual data) communications by email, for
example, to clarify content or approve next steps. In contrast, an automated
workflow integrates client application and/or Web-based dynamic elec-
tronic forms, business processes, engineering analytics and modeling, a com-
mon data model repository (which can automatically access various data
sources), and a self-service workflow application into a comprehensive sys-
tem that does not require interventions by staff and managers. Fig. 5.4 shows
that manual workflows are much less efficient and are prone to errors.
Al-Jasmi et al. (2013a,b,c) quantified a comparison between manual pro-
cesses and automated processes to evaluate and optimize the performance
of a well for electric submersible pumps (ESPs) and gas lift (GL), to model
the artificial lift and optimize the lift performance. On average, the manual
process required 7.2h per well of an expert’s (e.g., senior engineer) time,
whereas the automated workflow required 1.6h per well, by a staff produc-
tion engineer (PE) (less experienced than an expert) with less risk of data or
model errors.

5.2.2 What Kinds of E&P Engineering Processes
Should be Automated?
Workflow automation should focus on any tasks that can be done much
more efficiently by computers than by people. For example, computing