Abstract

Cities Service International and the Cities Service Southern Region Exploration and Production Department are using decision tree analyses as a tool for strategic planning. The analyses are tailored to the particular evaluation requirements, but all have a common goal: to reduce a complex pattern of varying expenditures to an easily understandable calendar of drilling activities, which highlight future expected manpower requirements, expenditures, and reserves discoveries. The decision tree models are also used to evaluate the impact of changes in parameters such as manpower and budget.

A significant number of real world situations operate as networks or can be conceptualized as networks for purposes of evaluation.

Network models have two fundamental components, activities and relationships. Activities are the discrete network segments that represent action or performance within the network. Relationships are the performance within the network. Relationships are the connectors which define the ordering of activities within the network identifying the precedences and consequences of each activity.

The IPEX Model, Cities Service International, consists of 56 activities connected by 38 nodes. The variety of path alternatives corresponds to the variety of actual work progressions that evaluation of a project may take. This network is the basis from which project may take. This network is the basis from which evaluation of the exploration program begins. This evaluation is performed using Monte Carlo simulation. The IPEX Model simulates the exploration process 500 times in order to draw a representative sample of the population of all possible exploration program population of all possible exploration program defined by the parameters of the model. From this sample, statistical inferences can be made regarding properties of the exploration program being evaluated. properties of the exploration program being evaluated. The IPEX Model has the capability of evaluating multiple years of program data. The multiple year simulation run provides a more realistic evaluation environment; the cumulative statistics generated represent a mix of expenditures on projects initiated in different years, similar to the actual exploration program pursued by International Exploration and program pursued by International Exploration and Production. Production

Introduction

Complex exploration programs consisting of numerous projects in various stages of evaluation present management with a major planning problem. Not only is it difficult to determine exactly what level of manpower and monetary support will be required to continue project evaluations begun in previous years, it is equally difficult to predict the future staffing and monetary requirements of an expanding exploration program, particularly if the project "mix" is changing program, particularly if the project "mix" is changing from the established historical pattern. A preliminary investigation of the problem suggested the use of preliminary investigation of the problem suggested the use of a multi-period linear optimization model.

Such a model could have maximized projects evaluated subject to resource constraints or minimized resources utilized. This approach was regarded as untenable, since it failed to adequately address the probabilistic nature of project evaluation with probabilistic nature of project evaluation with respect to resources utilized and timing of evaluation activities.

Further evaluation led to the adoption of decision tree analysis. Decision tree analysis was found to be particulary applicable, since the exploration program could be easily represented by a network structure.

Monte Carlo simulation of the exploration program was chosen over a mathematical solution for program was chosen over a mathematical solution for expected value, since a simulation could determine not only the mean or expected values, but the minimum and maximum values as well. This range information was of particular value to management in determining which particular value to management in determining which variables were subject to wide variation. The simulation approach also provided the data necessary for constructing probability distributions of reserves discovered.

The International Petroleum Exploration (IPEX) Model is intended for use as a management tool to aid in the evaluation of exploration programs.

This content is only available via PDF.
You can access this article if you purchase or spend a download.