Decision Analysis in Contaminant Geohydrology

 

Dr. George F. Pinder

College of Engineering and Mathematical Sciences

University of Vermont

 

Date: Monday October 17, 2005

Time: 12:20 p.m. - 1:10 p.m.

Location: 367 Votey

 

 

Abstract

 

While significant effort has been and continues to be expended in the development of more physically representative mathematical models of the subsurface, comparatively little work has been done in developing the  computational tools needed to effectively use these models to facilitate and enhance groundwater remediation management decisions.

 

Since the greatest return on investment for groundwater modeling research occurs when limited resources targeted for remediation are used more cost-effectively through the use of models, current budgetary constraints and the concomitant demand for accountability in scientific research suggest that investment in the effective use of models may be now more critical than the further development of modeling capability in and of itself. 

 

Demonstrable effectiveness in cost reduction in the field has been achieved using models in combination with mathematical methods adopted and adapted from areas of science, mathematics and engineering heretofore virtually untapped by the groundwater community. To introduce these concepts we first consider the problem of NAPL source identification. Next, using field examples, we demonstrate 1) the least cost design of an inward-gradient constrained pump and treat system, 2) the least cost design of a risk (concentration) constrained pump and treat system, 3) the definition of a plume boundary using minimal water-quality sampling, and 4) long-term monitoring using the least possible number of water-quality samples.