Oil and gas supply modeling

Strategic analysis for safeguards systems: a feasibility study. Volume 2. Appendix

“…(This reinforces our stress, in Section 3 above, on the possibilities for sensitivity and parametric analysis of game-theoretic models.) A third observation ( [154], pp. 3, [74][75] is that the severity of the "due process" criterion will naturally depend on the salience of model-use in arriving at the regulatory decision already substantially determined, to a maximum role in which model outputs become the determinate guidelines for decisions.…”

“…Another illustration is given by the large-scale modeling systemsmore or less descendants of the Project Independence Energy System (PIES)-developed and operated by the Department of Energy; cf. for example Gass [154] and Gass et al [155], Johnson [157] describes operational application in fisheries regulation. More generally, it seems clear that many quantitative issues involved in a regulatory judgment--the effect on competition of a proposed merger, the effect on regional pollution levels of a proposed change in processing fuels or technology, the adequacy of continuing current charges to provide a reasonable rate of return for a public utility-involve complexities and data-volumes that must be receiving formal mathematical treatment.…”

“…Chapters 6 through 8 of [154] deal specifically with model roles and presentation strategies in the context of "procedural due process." One piece of advice is to avoid the "oversell," of attempting to envelop the model in "an aura of verity."…”

“…Instead, model documentation and presentation should stress the rationality and reasonableness both of the entire approach and of the major choices made (necessarily on partly subjective "best judgment" grounds) at key points in the model-development process. A second implicit recommendation ( [154], pp. [87][88] is that the model should be capable of appropriate particularization to the specific cases at hand (e.g., by setting parameters to the relevant values) instead, for example, of being irrevocably wedded to aggregate average data that might be demonstrably significantly inapplicable to the particular instance.…”

“…Examples of an agricultural type (e.g., pest management and pesticide-use control) are given by Rovinsky and Shoemaker [161] . Energy-related models like those cited above [154,155] often are (or have major submodels) of the optimization type. Some members of the judicial branch may have acquired a sympathetic familiarity with optimization concepts from past expositions (e.g., Nagel with Neef [162], Nagel and Neef [163]) of those concepts in the context of applicability to the legal process, we note that some checks for "reasonableness" available for predictive models are no longer so clearly applicable to normative models, since observations at variance with model outputs may simply reflect non-optimal behavior, improvement on which may have been the very motive for developing the model.…”