EXPLICATING A GUT FEEL-BENCHMARKING THE CHANCE FOR EXPLORATION SUCCESS

Barry A. Goldstein
Parker & Parsley Australasia Ltd.
Sydney

There is a 50-50 chance for adequate trap, seal, source, reservoir, and the likelihood for oil. So, the probability for a successful oil find is 3%. Brief and to the point? Yes. Comparable to other estimates? Frequently, not. Obscure? Yes.

Standards can be set not only to insure that subjective assessments of exploration prospects are to a large degree comparable, but also to explicitly demonstrate that no relevant criteria have been the subject of neglect in the prospect assessment process. Indeed, expansion into new ventures with finite exploration budgets is incentive to formalize the prospect risk assessment process, no matter how reputable and well experienced are the available expert opinions.

A checklist of 33 inquiries is offered as a yardstick for measuring the adequacy of petroleum prospect factor conditions. Each of the inquiries is a part of an overall test that good prospects should pass. Application of the checklist results in probability assignments to the chance for commercial exploration success. The length of the checklist for factor assessment may at first appear daunting for what experienced explorationists commonly regard as a simple, quick risk assessment task. However, a thorough approach introduces an essential element of uniformity between prospect evaluations. The proposed approach goes to some length to explicate consequential judgments that otherwise may remain unrecorded.

The checklist can be a valuable addition to farm-in and farmout discussions. Buyers and sellers are offered some common ground to mold perceptions.

INTRODUCTION

Explorers are bound to ignore Samuel Goldwyn's advice, "Never prophesy, especially about the future." Nonetheless, realistic estimates of "probabilities for commercial success" (Ps) are essential to exploration strategies. Successful exploration strategies will match or exceed forecasts. A method is offered, one that can help provide comparable expected values for diverse prospect types.

Dissection of the prospect assessment process can reveal an appropriate suite of inquiries. "Unless one starts with a clear understanding of the right questions, one's analysis will inevitably be determined by whatever information is readily available." This counsel was focused on financial analyses but is nonetheless relevant to the petroleum prospect assessment and selection process.

The right questions should lead to a valuable predictive ratio; one that accurately reflects the probability that a perfectly positioned wildcat will reap commercial rewards. Indeed, an explicit set of appropriate questions can stimulate proactive acquisition of information that facilitates good judgments. Put into practice, this assessment method constitutes a do-it yourself prospect police. A small improvement in the accuracy of predictions can have a positive effect on corporate profits.

STANDARD INQUIRIE INTO FACTOR FAVORABILITY

An explicit checklist of 33 standard inquiries is offered as a yardstick for the measurement of factor favorability (Table 1). Each inquiry is a part of the test that good exploration prospects should pass. The favorability (adequacy) of each factor is rated on the basis of explicit standards. Application of this checklist should help to reduce subjectivity and provide a valuable level of consistency between prospect assessments.

The checklist is designed to be global. Where the keys to success for specific local play types are assessable, one should tailor these inquiries (and associated grades) to parallel the definition of locally effective factor conditions. Factor grades are not sacrosanct; definitions of effective factor conditions can vary (from basin to basin) for a generalized prospect description.

EXPLORATION FORECASTS

A few observations follow. These pertain to exploration forecasts:

An explorer's business is most always to be wrong.3
Optimism tends to be in inverse relationship to local knowledge.4
Probability judgments should be clearly documented.5
Compare outcomes to predictions to re-calibrate the appraisal system.6
We need to improve on the reliability of reserve forecasts and I am not ready to admit that we are unqualified to do so.7

In general terms, the petroleum exploration industry has not yet forgone the "one-rock method" (Fig. 1). The comparative worth of exploration ventures is still judged with aggregate estimates such as expected monetary values.

INDUSTRY STANDARDS

Many estimates lie behind every volumetric and monetary measure of merit ascribed to exploration ventures. Industry-standard cornerstone estimates are (1) discounted cash flow forecasts, and (2) estimates of probabilities for commercial success. These two cornerstone estimates are combined to define industry-standard expected values. As portrayed by Rose, the expected value (EV) calculation is simple and elegant:

EV = (Probability success x Project present value) + (Probability failure x Cost of failure)

The challenging task of estimating project present values is not addressed in further detail in this article. Estimates of Ps are the focus of the remainder of this text. Forecasts for reserves growth via exploration expenditure are an essential planning tool and should reflect realistic expectations. Valuable predictions for commercial exploration success should reasonably match the results of prolonged, multiwell drilling programs that cover a variety of play types.

PROSPECT ASSESSMENT CRITICAL PATHS

When faced with a new venture opportunity, most explorationists will undertake a loose screening assessment to reach rough favorability ratings. These preliminary investigations and subsequent undocumented prospect favorability assessments are frequently called a "gut feel."

An expert's gut feel about a particular exploration project may be an excellent distillation of all pertinent factors that define the attributes of a model play type. The inexplicit nature of any such brief description will nonetheless make comparisons between prospective ventures somewhat obscure.

The explication of what may otherwise be a somewhat obscure gut feel will hopefully lead to an incremental improvement in the exploration prospect selection process. High quality and comparable exploration prospect assessments are target outcomes.

The method will add value to the prospect evaluation and selection process, despite the seemingly daunting procedure for a simple risk assessment task.

The prospect assessment process can be segregated into three general but frequently iterative phases which are:

Data compilation and interpretation.
Prospect value and quality assessments, e.g. reserve and probability for success estimates.
Commercial forecasts and decision-making, e.g. comparing expected values and choosing commitments.

This process is dynamic. Non-geological factors (excluding tax regimes) may influence forecast expected values and stimulate geological analysis of basins previously ignored. Two outcomes are common in embryonic prospect assessments:

Geologic prospectivity, price, politics, or other prohibitions are so adverse that in practice very little data are compiled and interpreted before resolving to go no further.
The opportunity is perceived to be enticing and further assessment is considered worthwhile, and proceeds based on a gut feel.

The balance of this article focuses upon the rigor of estimating probabilities for potential commercial exploration success, or in other words, the explication of a gut feel.

The juxtaposition of perceptible factors that combine to define a proven to be profitable petroleum discovery is listed along with a scaled list of factor conditions that range from perfect to absolutely unfavorable. The ratings of each factor are combined into a single ratio designed to be proportional to realistic forecasts for exploratory success. The results of exploration drilling need be regularly revisited to locally recalibrate the method.

HISTORICAL RATES OF SUCCESS

Benjamin Disraeli said "there are lies, damn lies and statistics." Drilling statistics for the offshore Australia Carnarvon basin are a case in point. The Barrow and Dampier sub-basins contain all 37 commercial discoveries in the "greater" offshore Carnarvon basin (Fig. 2). The line that circumscribes those oil and gas fields defines a sweet spot wherein three in every 10 exploration wildcats drilled (to yearend 1992) were commercial discoveries.

Outside that sweet spot, the historical rate of profitable exploration is nil. Alternatively, it is "correct" to claim a 21% rate of commercial exploration success for the whole of the "greater" offshore Carnarvon (Fig. 2). A rate of 20% commercial exploration success in the "greater" Carnarvon is probably achievable, but not everywhere in that broadly defined setting. Stellar rates of success will be predicated upon a rigorous prospect assessment and selection process.

Obviously, the most useful drilling statistics are definitive. Commercial cutoffs and basin boundaries should be defined for specified play-types. The term (years) of reference can also affect drilling statistics. Most wildcat drilling statistics do not differentiate between wells cased on the basis of incremental economics (money forward, excluding sunk costs) from those that are profitable after accounting for all prior associated costs. This is an accepted, practical limitation for two reasons. First, few explorers ever reveal full-cycle project economics. Second, it is easy to distinguish between field discoveries and wildcats outside development schemes. One may also assume that wildcats are usually cased where and when expected rewards exceed the costs of associated production.

There are anomalies to this definition of a commercially successful wildcat. Some cased wells may perform below commercial limits. Bypassed pay in plugged and abandoned wildcats is a source of anomalies on the other side of the ledger. Historical rates of wildcat success are nonetheless practical constraints on the credulity of exploration forecasts.

Commercial rates of exploration success (discoveries/wildcats x 100) range upwards from zero to extraordinary levels. The long-haul, global, frontier rate of commercial wildcat success is probably below 20%. A few published bottom-line accounts for exploration outcomes are provided in Table 2. The high-end range o commercial success sustained in Australia speaks well for those explorers, and the basins.

Play-specific exploration rates of success can vary considerably within a single basin. Productive basins almost always contain both sweet spots and moose pasture. Play-specific rates of success also oscillate. The combined success percent for two play types in the U.S. Powder River basin during 1956-72 (U.S.) reached a peak of 21 % soon after the initial discovery and then gradually oscillated between 5-14%. The trends to lower rates of success has been attributed to spates of overoptimism."

EXPLORATION YARDSTICKS

Post-drill autopsies are excellent foundations for future exploration. The makings of a commercial discovery provide models to emulate. Likeness to proven successful play-types is a valuable prospectivity scale. Perceptible attributes of profitable discoveries define models we seek to match. Dry holes are the foe; know the enemy. This is the "analog method."

Knowing what does and what does not work one can develop an explicit set of meaningful questions (in regard to prospects). Making necessary questions explicit poses no harm and can be beneficial. Nonetheless, the following is an apt preamble to the list of gradational factor conditions in Appendix 1 .

"Back in the dark ages, a man whose name was not recorded in history spent a year dissecting the carcass of an elephant. He did not do it very well and contributed nothing to the knowledge of anatomy. However, he achieved a measure of local fame for being willing to tackle a big and odorous job."

Hendricks 11 introduced a systematic approach to the "analog method" with this anonymous fable, recognizing the perils of preaching uniformity to intelligent professionals. A routine set of questions need not stymie professional individuality. Indeed, a detailed checklist (Table 1) of important exploration factors can stimulate proactive acquisition of consequential information not already available.

The gradational descriptions presented in Appendix 1 represent a standard yardstick for measuring factor favorability. The grades ascribed to factor conditions are intended to be a guide; not a proscription. Teams, divisions, and corporations will debate and agree on a level of commonality in rating prospect factor attributes. That list of factor conditions should be calibrated to suit local needs. Indeed, frequent recalibration of the method is recommended. An alternative, more generalized scale of factor adequacy is shown in Table 3.

Each of the 33 inquiries detailed in Appendix I pertains to a single factor. Each factor relates to one of five key criteria (trap, seal, source, reservoir and the likelihood of oil versus gas).

The five key criteria (trap, seal, source, reservoir, and the likelihood for oil versus gas) are "fuzzy sets," and the factors are "fuzzy sub-sets." The least favorable factor attribute that has a bearing on each of five key criteria is carried as the "weakest-link" favorability rating for the associated key criteria. The perceived quality of key criteria and factor conditions are graded with linguistics variables. These "weakest-link," logical outcomes are the intersections of fuzzy subsets as defined by Zadeh 12 and lucidly explicated by Chen and Fang." Factors are not weighted and this follows the recommendation of Rose.8 "Weakest-link" ratings for the five key criteria are then entered into a serial multiplication.

THE CHECKLIST

Using Appendix 1, the attributes of a prospect are graded and then tabulated on the checklist (Table 1). The minimum factor rating assigned to any one of several factors that fall under one of the five key criteria is the probability assigned to the associated, dependent key criteria. The five key criteria are summarized in Table 4:

Serial multiplication of key criteria ratings is designed to be proportional to the aggregate probability geological adequacy (Pg) follows.

(1) trap * (2) seal * (3) source * (4) reservoir = Pg

A fifth key criterion is used to account for the probability that "at least" expected reserves will be found. This explicates the risk associated with reserves sales schedules assumed in present value scenarios. Appendix I provides a basis to establish the likelihood of oil versus combustible gas that may suit areas where gas markets are limited. Different inquiries can be tailored to suit other risks such as high levels of noncombustible gases. The aggregate probability geological adequacy (Pg) is multiplied by the fifth key criterion to estimate the probability for commercial success as expressed with an expected value scenario.

An abysmal (so low as nil) rating can result from the above process, though any such negative perception should be well-founded through the inquiry process. A sand-prone, very permeable stratigraphic setting poses such an extreme example, where seal adequacy is perceived to be dubious. Seismic interval attribute analysis will play a crucial role in such an assessment. Local success and failure may be cause to revise the range for descriptions of factor conditions as detailed in Appendix 1.

The product of adequacy ratings for these five key criteria provides a ratio designed to be proportional to the probability that a perfectly positioned wildcat will reap commercial rewards. As a general rule, wherever an aggregate probability for success (for gas or oil) exceeds 0.15, the assessment should include a historical basis for that claim. Put into practice, this method constitutes a do-it-yourself prospect police.

APPLYING THE CHECKLIST

The list of inquiries detailed in Appendix 1 and the checklist (Table 1) have been used to assess diverse play types. Details of three case studies are provided in the full paper from which this article was excerpted." The assessment of a prospect in a beckoning frontier, the Browse basin off Australia, is next recapped here in brief.

There are no commercial discoveries in the southern Browse basin. Significant gas accumulations have been established with Scott Reef 1 (12 tcf) and Brecknock 1A (5 tcf), but neither is currently commercial. The only significant oil shows in the southern Browse basin are reported from the Cretaceous in Caswell 1 and Caswell 2. A 15% probability for success is a prudent upper limit in this setting.

BROWSE BASIN CASE STUDY

The Galapagos feature (Fig. 3) is summarized below:

Trap rating: 0.70-Four-way, dip-closed culmination related to transpression.

Seal rating: 0.70-Expect 100+ m of basal Lower Cretaceous marine shales with some chance for interbedded sandstones over the primary, Upper Jurassic sandstone reservoir target.

Source rating: 0.45-Triassic-age, marine facies are probably at optimum organic maturity in undrilled half-grabens immediately down dip from the Galapagos prospect. The adequacy of these source rock kitchens is speculative.

Reservoir rating: 0.50-Jurassic sandstones: 15-75 m net reservoir with 15-20% PHIE.

Chance for oil rating: 0.30-Traces of oil are reported from local wells.

Probability for geological adequacy (Pg): 11 %.

Probability for commercial success (Ps): 3.3%.

Expected value (per NPV @ 15% and 0.74 A$/U.S.$) (U.S.) $4,692,000.

In brief, the least favorable attributes of the Galapagos prospect are oil source capacity and the likelihood of expected reserves. Source rock capacity to yield oil is perceived to be the key risk in this case study. At present product prices, Galapagos probably needs be an oil (rather than gas) accumulation to be commercial. Gas has been the most common trap-filling hydrocarbon found to date in the Browse basin.

Post-Lower Jurassic strata are considered insufficiently (organically) mature to account for much if any oil expulsion in the "fetch" to Galapagos, and favorable source scenarios depend on the oil expulsion capacity of the Triassic-age half-grabens that lie downdip from the Galapagos prospect (Fig. 3).

Available evidence for the quality of Triassic Northwest Shelf source rocks is shown in Fig. 4. Assuming an aggregate of 100 m of shale, with average hydrogen index of 350, the plausible ratio between oil migrated towards Galapagos, and the potential OOIP of Galapagos, is between 10 and 20. This is cause to assign a 0.45 grade to the adequacy of expelled volumes versus trap capacity (Appendix 1-factor 3C).

CHANCE FOR COMMERCIALITY

Factor favorability assignments for the Galapagos prospect are shown in Fig. 5.

Well defined structural traps have a tangible quality feel. Source rock proclivity to yield oil in a frontier setting is always an intangible quantity, even if analogies can be drawn with "global" oil source kitchens. Triassic strata on the Northwest Shelf of Australia may contain significant oil source capacity, but proof has yet to be published."

A plausible, profitable development scenario for a gas find at Galapagos could suffice to lift the chance for success more than threefold, up to 11 %.

BUYER BEWARE, SELLER DILIGENT

The suggested "greater-than-one-rock" (Fig. 1) method is the standard approach to gauge prospectivity. A somewhat more explicit checklist of inquiries to gauge factor-specific conditions has been offered. Disciplined application of such a checklist can insure that no consequential "stones-are-unturned" in the prospect assessment and selection process.

Taken one step further, histograms of factor ratings can be used to illustrate the risk profile of prospect attributes (Fig. 5). These profiles are graphic means to explicate the relativity between "gut feels." The relative merits and weaknesses of quite different play-types are thus, made comparable. The 'favorability curves" defined by Chen and Fang" are alternative graphic means to display a "gut-feel." Indeed, the methodology offered by Chen and Fang" inspired this effort.

Assessment of critical exploration factors should remain within practical limits.

The checklist is not designed to be a guide to academic research. A checklist methodology will not add measurably to prospect assessment costs but may add value, at least in terms of consistency. Perhaps the checklist is too short.

CONCLUSIONS

Application of the proposed guidelines can benefit the prospect assessment and selection process by:

Reducing subjectivity in judging the relative certainty of play type ingredients,
Boosting internal consistency between exploration prospect assessments, and
Providing inexperienced explorationists with an explicit basis to guide prospect assessments.

A strategy to achieve higher commercial exploration success rates should include benchmarks for the prospect risk assessment process.

ACKNOWLEDGMENTS

Thanks are extended to Bridge Oil Ltd., in particular Greg Roder for directing that this work be undertaken. Gratitude is extended to BHPP, in particular Peter J. Watson, for providing their adaptation of the Kent-Sherman scale. Consent to publish the case-study account was graciously provided by the WA-207-P joint Venture. Editorial advice from Andy Rigg (APEA-Ampolex) and Murray Johnstone (APEA) was very valuable.

REFERENCES

1. The Economist, 150 Economist years. The future surveyed, September 1993, p. 3,

2. Butters, J.K., et al. (ed.), Case problems in finance, 8th edition, Richard D. Irwin Inc. (publisher), 1981, pp. 3-15, in Goudzwaard, M., Finance for non-financial managers, course notes, 1993, Sydney.

3. Capen, E., Dealing with exploration uncertainties, in Steinmetz, R. (ed.), The Business of Petroleum Exploration, AAPG, Tulsa, Okla., 1992.

4. Ivanhoe, L.F., Oil discovery index rates and projected discoveries, in Rice, D.D. (ed.), Oil and gas assessment-methods and applications, AAPG Studies in Geology No. 21, 1986, pp. 77-84.

5. Bird, K.J., A comparison of the play analysis technique as applied in hydrocarbon resource assessments of the National Petroleum Reserve in Alaska and the Arctic National Wildlife Refuge, in Rice, D.D. led.), Oil and gas assessment-methods and applications, AAPG Studies in Geology No. 21, 1986, pp. 133-142.

6. Sluijk D., and Parker, J.R., Comparison of pre-drilling predictions with post-drilling outcomes, using Shell's prospect appraisal system, in Rice, D.D. (ed.), Oil and gas assessment-methods and applications, AAPG Studies in Geology No. 21, 1986, pp. 5558.

7. Weeks, L.G., Potential petroleum resources - classification, estimation and status, in Haun, J.D. (ed.), Methods of estimating the volume of undiscovered oil and gas resources, AAPG Studies in Geology No. 1, 1975, pp. 31-49.

8. Rose, P.R., Chance of success and its use in petroleum exploration, in Steinmetz, R. (ed.), The business of petroleum exploration, AAPG, Tulsa, Okla., 1992.

9. Common quotation attributed to Benjamin Disraeli, in The Oxford Dictionary of Quotations, 3rd edition, Oxford University Press, London, 1979.

10. Drew, L.J., Oil and gas forecasting-reflections of a petroleum geologist, International Association for Mathematical Geology - Studies in Mathematical Geology NO. 2, Oxford University Press, New York, 1990.

11. Hendricks, T.A., Estimating resources of crude oil and natural gas in inadequately explored areas, in Haun, J.D. (ed.), Methods of estimating the volume of undiscovered oil and gas resources, AAPG Studies in Geology No. 1, 1975, pp. 19-22.

12. Zadeh, L.A., Fuzzy sets, in Information and Control, Vol. 8, 1965, pp. 338-353.

13. Chen, H.C., and Fang, J.H., A new method for prospect appraisal, AAPG Bull., Vol. 77, No. 1, 1993, pp. 9-18.

14. Goldstein, B.A., Explicating a gut feel - Benchmarking the chance for exploration success, APEA journal, Vol. 34, 1994, pp. 378-404.

15. Bradshaw, M., Unpublished comment following presentation of Bradshaw, M., The development of Australian Phanerozoic basins, Program & Abstracts, Selwyn Memorial Symposium, Melbourne, Sept. 30, 1993.

BIBLIOGRAPHY

Bureau of Resource Sciences, Oil and gas resources of Australia-1992, published by the Petroleum Resources Branch, Canberra, 1993.

Foreman, D.J., and Hinde, A.L., Improved statistical method for assessment of undiscovered petroleum resources, AAPG Bull., Vol. 69, No. 1, 1985, pp. 106-118.

Goldstein, B.A., Waxings and wanings in stratigraphy, play concepts, and perceived prospectivity in the Canning basin, APEA journal, Vol. 29, 1989, pp. 466509.

Heath, R., Exploration in the Cooper basin, APEA journal, Vol. 29, 1989, pp. 366-378.

Hedberg, H.D., False precision in petroleum resource estimates, in Haun, J.D. (ed.), Methods of estimating the volume of undiscovered oil and gas resources, AAPG Studies in Geology No. 1, 1975, p. 160.

Horace, pre-65 B.C.-Epistles, Book III (Ars Poetica), in Oxford Dictionary of Quotations, 3rd Edition, 1979, Oxford University Press. Langford, F.F., and Blanc-Valleron, M.-M., Interpreting Rock-Eval pyrolysis data, AAPG Bull. Vol. 74, No. 6,1990, pp. 799-804.

Mackenzie, B.W., and Cai, Z., Economics of petroleum exploration in Australia, Australian Mineral Foundation, 1993, p.27.

Meyer, M., and Butcher, J., Eliciting and analyzing expert judgement-a practical guide, Academic Press, London, 1991.

Newendorp, P.D., Decision analysis for petroleum exploration, Pennwell Publishing Co., Tulsa, Okla., 1975.

Thompson, M.A., Success rates in the Carnarvon basin, unpublished Bridge Oil Ltd. analysis, 1993.

Watson, P.J., Sherman-Kent scale, unpublished correspondence from BHP Petroleum to Bridge Oil Ltd,, 1993.

Issue date: 12/26/94