Floating spatial domain areal averaging: A multi-objective, multi-criteria decision support model for the development of cleanup goals at contaminated sites

Date of Completion

January 2000


Environmental Sciences|Engineering, Environmental




Cleanup plans for contaminated surface soils are frequently developed with the goal of ensuring that regulatory criteria for risk exposure are met after remediation. Currently, some of the plans for Superfund sites are based on the premise that a concentration equal to the average goal or based on a confidence level expression of the concentration measurements is not to be exceeded (NTE). However, in cases of wide-spread, low-level contamination, the NTE constraint may result in lowering risk well below the regulatory criteria and substantially increasing remediation costs. An alternative method, floating spatial domain averaging (FSDA), is developed and expanded. FSDA is a multi-criteria decision making tool in which concentrations or risk factors are averaged over domains which represent possible human exposure areas. The domains are variable in space. The averaging domain floats over the surface in increments representative of a field sample size (cell) or Geographical Information System (GIS) spatially interpolated raster unit. This allows for the incorporation of domain-wide topological information into the averaging operation around each cell and also the consideration of flexibility in the choices for the future development of the site. ^ The FSDA model is based upon linear averaging equations and can be analyzed both manually and also with the aid of linear programming optimization routines. The resultant solutions can be compared to the NTE approach. This modeling effort is further expanded to include multiple risk factors, maximum contaminant level (MCL) constraints, spatial domain or lot size variations and sampling limitations. ^ The multi-component FSDA model portraying a hypothetical site with low-level, wide-spread surface soil contamination and MCL constraints, resulted in the determination of several remediation plans. Many of these alternatives allow for a marked reduction in the extent of required remediation, while complying with the pre-determined cleanup goals and considering flexibility in future land use decisions. The reduction was particularly evident in cases where the number of cells within each area was limited by sampling constraints. ^