Goto

Layer | Fill | Outline |
---|

Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|

Fill | Stroke |
---|---|

Summary When exploring nearby prospects in a common area, the outcome of drilling a well can change the chance of success in nearby prospects, affecting their economics and drilling decisions. Here, besides possibly discovering hydrocarbons, a single well could also supply information about other wells. For such a cluster of exploration prospects, which well should we drill first, and which next? More importantly, what is the economic value of this group of prospects? The answers are multidimensional; they depend, at least, on geological dependencies and economic dynamics. Because it takes time to interpret each drilling outcome and update our understanding regarding neighboring prospects, the varying hydrocarbon prices also affect the economics of the upcoming wells. Therefore, our sequence of drilling decisions should consider both geological dependencies and uncertainty in prices. In this paper, we develop a valuation model for a group of interdependent prospects. We use a dynamic programming model that combines the binomial representation of prices with the conditional probability of success or failure at each drilling site. The software implementation of the algorithm accompanies this paper and is a useful valuation and decision‐support system.

Artificial Intelligence, asset and portfolio management, assumption, Bayesian Inference, Binomial Lattice, cash flow, decision support system, discovery, drill cost, Drilling Decision, drilling operation, exploration, forward curve, forward price, geological dependency, information, machine learning, NPV, october 2020, price process, probability, prospect, Reservoir Characterization, risk management, Scenario, Upstream Oil & Gas, valuation

Country:

- North America > United States (0.46)
- Europe (0.46)

SPE Disciplines:

Technology:

- Information Technology > Decision Support Systems (1.00)
- Information Technology > Artificial Intelligence > Representation & Reasoning > Uncertainty > Bayesian Inference (0.66)
- Information Technology > Artificial Intelligence > Machine Learning > Learning Graphical Models > Directed Networks > Bayesian Learning (0.66)

Abstract Geological properties are likely to be shared for different prospects in a common basin. If, for example, source rock is found in one location, this increases the chance of finding it in other locations. Other geological characteristics will have similar correlation structures. When developing a drilling plan, these correlations should be considered because they can affect the optimal drilling sequence. Modeling and accounting for these dependencies pose a significant challenge. The strength and structure of the dependencies must be assessed and specified through the use of conditional probabilities or correlations. Even for a relatively simple exploration plan, the resulting decision tree may have hundreds or even thousands of endpoints. The sequential exploration problem with geological dependence has been addressed by several authors in the past few years. In this paper we extend this work by (1) simplifying the structure and model of the problem through the use of Influence Diagrams (Bayesian Decision Networks) and (2) using stochastic reservoir models to assess the dependence structure and strength. We also illustrate how to use the Maximum Entropy principle to construct joint probabilities when we have incomplete information about the correlation structure. Using Bayesian Decision Networks informed by stochastic reservoir models simplifies the modeling of exploration plans with a realistic number of wells. Applying well known geostatistical techniques to aid in the assessment of the geological dependence structure and strength makes the approach more accessible to geoscientists wanting to account for these dependencies in their plans. Similarly, using Bayesian Decision Networks with their relatively simple and transparent structure will better facilitate the understanding and communication of the impacts of dependencies on the exploration plan. Given the typical exploration well cost, an optimal exploration program will in most cases generate significant savings.

Artificial Intelligence, bayesian decision network, Bayesian Inference, Bayesian network, correlation coefficient, decision tree, dependence, dependency, evaluation, geologic modeling, geological dependency evaluation, geological modeling, influence diagram, information, joint probability, machine learning, pairwise probability, probability, prospect, Reservoir Characterization, risk management, spe 147062, three-well example, Upstream Oil & Gas

Oilfield Places:

- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6507/8 > Heidrun Field > Åre Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6507/8 > Heidrun Field > Tilje Formation (0.99)
- Europe > Norway > Norwegian Sea > Halten Terrace > Block 6507/8 > Heidrun Field > Ile Formation (0.99)
- (5 more...)

SPE Disciplines:

Abstract Prospects in a common basin are likely to share geologic features. For example, if hydrocarbons are found at one location, they may be more likely to be found at other nearby locations. When making drilling decisions, we should be able to exploit this dependence and use drilling results from one location to make more informed decisions about other nearby prospects. Moreover, we should consider these informational synergies when evaluating multi-prospect exploration opportunities. In this paper, we present a practical approach for modeling the dependence among prospects and determining an optimal drilling strategy that takes this dependence into account. We demonstrate this approach using an example involving five prospects. This example illustrates the value of modeling dependence among prospects and the value of learning about individual geologic risk factors when choosing a drilling strategy. Introduction When engineers and geoscientists consider a new prospect, it is important to consider its probability of success. In practice, this assessment is often decomposed into success probabilities for a number of underlying geologic factors. For example, one might consider the probabilities that the hydrocarbons were generated, whether the reservoir rocks have the appropriate porosity and permeability, and whether the identified structural trap has an appropriate seal (see, e.g., Magoon and Dow). The overall probability of success is the product of these individual probabilities. For any single prospect, though these assessments may be difficult, the process is straightforward. However, when considering multiple prospects in a common basin or multiple targets in a single formation, in addition to considering the success of each prospect in isolation, we need to consider the dependence among prospects. For example, if hydrocarbons are found at one location, they may be much more likely to be found at another nearby location. Conversely, if hydrocarbons are not found at the first location, they may be much less likely to be found at the other. If we are making sequential drilling decisions, we should be able to exploit this dependence and use the early well results to help us make more informed decisions at other locations. When evaluating multi-prospect exploration plays, we should recognize these informational synergies, design drilling strategies that exploit these synergies and value the opportunities appropriately. Unfortunately, this can be quite challenging to do in practice. To fully model dependence among prospects, we must specify the joint probability distribution over all possible combinations of outcomes for the individual prospects. For example, if we consider five prospects where each well may be either a success (productive) or a failure (unproductive), there are 2 = 32 possible outcomes whose probabilities must be specified. Many of these probabilities will be very difficult to assess. For example, what is the chance that a well at location 5 would be productive given that 1 and 4 failed and 2 and 3 succeeded? If we have decomposed the individual risk assessments into underlying geologic factors, the task becomes even more complicated: we must assign a joint probability distribution for all of the possible combinations of outcomes of all of these factors, at each location. With three factors each of which may succeed or fail and 5 wells, there are (2´2´2)» 33,000 different possible outcomes. Then, if we did somehow manage to specify a joint probability distribution over all possible well outcomes, it is difficult to determine an optimal drilling strategy that takes advantage of the informational synergies provided by this dependence. A traditional decision tree model would begin by considering which well to drill first, if any. We would then learn the results for that well and decide which well (if any) to drill next and so on, for all possible well outcomes and possible sequences of wells. Though conceptually straightforward, these decision trees can be quite large, even with a modest number of wells. For example with 5 wells, if we only learn whether a well failed or succeeded, the straightforward tree would include 9,496 scenarios. If we consider success or failure of three underlying geologic factors, then the decision tree would contain approximately 5 million scenarios.

Artificial Intelligence, correlation coefficient, decision tree learning, dependence, drilling operation, Drilling Strategy, equation, geologic factor, information, joint distribution, joint probability distribution, Lagrange multiplier, machine learning, Petroleum Engineer, probability, prospect, risk and uncertainty assessment, risk management, Scenario, smith 14, Upstream Oil & Gas, well 2, well 3

SPE Disciplines:

Technology:

Abstract The paper presents a new approach for modeling important geological elements, such as reservoir, trap and source, in a unified statistical model. This joint modeling of these geological variables is useful for reliable prospect evaluation, and provides a framework for consistent decision making under uncertainty. A Bayesian Network, involving different kinds of dependency structures, is used to model the correlation within the various geological elements, and to couple the elements. Based on the constructed network, an optimal sequential exploration strategy is established via dynamic programming. This strategy is useful for selecting the first prospect to explore, and which decisions to make next, depending on the outcome of the first well. A risk neutral decision maker will continue exploring new wells as long as the expected profit is positive. The model and choice of exploration strategy is tailored to a case study represented by five prospects in a salt basin, but it will also be useful for other contexts. For the particular case study we show how the strategy clearly depends on the exploration and development cost, and the expected volumes and recovery factors. The most lucrative prospect tends to be selected first, but the sequential decisions depend on the outcome of the exploration well in this first prospect.

Artificial Intelligence, basin, Bayesian Inference, Bayesian network, case study, discovery, evaluation, exploration campaign, exploration strategy, geological element, information, machine learning, marginal probability, mechanism, node, node 2, node 3, parent node, probability, prospect, Reservoir Characterization, Revenue, risk management, Scenario, strategic planning and management, structural geology, Upstream Oil & Gas

SPE Disciplines:

- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Management > Strategic Planning and Management > Exploration and appraisal strategies (1.00)
- Management > Risk Management and Decision-Making (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (1.00)

Summary Prospects in a common basin are likely to share geologic features. For example, if hydrocarbons are found at one location, they may be more likely to be found at other nearby locations. When making drilling decisions, we should be able to exploit this dependence and use drilling results from one location to make more informed decisions about other nearby prospects. Moreover, we should consider these informational synergies when evaluating multiprospect exploration opportunities. In this paper, we describe an approach for modeling the dependence among prospects and determining an optimal drilling strategy that takes this information into account. We demonstrate this approach using an example involving five prospects. This example demonstrates the value of modeling dependence and the value of learning about individual geologic risk factors (e.g., from doing a postmortem at a failed well) when choosing a drilling strategy. Introduction When considering a new prospect, it is important to consider its probability of success. In practice, this assessment is often decomposed into success probabilities for a number of underlying geologic factors. For example, one might consider the probabilities that the hydrocarbons were generated, whether the reservoir rocks have the appropriate porosity and permeability, and whether the identified structural trap has an appropriate seal [see, e.g., Magoon and Dow (1994)]. The overall probability of success is the product of these individual probabilities. Although these assessments may be difficult, for a single prospect, this risk analysis process is straightforward. When considering multiple prospects in a common basin or multiple target zones in a single well, in addition to considering the probability of success for each prospect, we need to consider the dependence among prospects. For example, if hydrocarbons are found at one location, they may be much more likely to be found at another nearby location. Conversely, if hydrocarbons are not found at the first location, they may be less likely to be found at the other. When evaluating opportunities with multiple prospects, we should consider decision processes and workflows that exploit this dependence and use results from early wells to make more informed decisions about other locations. For example, if a postmortem analysis of core samples from a failed well reveals that there were no hydrocarbons present, then we may not want to continue drilling at nearby sites. On the other hand, if the postmortem analysis reveals that hydrocarbons were present, but the reservoir lacked a seal, then we may want to continue to explore other nearby sites. In this paper, we describe an approach for modeling dependence among prospects and developing a drilling strategy that exploits the information provided by early drilling results.

Artificial Intelligence, assessment, asset and portfolio management, conditional probability, dependence, drilling operation, Drilling Strategy, geologic factor, information, joint distribution, joint probability distribution, Lagrange multiplier, machine learning, marginal probability, optimal strategy, probability, prospect, Reservoir Characterization, risk management, Scenario, SPE Reservoir Evaluation, Upstream Oil & Gas, well 1, well 2, well 4

SPE Disciplines:

Technology:

Thank you!