Portfolio Management in an Upstream

Oil & Gas Organization

INFORMS Conference
Philadelphia 1999

DA Society Practice Award Finalist Presentations

8 November 1999

Mazen A. Skaf, Ph.D. Donald W. Spillman

Strategic Decisions Group Portfolio Advisor
745 Emerson Street Shell Offshore, Inc.
Palo Alto, CA 94301 One Shell Square
New Orleans, LA 70161
 Agenda

• Background and Needs that Led to The Project

• The Implemented Solution:

– The Portfolio Management Process
– The Portfolio Management System

• Benefits and Impact on The Organization

• Learnings and Conclusions

17-990—INFORMS—8 November 1999 1 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The Gulf of Mexico has been a key growth area for
exploration and production for several oil and gas
companies in the U.S.
• Companies have increased their investments in the Gulf:
– Decline in reserves in onshore areas
– Advancements in offshore exploration and production technology
– Partnerships allow for the sharing of risk and critical expertise

• Companies lease blocks from the Minerals Management Service (MMS)

through participating in lease auctions

• Lease holders must drill prospects before lease expiry (mostly 10-year
terms).
– Discoveries can then be held for production beyond the lease expiry
date
– Lease expirations mean that undertaking the highest value projects
first is not always the best action.
• Shell Offshore Inc. (SOI) is the Shell Oil subsidiary responsible for
offshore exploration and production in the Gulf of Mexico.
17-990—INFORMS—8 November 1999 2 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
SOI has a large portfolio of assets spread across mature
and frontier areas in the Gulf of Mexico.
Mobile

Chandeleur Chandeleur Viosca

Sound Addn Knoll
Breton
Sound Main
Sabine Pass
Pass
Grand
Isle West
South Delta South
West East Marsh South Pass Viosca Knoll
Vermilion Eugene Ship
Cameron Cameron Island Pelto
Island Shoal
High South
Island Timbalier

Brazos
Galveston
Mississippi
Canyon
Matagorda Ewing Bank
Island

Mustang
Island
East Breaks Garden Banks Green Canyon Atwater
Corpus
Christi
North
Padre
Island

Port
South Isabel Alaminos Canyon Keathley Canyon Walker Ridge
Padre
Island

SOI is one of the largest leaseholders in the Gulf of Mexico and owns 15% of
the leased acreage in water depths over 1500 feet.

17-990—INFORMS—8 November 1999 3 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
A typical asset that reaches production goes through a
lifecycle that takes up to ten years, with increasing
investment as it progresses.

1. Screening & Scoping: Lease 2. Purchase Lease:

< $0.5 million $0.2 - $30 million

3. Exploratory Drilling: 4. Development &

$10 - $40 million Production:
$0.3 billion - $1.5 billion

17-990—INFORMS—8 November 1999 4 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
Managing a growing portfolio of opportunities in the
face of resource constraints, lease expiries, and
technical uncertainty has become very complex.
• Some critical resources are often severely constrained:
– Drilling rigs
– Scientific staff for prospect evaluation

• Projects at different stages in the asset life cycle create uncertainty about
future resource needs.

• Asset teams within the same business unit compete and battle for the
same resources:
– In several organizations, there is a bias to overanalyze individual
prospects without understanding portfolio implications
– Difficulty in comparing asset plans across teams without a formal
portfolio management process

• Assets are continually added to the portfolio as new prospects are

identified.

17-990—INFORMS—8 November 1999 5 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
In 1996, the SOI Leadership Team asked us to jointly
develop and implement a portfolio management
process and system that would achieve the following:
• Facilitate the allocation of resources across the portfolio:
– Provide comparability among decisions and assets across the
portfolio
– Help determine optimum allocations of resources to meet various
portfolio objectives
– Evaluate whether present resource levels are appropriate and
support resource planning activities

• Provide insights into numerous portfolio-wide issues:

– Leasing strategy, portfolio balancing, technology investments, etc.
– Identify critical portfolio issues requiring management’s focus

• Be easy to use and update

17-990—INFORMS—8 November 1999 6 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 Agenda

• Background and Needs that Led to The Project

• The Implemented Solution:

– The Portfolio Management Process
– The Portfolio Management System

• Benefits and Impact on The Organization

• Learnings and Conclusions

17-990—INFORMS—8 November 1999 7 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The solution we developed and implemented built on
several years of strategy development projects in the
industry and joint efforts in decision analysis.
• The roll-out of the process and the system required a strong familiarity
with the dialogue decision process
– Additional training was necessary in some cases

• The design and development of the system required extensive

assessments that built on the results of previous joint SDG-SOI efforts:
– For example, assessment of functional relationships that use asset-
specific parameters instead of direct assessment of cost variables
– Required an in-depth understanding of the internal exploration and
development process

• During the last phase of system development, we tested the system in

two separate portfolio strategy efforts

• We formed and trained a portfolio management core team that is now the
keeper of the process and system within SOI
17-990—INFORMS—8 November 1999 8 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 The designed portfolio management process provides a
forum for decision-focused dialogues between senior
management and asset teams.

Decision Team Agree on yearly

Review Preliminary Compare alternative
(Portfolio Decision Evaluation of Team portfolio plans and
business
and Business Unit resource allocation
Review Meetings and challenges
Portfolio Alternatives options
Workshops)

Preliminary Evaluation Final Evaluation

Portfolio Advisor & of Team and B. Unit of Team and B. Unit
Summary Alternatives Alternatives
Portfolio Management of Peer Challenges Alignment on
Portfolio Plan 1
Core Team Review for Teams in Team 1 Portfolio Plan
and Developing Portfolio Plan 2 and Resource
Facilitating the Diagnostic Alternatives Team 2 Allocations
Portfolio Plan 3
Decision-Focused Analysis
Portfolio Plan 4
Team 3
Dialogue

Asset Data Agree on best set Other Business

Asset Base Peer
Teams develop
of alternatives for
Detailed refinement of
Planning
alternative plans chosen plan
Teams Plans Review final evaluation Processes
• Operating plan
• Leasing strategy
Other Activities by Conduct Diagnostic Facilitate
Conduct final Update division • Rig scheduling
the Portfolio Analysis and development of
evaluation for portfolio system and • Staffing priorities
Feedback to Asset team alternatives &
Management Core business unit prepare for operating
Teams and roll-up B. Unit level
Team Decision Team alternatives
portfolio alternatives plan development

17-990—INFORMS—8 November 1999 9 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 The process for managing a portfolio of assets is
iterative and requires the evaluation of several
alternative plans.
Value of Increasing
Resource Resource Levels
ResourceLevels
Levels
97 98 99 00 01 02 03 04 • • • •
Staff Resources
G&G Staff
Petrophysicist
Res. Eng.

Rigs
Gen II
Plan
Plan33 1996 1997 1998 1999 2000 ••••
•••
Gen IV Plan
Plan22 Revenues
OpEx
29
4
36
4
150
22
158
25
148
25
Overhead 4 5 8 23 36
Budget Plan
Plan11 Tax (31) (4) 37 26 (6)
Income (52) (7) 62 43 (11)
Net Cash Flow (4) (38) (4) (201) (377)
W.I. Op Scope Bid Risk Drill Develop

List
List of
of Assets/Opportunities
Assets/Opportunities Alpha Annual CapEx &
Cash Flow
Oil & Gas Production
Gamma
Corridor Play Type Water Current ••••
Depth Stage Poseidon

Alpha GC Amp 3100 2 Zeus

Gamma GC Amp 4200 1
Ithaca
Poseidon GC Amp 1500 4
Zeus GC Amp 4450 1
Ithaca GC Amp 2500 3
•
•
•
Beryte GC Amp 1600 4
Byblos GC Amp 2900 4

• Map Out Opportunities • Develop Alternative Plans • Evaluate Plans

• Check Resource
Requirements and
Constraints
Review/Modify
Plans
17-990—INFORMS—8 November 1999 10 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 Agenda

• Background and Needs that Led to The Project

• The Implemented Solution:

– The Portfolio Management Process
– The Portfolio Management System

• Benefits and Impact on The Organization

• Learnings and Conclusions

17-990—INFORMS—8 November 1999 11 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The portfolio management system supports the
following set of decisions and analyses:
• Evaluation of portfolio strategic alternatives
• Lease sale decisions:
– Evaluating different leasing strategy alternatives

• Asset activity timing decisions:

– When to scope, lease, risk, drill, develop or start producing a specific
asset / collection of assets

• Development configuration decisions

• Ownership structure and operatorship decisions

• Resource level decisions:

– Determine critical staff resources required for a given portfolio strategy
– Facilitate rig contracting decisions

17-990—INFORMS—8 November 1999 12 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 The system architecture builds on a model of the asset
life cycle and the key decisions in the life cycle.
Wildcat
Key Decisions in Scope Bid Risk Drill and Develop
Asset Life-Cycle Appraise

Stages of
Asset
Lifecycle:
P0-1 Viable P1-2 P2-3 P3-4 Commercial Producing
Lead Scoped Leased Drill-Worthy
Discovery Asset
Lead

Farm Out/ Non-

Not
Partner Commercial/
Culled Leased
Opportunity Dry Hole

Farm Out/
Partner
Lease
Expires/
Relinquish Produce
Project Execution
Process

Facility Design
Work

Fabrication

Installation

Start-up
Wildcat Drilling Production
Process: and
Screen Scope Leasing Risking and Appraisal
Process Surveillance

17-990—INFORMS—8 November 1999 13 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The basic structure of the system consists of asset-by-
asset computations.

• The basic unit of activity is an asset

– Stages: Lead, Scoped Lead, Leased prospect, Drill-Worthy, …

• Production logic includes both gas and oil capacity constraints for
processing:
– All wells within a development are treated similarly

• Development options include TLP, FPF, Shallow Water Hub Facility,

Subsea, and other future proprietary development systems:
– Specify tieback location for subseas
– Specify processing capacity (oil and gas) for hubs

• The system is linked to decision analysis software packages for

conducting range sensitivity analysis and probabilistic analysis.

17-990—INFORMS—8 November 1999 14 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The system creates a plan for a single asset team or an
entire business unit and uses portfolio globals to
compute total economics and resource use.
Production ECON (Economic Evaluation sheet)
General Inputs ’97 ’98 ... ’97 ’98 ’99 ...
By Asset:
• Tax Rate & Royalty • New Well Drilled • Revenues
• WACC • Costs
• Portfolio-wide Uncertainties • .....
Portfolio-level • Financial Parameters • Unconstrained Oil •
Assessments • Cycle Times (defaults) Production
& Probabilities • Unconstrained • Net Cashflows
• Platform & Well OpEx Gas Production
• Platform & Well CapEx • Resource
• ....... Requirements
• Functional Relationships • Constrained Oil
• Resource Consumption Rates Production
• Oil & Gas Price Premise • Constrained Gas
Production
Calculated Plan*
Initial Resource Plan Inputs (Decisions) Name W.I. Scope Bid Risk Dev.
Guidelines Alpha
I. Working Interest Gamma
Options for II. Activity Timing Poseidon
Asset Timing Decisions Cesar
from Corridor Omega
III. Development, • SUMM (by Asset Measures)
Teams configuration, capacity, •
etc. • Resources Required Financial •Resources Required
Measures •Bang-for-the-Buck Measures
Staff and Trade-
offs
* Reflects cycle times, ensures development
Asset Inputs Rigs
start dates for subseas in synch with
capacity becoming available at hubs . . .
Capital Budget
I. Asset-specific parameters
•Reserves
•Probabilities
From •GOR, # of Wells
Corridor •TVD, Water Depth
Teams & •. . . Yes Yes
Asset I II Select plan and
Database link to
operating plan
No No
model

I. Is portfolio plan the highest-

value creating given:
Adjust Portfolio Plan / Consider Another Plan •Resource constraints
•Value trade-offs
II. Are all constraints met? If no, is
it worth it to relax unmet
constraint(s)?

17-990—INFORMS—8 November 1999 15 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
Today’s discussion will touch upon a sample of uses
and features of the system.

• Tracking measures across the portfolio by asset or any group of assets

• Allocating resources and using bang-for-the-buck measures

• Identifying key value drivers

• Business unit strategy development

• The use of the system as a workbench for decision analysts to conduct

tailored analyses such as valuing a technology investment.

17-990—INFORMS—8 November 1999 16 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The system calculates expected requirements of
various resources, for each portfolio plan, and
compares requirement levels to resource availability.
Expected Usage and Availability of Resources X and Y
for Region ABC Unconstrained Portfolio Plan

Yearly Requirements for Resource X

Yearly Requirements for Resource Y
Yearly Available Levels of Resource X
Yearly Available Levels of Resource Y
Resource X

Resource Y
1996

2000

2002

2004

2006

2008

2009

2010
1997

1998

1999

2001

2005

2007
2003

17-990—INFORMS—8 November 1999 17 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 Given resource constraints, the system facilitates the
development of an optimal plan through prioritization
of assets within each stage of the life cycle.
Corridor ABC Reference Case Corridor ABC Constrained Alternative
Expected Requirement and Availability of Resources Expected Requirement and Availability of Resources
X and Y for Corridor ABC Reference Plan X and Y for Corridor ABC Constrained Alternative

Yearly Req. of Resource X

Yearly Req. of Resource X
Yearly Req. of Resource Y
Yearly Req. of Resource Y
Yearly Available Levels of X
Yearly Available Levels of X
Yearly Available Levels of Y
Yearly Available Levels of Y

Resource X

Resource Y
Resource X

Resource Y
2007

2008

1996

1999

2000

2001

2007
1996

1997
1998

1999
2000

2001
2002

2003

2004
2005
2006

2009

2010

1997

1998

2002

2004

2005

2006

2008

2009

2010
2003
17-990—INFORMS—8 November 1999 18 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The portfolio management system supports range
sensitivity analyses on any set of portfolio-wide and
asset reserve uncertainties.
Expected PV of NCFAT 1996-YYYY, $MM

Cycle Time for Activity X (years)

CAPEX for Item ABC
CAPEX for item DEF
Cycle time for Activity Y
Reserves of Asset x7
Costs for Process Z
Reserves for Asset y5
Cycle Time Activity W
OpEx Variable for ...
Reserves of Asset s8 Low
Operational Parameter Alpha High
Cycle time for Activity Q
Reserves of Asset t4
Improvement in Use of Resource XD
Reserves of Asset t6
Reserves of Asset x3
Reserves of Asset s4 Disguised for
Reserves of Asset y7
Reserves of Asset s10
confidentiality purposes
Reserves of Asset x2
Base Value: $X,XXX

17-990—INFORMS—8 November 1999 19 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 Agenda

• Background and Needs that Led to The Project

• The Implemented Solution:

– The Portfolio Management Process
– The Portfolio Management System

• Benefits and Impact on The Organization

• Learnings and Conclusions

17-990—INFORMS—8 November 1999 20 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
The portfolio management process and system have
provided many benefits to decision-makers and
facilitated the creation of significant shareholder value.
• Enabled a systematic approach to business unit strategy development:
– Senior management and asset teams are both engaged in the process
– Commitment to action increases as a result of shared understanding
of the value implications of the different alternatives
– Decisions impacted investments in the hundreds of millions of
dollars

• Cut the cycle time in developing regional strategies (team-level

strategies) from about three months to two weeks in some cases.

• Replaced non-value-based decision rules driven by expiry dates or other

urgencies.
• Helped management ensure implementation success by tracking resource
requirements.

17-990—INFORMS—8 November 1999 21 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
Since mid 1997, the portfolio management process and
system have been integrated in SOI’s business
processes and used in strategy development and
implementation.
• Annual strategy renewal and budget building for SOI are facilitated by
using the process and system
• The process and system are an integral part of day-to-day decision-
making:
– Planning of drilling inventory, leasing decisions, divestitures, etc.
• This has led to keeping the system up-to-date on a continuous basis:
– Enabled the organization to examine any urgent investment decision
or acquisition opportunity from an overall portfolio perspective.
– Encouraged asset team leaders to consult frequently with the
portfolio advisor.
• The redesign of the decision-making process at multiple levels of the
organization has led to significant cultural change.

17-990—INFORMS—8 November 1999 22 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
 Agenda

• Background and Needs that Led to The Project

• The Implemented Solution:

– The Portfolio Management Process
– The Portfolio Management System

• Benefits and Impact on The Organization

• Learnings and Conclusions

17-990—INFORMS—8 November 1999 23 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.
Successful implementation of a decision analytic
solution for ongoing use requires special attention to
the needs of multiple stakeholders in the organization.

• “It is easy to make a case for action for implementing portfolio

management at the strategic level, but to keep the data up-to-date, you
need to link the portfolio management effort to implementation and
tactical level decisions.”
– We recognized this upfront and designed the solution accordingly
• A DA solution has to be useable at multiple levels in the organization by
design.
• The total solution consisted of:
– The tailored process
– The system including dedicated analysis modules
– Training, user support, and detailed user manual
– Other user-specific applications that were built to link to the system.

17-990—INFORMS—8 November 1999 24 © 1999 by M. A. Skaf and D. W. Spillman. All rights reserved.