Steve Begg

I would rather be vaguely right than precisely wrong: A new approach to decision making in the petroleum exploration and production industry

This article addresses the need for a holistic, integrated approach to assessing the impacts of uncertainty on oil and gas investment decision making. We argue that this cannot be accomplished effectively by just adding a capability to deal with uncertainty to classical, comprehensive, and rigorous models of all the components that contribute to an investment decision evaluation. Furthermore, we suggest that such an approach, even if feasible, is not desirable. Instead, we propose the concept of a holistic and probabilistic approach embedded in a decision-support system. This holistic approach has two major components. One is a technology component that integrates a variety of evaluation and decision-making tools. The second component is a modeling philosophy that fully recognizes the breadth and magnitude of uncertainty. It involves trading off some technical precision and detail for a more complete, accurate, and rigorous assessment of the impacts of uncertainty on the investment decision-making process. The main elements of such a system are simplified component models for each domain; a Monte Carlo simulation engine; and a modeling language for customization, incorporation of interdependencies between components, implementation of decision logic, and updating information as a result of learning. We illustrate how such a system identifies which uncertainties impact the decision the most, values the acquisition of information (data, technical analysis), and encourages flexibility in forward plans to mitigate and/or exploit uncertainties. Further applications are for the optimization of development plans, real options valuation, and the generation of consistent, risked cash flows for input to portfolio analysis. The application of such a system results in a true value-driven focus to the work of multidisciplinary asset teams through its ability to integrate the technical and business aspects of decisions.

Using the value of information to determine optimal well order in a sequential drilling program

Offshore drilling programs on complex reservoirs carry inherent risks. Subsurface uncertainty can lead to costly mistakes being made, and, therefore, being able to gain information during the course of a sequential drilling program, and use it effectively, can have sizeable capital importance. Based only on predicted drilling costs and production rates, wells closest to the platform would commonly be drilled before longer offset wells by virtue of their cheaper drilling costs. However, there is potential during the drilling program for learning to occur between the drilling of two wells that would either provide encouragement to go ahead and drill the second well, change it in some way, or, indeed, remove it from the program altogether. Analyzing the predicted value of this learning, before drilling commences, can reveal the conditions under which the longer offset well could be drilled first because of its potential to impact decisions regarding the second well. The various scenarios and their sensitivities are analyzed using a value of information (VoI) approach, providing an example of how VoI can be used proactively in the construction of learning drilling strategies. The same approach can be extended to most other data acquisition situations.

Uncertainty vs. variability: what's the difference and why is it important?

Technical professionals are often asked to estimate “ranges” for uncertain quantities. It is important that they distinguish whether they are being asked for variability ranges or uncertainty ranges. Likewise, it is important for modelers to know if they are building models of variability or uncertainty, and their relationship, if any. We discuss and clarify the distinction between uncertainty and variability through strict definition, illustrative analogy and numerical examples. Uncertainty means we do not know the value (or outcome) of some quantity, eg the average porosity of a specific reservoir (or the porosity of a core-sized piece of rock at some point within the reservoir). Variability refers to the multiple values a quantity has at different locations, times or instances - eg the average porosities of a collection of different reservoirs (or the range of core-plugs porosities at different locations within a specific reservoir). Uncertainty is quantified by a probability distribution which depends upon our state of information about the likelihood of what the single, true value of the uncertain quantity is. Variability is quantified by a distribution of frequencies of multiple instances of the quantity, derived from observed data. That both are represented by ‘distributions’ is a major source of confusion, which can lead to uncritical adoption of frequency distributions to represent uncertainty, and thus to erroneous risk assessments and bad decisions. For example, the variability of natural phenomena is sometimes well-approximated by normal or log-normal distributions, but such distributions may not be appropriate to represent the uncertainty in outcome of a single occurrence. We show there is no objectively ‘right’ probability distribution for quantifying the uncertainty of an unknown event – it can only be ‘right’ in that it is consistent with the assessor’s information. Thus, different people (or teams or companies) can legitimately hold different probabilities for the same event. Only in very restrictive, arguably unrealistic, situations can we choose to use a frequency distribution derived from variability data as a probability distribution to represent our uncertainty in an event’s outcome. Our experience as educators of students and oil & gas industry personnel suggests that significant confusion exists in their understanding of the distinction between variability and uncertainty. This paper thus provides a resource for technical professionals and teachers to clarify the distinction between the two, or to correct it where it has been wrongly taught, and thereby help to improve decision-making.

Valuation of swing contracts by Least Square Monte Carlo simulation

This paper was accepted for presentation at the 2010 Asia Pacific Oil & Gas Conference and Exhibition held in Brisbane, Queensland, Australia, 18–20 October 2010, and revised for publication. Original manuscript received for review 2 December 2010. Revised paper received for review 2 August 2011. Paper peer approved 15 September 2011 as SPE paper 133044. Summary Natural gas and electricity are commonly traded through swing contracts that enable the buyer to exploit changes in market price or market demand by varying the quantity they receive from the producer (seller). The producer is assured of selling a minimum quantity at a fixed price, but must be able to meet the variable demand from the buyer. The flexibility of such contracts enables both parties to mitigate the risks and exploit the opportunities that arise from uncertainty in production, demand, price, and so on. But how valuable are they? Traditional net present value (NPV), based on expected values, cannot value this flexibility, and the traditional options/valuation techniques could not model the complexity of the terms of such contracts. Taking gas contracts as an example, this paper seeks to (a) raise awareness of how flexibility creates value for both parties and (b) show how least-squares Monte Carlo (LSM) simulation can be used to quantify its value in dollar terms, from the perspective of both producer and buyer. Because the value of flexibility arises from the ability it gives to respond to fluctuations (e.g., in commodity prices), a useful model of swing contracts needs to reflect the nature of these fluctuations.

Would You Know a Good Decision if You Saw One

Ever since the early publication by Grayson, we have seen an increasing interest in decision-analysis in the oil and gas industry. There have been numerous studies and publications discussing methods and models for rational decision-making. Given the inherent limitations of intuition and heuristics, one might expect decision-makers to be delighted with the more consistent approach provided by rational decision-making techniques. Modern decision models can help process large amounts of information without losing valuable pieces. They never suffer from distraction, fatigue, boredom or random error. They are consistent, week after week. An important contributor of the complexity in decisionmaking stems from the fact that human beings are imperfect information processors, we are not always rational. Personal insights about uncertainty and preference can be both limiting and misleading, even while the individual making the judgments may demonstrate an amazing overconfidence. An awareness of human limitations is critical in developing good decision-making procedures. To our knowledge, very little has been published in the exploration & production literature on the “cognitive”; i.e., the “thinking”, social, judgmental and emotional aspects of decision-making. Yet, in spite of these obvious advantages, people treat these with a nearly instinctive distaste. They are particularly resistant to the idea that simple models can validly make such subjective evaluations. For some, that resistance may stem simply from unfamiliarity with the statistics and probability involved. Many more, however, subscribe to the widespread assumption that human judgment is more discerning than a model. They are reluctant to believe that simple mathematical calculations can match the complexity of the human mind. “Take away an ordinary person’s illusions,“ says Dr. Relling in Henrik Ibsen’s Villanden, “and you take away happiness at the same time.” Even today, after decades of research on the psychological aspects of judgment and decision-making, people continue to assume that intuition, repeated experience and their general intelligence will see them through. Unfortunately, intuition and repetition are unreliable teachers at best. Research shows that the less competent people are, the less likely they are to know it. Overconfidence is a deeply rooted human characteristic. Not only do most people tend to hold overly favorable views of their intellectual and interpersonal abilities, but those who are the least accomplished overestimate their performance and ability the most. In other words, those who most need training to improve their decision-making abilities are the least likely to recognize it. Instead, like drunken drivers who are certain that their reflexes are unimpaired, they proceed with the mistaken impression they are doing just fine. A particularly interesting aspect of human judgment and decision-making are the traps we unknowingly step into. Many decision-makers believe that intuition, repeated experience and their general intelligence will see them through. Unfortunately, as will be discussed in the paper, intuition and repetition are unreliable teachers at best. In this paper we will illustrate how a cognitive perspective can offer practical suggestions on how to deal with many 2 BRATVOLD, BEGG AND CAMPBELL SPE 77509 common problems in decision-making. The findings presented here are not novel and we have borrowed liberally from the pioneers and prominent researchers in the exploration of the minds of the decision-maker such as Kahneman and Tversky, Russo and Schoemaker, Thaler, March, and Plous. Yet, given the tendency of geo-scientists and engineers to focus on the “hard” elements of decision-making such as techniques and tools for quantifying uncertainty and risk, and not so much on the qualitative elements such as the cognitive and judgment aspects, we believe there is a need to broaden the understanding and appreciation of cognitive aspects. In this paper we are merely “touching the surface” of this field and we strongly encourage the interested reader to further explore this fascinating field by reading the original references. A good understanding of the behavioral elements of decisionmaking will lead to improved decision quality and, ultimately, to improved corporate performance. Good Decisions Process vs. Outcome Most decision makers focus on outcome. This is not surprising as most organizations reward – or penalize – people based on the outcomes of their decisions. Results are what matter. Indeed, many people believe that good outcomes are synonymous with a good process, that good outcomes necessarily imply that a good process was used. They often assume the converse is true as well: that a poor outcome necessarily signals a poor or incompetent process. Furthermore, if a company’s or a managers track record is based on just a few “big” decisions instead of numerous small ones, a focus on outcomes carries the risk of rewarding good luck – or penalizing bad luck. The best hope for a good decision outcome is a good decision process. The main reason for this is that this forces the decision-makers to focus on what actually is under their control. Three things influence outcomes:

Why are decisions for oil and gas projects not always made the way they ‘should’ be?

The outcomes of many business decisions do not live up to expectations or possibilities. A literature review of neuroscience and psychological factors that affect decision making has been undertaken, highlighting many reasons why it is hard for people to be good decision makers, particularly in complex and uncertain situations such as oil and gas projects. One way to diminish the impact of these human factors is to use the structured methodology and tools of Decision Analysis, which have been developed and used over 50 years, for making good decisions. Interviews with senior personnel from oil and gas operating companies, followed up by a larger-scale survey, were conducted to determine whether or how Decision Analysis and Decision Quality are used and why they are used in particular ways. The results showed that Decision Analysis and Decision Quality are not used as often as the participants think they should be; some 90% of respondents believed that they should be used for key project decisions, but only ~50% said that they are used. Six propositions were tested for why Decision Analysis and Decision Quality are not used more, and the following three were deemed to be supported: • Decision Analysis and Decision Quality are not well understood. • There is reliance on experience and judgment for decision-making. • Projects are schedule-driven. Further research is proposed to determine the underlying causes, and tackle those, with the aim being to improve business outcomes by determining how to influence decision makers to use Decision Analysis and Decision Quality more effectively.

Investigation of permeability change in ultra-deep coal seams using time-lapse pressure transient analysis: A pilot project in the Cooper Basin, South Australia

Very limited literature is available relating to gas production from ultradeep (>9000 ft [>2700 m]) coal seams. This paper investigates permeability enhancement in ultradeep coal seams of the late Carboniferous and early Permian to Late Triassic Cooper Basin in central Australia, using a time-lapse pressure transient analysis (PTA) approach for a pilot well. The gas production history and three extended shut-in periods are used to construct the time-lapse PTA for the study well. A new approach is introduced to construct a permeability ratio function. This function allows the calculation of permeability change resulting from competition between the compaction and coal-matrix shrinkage effects. Pressure transient analysis indicates that gas flow is dominated by a bilinear flow regime in all extended pressure buildup tests. Hence, reservoir depletion is restricted to the stimulated area near the hydraulic fracture. This implies that well-completion practices that create a large contact area with reservoirs, such as multistage hydraulically fractured horizontal wells, may be required for achieving economic success in these extremely low-permeability reservoirs. The permeability ratio is constructed using the slope of the straight lines in bilinear flow analysis. Because of uncertainty in average reservoir pressure, probabilistic analysis is used and a Monte Carlo simulation is performed to generate a set of possible permeability ratio values. The permeability ratio values indicate that coal permeability has increased during the production life of the wellbore because of the coal-matrix shrinkage effect. Permeability enhancement in this ultradeep coal reservoir has offset the effect of permeability reduction caused by compaction, which is beneficial to gas production.