Useful Measures of Exploration Performance
- R.V. Clapp (Arco Oil and Gas Co.) | R.D. Stibolt (Arco Oil and Gas Co.)
- Document ID
- Society of Petroleum Engineers
- Journal of Petroleum Technology
- Publication Date
- October 1991
- Document Type
- Journal Paper
- 1,252 - 1,257
- 1991. Society of Petroleum Engineers
- 5.7.5 Economic Evaluations, 4.1.5 Processing Equipment, 5.7 Reserves Evaluation, 1.6 Drilling Operations, 4.1.2 Separation and Treating, 5.1.5 Geologic Modeling
- 1 in the last 30 days
- 207 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
Uncertainty about the outcomes of individual wells is a complicating factor in the development of performance measures for an exploration program. Using some basic concepts of probability and statistics to deal with this uncertainty, we have developed some tools to help exploration management compare results of an exploration program with expectations in a meaningful way.
The collapse of oil and gas prices in the mid-1980's brought about a radically different investment climate for exploration. The industry had enjoyed a period when the oil and gas price scenarios that most people thought reasonable made almost any prospect look economically attractive. In that world, prospect economics seemed far less important than the explorationist's intuition about the prospect's desirability. Quantifying these beliefs and using historical data to substantiate them seemed unnecessary. After the fall in prices, most managements revised their future price scenarios and found that many of their exploratory plays were uneconomic or marginal. Exploration budgets were cut. We began to look more closely at the economics of individual plays or prospects before drilling. Explorationists' beliefs came under closer scrutiny, and at Arco we felt an urgency to measure how well we were doing with our exploration program, which used a substantial share of limited capital resources. program, which used a substantial share of limited capital resources. Measuring the results or performance of an exploration effort in ways that enhance future decision-making capability is difficult. In this paper, we focus on a few aspects of performance measurement that we believe paper, we focus on a few aspects of performance measurement that we believe have crucial implications for intelligent decision-making in exploration. Here we focus on the estates of key parameters made by the explorationists before a prospect is drilled. We believe that, by far, the two most important parameters are the probability of finding hydrocarbons (the "chance factor") and the reserves discovered if the well is successful. Explorationists have historical data, geologic models, and their own experience as aids, but the estimates they make of these key parameters are largely subjective, resulting from an informed judgment about what is reasonable. The uncertainty inherent in exploration further complicates the measurement of performance. We need to allow for this uncertainty when we compare actual outcomes with our predrilling estimates of key parameters. Most of us do not reject the idea that the probability of heads on the flip of a coin is one-half, even if we flip twice and get two tails in a row. Likewise, it doesn't make much sense to condemn the explorationist who had estimated the probability of success on each of two wells at 50% when both turn out to be dry. We need to use and understand some probability concepts to compare actual and estimated results properly. But probability concepts to compare actual and estimated results properly. But before we look at methodology, let us test our intuition in a hypothetical situation.
Intuition and Exploration Results
Let us assume that our exploration department has drilled 20 exploratory wells in the first 6 months of this year. To keep things simple, let us assume that the explorationists assigned a probability of success (chance factor) of 20% to each of these wells. Also in the interest of simplicity, let us assume that their estimate of the reserves (given a success) on each well was identical and equal to 10 million bbl. It is easy to calculate the chance-weighted value or "expected value" of reserves for each well as 2 million bbl. (We discuss the concept of expected value in the next section). Because we have drilled 20 of these identical wells, we can multiply the expected reserves on each well by 20 and calculate the expected reserves for the total program as 40 million bbl. Let us assume that we actually discovered 24 million bbl alter drilling 20 wells. Thus, our actual reserves discovered are 40% less than the expected amount. This, of course, is cause for concern. But is the result strong evidence that the estimates of our explorationists were biased to the high side, or could the actual outcome reasonably be ascribed to chance (bad luck)? Is it time to overhaul our whole estimating process? Intuition might suggest that it is, but we need to delve into the world of probability to gain the understanding needed to answer these questions properly. properly. Reserve and Chance Probability Assessment Probability Assessment One thing we can be sure of in any exploration program is that outcomes will never be exactly as predicted.
|File Size||840 KB||Number of Pages||6|