Open AccessA simulation and decision analysis approach to locating DNAPL in subsurface sediments
Author(s) -
Brian Borchers,
Stephen H. Conrad,
Erik K. Webb
Publication year - 1997
Language(s) - English
Resource type - Reports
DOI - 10.2172/532658
Subject(s) - residual , sampling (signal processing) , probabilistic logic , percolation (cognitive psychology) , process (computing) , computer science , environmental science , petroleum engineering , operations research , data mining , geology , engineering , algorithm , artificial intelligence , filter (signal processing) , neuroscience , computer vision , biology , operating system
This report presents a strategy for delineating the location of residual dense non-aqueous phase liquids (DNAPL) that combines probabilistic simulations of DNAPL spill location and volume, geologic texture constraining migration pathways, migration physics through percolation modeling, and a decision analysis model to pick optimal locations for sampling wells. The authors` strategy is an iterative process of simulating the residual DNAPL location, selecting new locations for data collection, gathering data, and then using the data to condition further simulations of DNAPL migration. As they iterate through this process, data worth analysis is used to determine an appropriate stopping point. The authors present the results from a preliminary version of their model, showing how the process was used to delineate hypothetical DNAPL spills