Download A Knowledge System for Decision Support in the “Digital Oil... Roar Fjellheim Abstract

A Knowledge System for Decision Support in the “Digital Oil Field”
Roar Fjellheim
Computas AS
Vollsveien 9, Postbox 482
1327 Lysaker, Norway
E-mail: [email protected]
Abstract
The AKSIO project is developing an integrated operations system to achieve cost efficient drilling and increased recovery from offshore oilfields. The system will provide timely and contextual knowledge for work processes. It supports decision-making by continuously updating an assessment of the drilling situation and ranking the available options for actions. AKSIO supports collaborative
work in teams with members from the offshore platform, onshore operations centre, individual specialists, and suppliers. It links databases, applications, specialist knowledge networks, and real-time data from the field to a visual representation of the work process,
situation, and decision context. The concept is to be demonstrated at an Onshore Support Centre (OSC), equipped with fiber-optic
links to the field and state-of-the-art display and communication facilities.
Introduction
Knowledge and decisions
The management of offshore oil & gas fields, in the North
Sea and elsewhere, is moving towards integrated operations, i.e. integrated work processes (including planning,
drilling & well operations, production, and maintenance),
real-time data from the field, on-shore operation centres,
and massive use of IT for on-line monitoring, analysis,
and decision tasks. This concept is also known as the
“digital oil field” or “e-field” REF. For integrated operations to succeed, it is critically important to provide appropriate and timely knowledge to decision-making personnel involved in drilling operations. The challenge is to
provide support in a situation with decreasing personnel
levels and higher reliance on individual expert resources.
AKSIO focuses on knowledge and IT-enabled knowledge
management (KM) for successful implementation of the
integrated operations concept. Knowledge is recognized
as a critical resource for achieving business results in the
oil & gas industry, but KM is often not well connected to
core work processes (Fjellheim 2003). KM research focuses on issues like communities-of-practice, narratives,
innovation, organizational issues, as well as purely IToriented aspects. In the AKSIO project, Semantic Web
technology is being explored as a foundation for building
more intelligent knowledge support to operations (Gil et
al. 2004).
Improved drilling technology and more cost efficient
drilling are keys to increasing the recovery rate and lifetime of an oil field. Better utilization of drilling data can
be obtained through more intelligent, multi discipline well
construction processes – in planning, operations, operation follow up and support, and post drilling analysis. To
achieve this one needs good data availability, userfriendly systems, and trend-breaking work processes.
Existing systems can be linked together to form the fundament for tomorrow's work processes, each part representing an "island" waiting to be linked to fully utilize
integrated operations, onshore operations support centre
or visualization rooms regarding drilling.
Particular attention is paid to models for rational decision-making in the drilling process. Current practice
overemphasizes short-term benefits (e.g. short drilling
duration), at the expense of longer term benefits (expected lifespan and oil recovery from a well). One reason
is that longer term aspects are not well supported by existing tools and work processes. In a study made for an oil
field in the North Sea, a decision model for value creation
based on real option theory was developed (Hanssen,
Bakken, and Nordby 2003). Conventional value-based
management metrics focus on calendar-year profits, shortterm production and increased reserves booked. Such
measures do not honor the critical elements of risk assessment, decision robustness and real option identification. In the study, the predominant value was created
through assessment of technical feasibility, timing optimization, and tailoring well design.
Oil companies must actively seek, develop, monitor
and act upon their future opportunities, i.e. cultivate their
real options portfolio in an adaptive manner. Real Option
valuation is to appreciate the value of flexibility inherent
in a project. It can be used to benefit from the time flexibility of a decision or, to identify and secure a back-up
location for a high-risk drilling target. To secure that the
project teams constantly focus on robust and flexible
planning, management metrics are recommended to have
a better balance between cash flow monitoring and opportunity monitoring. This is increasingly important as fields
mature (Saputelli et al. 2003).
Design concepts
The major aspects of the AKSIO concept are illustrated in
Fig. 2. Engineers and other decision-making staff at the
OSC perform work tasks as part of certain work processes. To make the best decisions, they access data
sources (historical and real-time data), use specific IT
tools, and interrogate colleagues in knowledge networks
for specific pieces of knowledge. Knowledge must be
timely and contextual relative to the decision task and
work process at hand (Gollery 2002).
Knowledge
networks
IT tools and
data sources
Work processes
Planning and executing a drilling operation is a highly
complex undertaking, and is governed by an overall work
process, as indicated in Fig. 1.
Decision making
drilling team
Real-time
data
Figure 1. The overall work process for a drilling project.
The three first stages of the process are all concerned
with planning, starting with a business proposal to drill
for oil in a certain geological area, and ending with a
detailed plan for the operation. Several types of expertise
need to cooperate to make a high-quality drilling plan,
such as geologists, geophysicists, petrophysicists, drilling
engineers, reservoir geologists, etc.
The operation phase uses the detailed plan to carry out
the actual drilling work. Depending on experience gained
during drilling, the detailed plan may be revised “on-thefly”. Such experience is documented in an experience
data base and can be used as a basis for the reporting
phase as well.
The definition of the drilling work process seems to
indicate a sequence of steps. However, many of the stages
are overlapping, often in order to save time for the subsequent stages. The potential for better integration of planning and operation activities are considerable. In particular, studies have shown that there is a need for improved
feedback of knowledge from the operations phase to
planning of subsequent drilling projects, as well between
the operations phases of different projects.
Figure 2. AKSIO combines knowledge management, decision
support, and real-time data tracking
Functional properties
The services to be offered by the AKSIO system include:
•
Manual and autonomous monitoring of the drilling
process – situation assessment
•
Dynamic and adaptive display of the decision situation and a ranking of available options for actions
•
Solicited as well as unsolicited (“active”) guidance
during performance of work process operations
•
Assistance in identification of needed information
sources (human, databases, documents)
•
Sustained knowledge creation and sharing (between
different fields and work process phases)
Services are to be embedded in the tools already used by
oil company personnel to carry out their daily duties.
Operations scenario
References
The following brief scenario describes how AKSIO will
be used. A team is in the midst of a drilling operation,
working at the OSC. The sequence of drilling actions is
modeled in AKSIO acting as a work process tool, containing information previously documented in a Main
Drilling Program (MDP). AKSIO is kept up to date with
actual drilling by being fed real-time data from the field,
and is therefore able at all times to give the engineers an
up-to-date picture of how actual progress compares to
planned progress.
Fjellheim, R. 2003. Knowledge Management in the International Oil&Gas Industry – Best Practice. 2nd KM Forum
Annual Conference, Oslo, September 2003
At a certain moment, a deviation is detected in drilling
progress (reduced speed). Based on displayed information
and with diagnostic advice from the system, the team
engineers suspect that the cause is an unforeseen change
in geology. Since AKSIO is aware of the actual situation,
it can quickly formulate a focused search for relevant
information in internal databases and experience reports.
The engineers may also wish to discuss the situation
with other experts, and ask AKSIO to locate experts
based on recorded expertise profiles and availability (using “on-line presence” indicators in the IT infrastructure).
Communication can be via telephone, video, e-mail, instant messaging, etc. The collected information is fed to a
decision support module, which actively tracks the options available to the drilling team, and helps in ranking
the options and selecting the best action to take.
Conclusions
Key innovative elements of AKSIO are real-time knowledge and process management, combined with adaptive
decision support. The underlying tools will borrow from
emerging Semantic Web technology. AKSIO will enable
drilling engineers to make the best informed and optimal
decisions, and will bring state-of-the-art decision technology to a new arena.
Acknowledgements
The AKSIO project is carried out by a team consisting of
researchers from Norwegian organizations Statoil ASA,
Hydro ASA, Computas AS, Institute of Energy Technology, Det Norske Veritas, The Norwegian University of
Science and Technology, and the University Graduate
Studies at Kjeller. The project is supported by The Norwegian Research Council through grant PETROMAKS
163365 (2004-2007).
Gil. Y.; Deelman, E.; Blythe, J.; Kesselman, C.; and
Tangmunarunkit, H. 2004. Artificial Intelligence and
Grids: Workflow Planning and Beyond. IEEE Intelligent
Systems, 26-33, January/February 2004
Gollery, S.J. 2002. Context Building Information-Centric
Decision-Support Systems. 2002 Command and Control
Research and Technology Symposium, Monterey, California, June 2002
Hanssen, T.H.; Bakken, E.: and Nordby, L.H. 2003.
Time-lapse Seismic and Real Options: New Measures are
Required to Show Value Creation. AAPG International
Conference, Barcelona, September 2003
Saputelli, L.; Economides, M.; Nikolaou, M.; and V.
Kelessidis, V. 2003. Real-time Decision-making for
Value Creation while Drilling. SPE/IADC Middle East
Drilling Technology Conference, Abu Dhabi, October
2003