American Oil and Gas Reporter - September 2017 - 39

SpecialReport: Industry In Motion
It captures knowledge that otherwise is
lost when people leave an organization
for whatever reason. It leverages technical
domain experts, including those with
decades of experiences who have learned
from previous successes and failures in
specific fields, reservoirs or operations.
In considering the specific capabilities
an expert system could bring to field operations, one practical example would be
managing one of the industry's longeststanding and costliest problems: produced
water. One of the holy grails in oil and
gas production long has been the ability
to practically and affordably separate
water down hole so that only hydrocarbons
come to the surface. Expert systems could
be the key to enable to reliable downhole
water separation, automatically adjusting
to changing reservoir conditions to continually separate and reinject water into
a nonproductive zone.
Production would be optimized and
there would be less surface equipment,
less corrosion, and less complicated piping,
storage and transportation logistics. The
reservoir even may benefit by experiencing
less dramatic fluid volume changes. This
scenario is realistic and likely in the
future. It is a natural progression from
today's automation to tomorrow's expert
systems. It also is the kind of technological
project that may help inspire a new generation of process engineers and computer
scientists.
Expert System
Building an expert system (Figure 2)
begins with domain expertise that is
painstakingly gathered and engineered
into a knowledge base. Both the expertise
and the engineering are critical to achieving
a meaningful output.
Knowledge is input into expert systems
as a series of rules or patterns that can be
recognized. There may be hundreds of
rules defined, derived from the input of
hundreds of experts, from earth scientists
to drilling engineers and production operators. Rules also can be driven from
data and models, such as reservoir models
and well production profiles.
Like humans, computers have the ability to reason. Humans usually reason
using specific rules, categorization, heuristics, past experiences and expectations.
Computers have no expectations. Instead
they reason based on defined rules, frame
concepts, cases, pattern recognition, neural
networks or generic algorithms.
Consistency is one of the quickest

benefits of expert systems, as they formalize knowledge. Some of the most
mundane tasks can be transferred to expert
systems, standardizing the work and freeing people for higher-value activities.
Technology is not the biggest hurdle
in establishing expert systems, nor is
budgeting. The biggest issue is extracting
knowledge from experts. Expertise is difficult to capture. There is an art to gathering
meaningful and complete answers, beginning with asking the right questions.
It is also critical that knowledge gaps be
identified and filled. Sometimes, the
knowledge is simply not available.
Expert systems already have been deployed successfully in oil field applications,
including fault diagnosis, process hazards
review, emergency response, chemical
spill response and maintenance scheduling.
These applications illustrate the breadth
of benefits that come from this next level
of automation (Table 1).
The first and most obvious benefit is
efficiency. Well-trained machines can
make better, faster, more consistent decisions than humans. For example, in field
maintenance, an expert system can monitor
thousands of sensors and other input devices, look for patterns and identify potential problems before the pumper drinks
his first cup of coffee. The pumper is
then free to focus on those wells that
truly need attention.
Similarly, expert systems can use data
from the well, equipment manufacturers,
other similar wells, global positioning
systems and production operator rules to
recommend the most efficient preventive
maintenance schedule. Field personnel
log less windshield time and spend less
time waiting for parts. Just about everything and everyone can become more efficient with expert systems.
Interestingly, wells also can become
more productive with expert systems.
First, production can be increased by reducing or eliminating downtime. Second,
fine-tuning every aspect of production
operations can result in overall productivity
increases. When expert systems are combined with four-dimensional virtual operation centers, they enable sharing and
synergy of efforts. Decision makers in
the center receive ultrafast expert recommendations and can run fast alternative
scenarios. Operations people in the field
can take immediate action, enabling optimal production.
Another great opportunity is in the
area of safety. When expert systems

TABLE 1
Beneﬁts of Artiﬁcial IntelligenceBased Expert System
*
*
*
*
*
*
*
*
*
*
*
*

Increased timeliness in decision
making
Increased productivity of
technical/operational experts
Timelier and better informed decision
making
Improved understanding of complex
issues
Improved management of uncertainty
Formalization of organizational
knowledge
Faster problem solving
Consistent and easy to replicate
No cost when not used
Better use of staff and time
Integration of multidisciplinary experts
Aids in training novices

enable remote monitoring and control,
they reduce miles driven and the number
of site visits, decreasing the risk of accidents or worksite incidents. Training is
improved and maintenance can be streamlined. In addition, expert systems can
detect and shut off leaks faster than possible with human intervention. They even
can identify conditions that may lead to
leaks and avert an environmental problem
before it can happen.
Most remarkably, implementing the
intelligent oil field presents the opportunity
for renewing young workers' interest in
the industry. This recruitment tool should
not be underestimated. Expert systems
optimize, engineer, operate and control
operations at an entirely new level. Coupled with a trend toward better environmental stewardship, these systems will
play an important role in recruitment and
retention in the years to come.
r
FrEd CzuBBA is founder and president of Ottawa, Ontario-based AXE
Consulting Ltd. He began his career in
1974 as a commissioning engineer at
ABB Lummus, and worked for the company for 26 years in project management
and senior business development positions. Czubba also previously served
as a regional manager at Siemens and
as senior business development manager
of oil and gas, refining and petrochemicals at Phoenix Contact. He is a graduate of Lambton College of Applied
Arts & Technology, Worcester Polytechnic, and the Thunderbird School
of Oil & Gas Management.
SEPTEMBER 2017 39

American Oil and Gas Reporter - September 2017

Table of Contents for the Digital Edition of American Oil and Gas Reporter - September 2017