Multi-Criteria Decision Making in Strategic Reservoir Planning Using the Analytic Hierarchy Process

Abstract

The paper describes the use of the Analytic Hierarchy Process in the strategic planning for the development of a series of reservoirs in the same field. A department of the Kuwait Oil Company is responsible for the development of these reservoirs and needed to determine the best order in which to develop these assets. In order to do this it was necessary to make judgements regarding how well any development option met a set of strategic criteria. At the individual reservoir level, it was also necessary to judge which development option best met a set of criteria for the development of specific reservoirs. A decision hierarchy was developed that reflected the ordering of the criteria and the specific alternatives. Pair-wise comparisons of the various criteria were made in order to determine how well they satisfied the overall objective. These judgments were evaluated using a judgment matrix to determine weights for the criteria. These criteria were applied to decisions to be made at the next lower level in the hierarchy. In this way, the decisions, which were judged to best meet the overall objective, were given the highest priority. This influence cascaded down through the hierarchy to rank the development options. With this structure it was simple to substitute one overall objective for another to determine whether this would cause a change in decision policy. The sensitivity of the decision outcome was also evaluated for its sensitivity to the range of values of the input factors. Since there were thirty-five factors to evaluate, a statistical design approach was used to evaluate factor effects. It is shown that the planning factors, planning horizon and the demand scenario used for planning purposes were significant in all cases evaluated, and highlighted the importance of testing the assumptions, criteria and objectives in a strategic planning exercise. Introduction Uncertainties in Strategic Planning Strategic planning is necessary for the proper development and management of any oil producing asset. However, there are many sources of uncertainty in any planning exercise. Uncertainties in input parameters, such as economic factors and production potentials, lead to uncertainties in predicted results. There are well documented means of addressing these uncertainties. We do not often give cognizance to the uncertainties in our planning parameters themselves, however, or test our strategic plans given different strategic objectives. It is not always clear what is the appropriate objective for our strategic planning. The executives of publicly traded companies will say it is to maximize NPV (net present value), or a similar metric. However, even publicly traded companies are subject to demands that may not be congruent to maximizing NPV, health, safety and the environment objectives, community improvement, security of management tenure, employee job security, investor interests, and other objectives, all have their constituencies. A system that provides a conceptually consistent means of weighting the set of appropriate objectives for any development would allow the explicit incorporation of the potentially disparate objectives into the ranking of projects. In addition, if we are uncertain as to what is the proper strategic objective, or if the objective changes due to changing circumstances, it would be beneficial to have a system that allows easy substitution of objectives for reevaluation. We could then determine how robust is our plan.

Application of fuzzy comprehensive evaluation to evaluate the effect of water flooding development

“…Based on the hierarchical structure model, we can obtain the judgment matrix. The maximum eigenvalue and the eigenvector of the judgment matrix X are calculated by using the power method (Wen et al 2017;Gerbacia and Hanaa 2001). Firstly, the factor set U = u 1 , u 2 , … , u n is determined, and then the judgment matrix P = P ij m×n is constructed by the following method, which P ij reflects the relationship between the two factors.…”