Design Optimization for Subsurface Characterization and Management

D. E. Dougherty and D. M. Rizzo

Formal mathematical optimization methods can be applied with profound effect on the design of subsurface investigation, control, and monitoring projects. While investigating a site, for example, each new datum increases information and hence changes the spatial distribution of uncertainties about a site. This change, in turn, can be used to modify a field investigation so as to maximize the reduction of uncertainty across a site by moving the location of the next investigatory sample. Similarly, each additional piece of information can be used to improve estimates of subsurface parameters, such as material property values (e.g., hydraulic conductivity) and forcing functions (e.g., infiltration rate). The design of control schemes for water supply or subsurface remediation is markedly affected by these parameters. In addition, subsurface fluid control schemes are significantly affected by the statement of and assignment priorities to objectives, goals, and constraints, plus the treatment technology trains available for use at a given site. Observations during and after operation of the control system need to be located and intermittently sampled to assess performance, to indicate errors in modeling assumptions or interpretations, and to help guide modifications to planned strategies. We describe and present formal optimization problems and solution methods for these problems. Limitations of current techniques and current research areas are presented. Example applications of hypothetical and real sites indicate not only the potential of these methods, but their utility today.

AAPG Search and Discover Article #91019©1996 AAPG Convention and Exhibition 19-22 May 1996, San Diego, California