| Modeling: 3D Geoscience Modeling Cleans Up
A toxic waste dump in Maryland undergoes the first step in an environment clean-up project.
By Ed Rychkun Aberdeen Site History
The Aberdeen Proving Ground has operated since 1917 providing conventional and chemical agent weapons testing research and production services for the U.S. defense program. Two areas, Englewood and Aberdeen, cover 73,000 acres in Harford and Baltimore counties in Maryland where more than 300 disposal sites were used to bury munitions, unexploded ordnance, chemical and radioactive products, solvents, pesticides, and fuel materials. Aberdeen provided dual functions of testing/production for weapon systems as well as research of toxic warfare substances. An impressive list of toxic substances, including blister agents, tear gas, and nerve agents have apparently been hidden, and after several decades, many of the burial sites are releasing pollutants, and hazardous substances into the environment. As a result, Aberdeen was detected as a serious problem and placed on the NPL in October of 1989.
Further complicating the problem is the site proximity to upper Chesapeake Bay. Half is covered in water, split by Bush, Gunpowder and Chesapeake rivers. Add a complex network of several streams and the result is a very difficult groundwater system to understand.
Currently there is another unique problem. Previous burial practices are much of a mystery, creating an unusual and extreme hazard to site personnel. Investigators must perform geophysical sweeps to expose potential unexploded devices, or wear gas masks in the event of releasing chemical agents. It is suspected that groundwater and soils are contaminated but the degree is unknown. Risk of exposure at surface appears to be the greatest problem for site personnel. Despite the awkward conditions, soil sampling, borehole drilling, monitoring wells, surface water, sediment and soil gas sampling have been conducted, much of it done through expensive remote drilling and borehole magnetometry to reduce risk to human presence. Site Characteristics
The Aberdeen site is in flat terrain underlain by a complex stratigraphy of alluvium and variable plastic sediments. Major lithology groupings are silt/clay, sand and other plastic mixes, deposited in complex interbedded bands ranging from a few feet to 30 feet. The water table can be found approximately 20 to 50 feet below surface, its gradient undulating with no apparent correlation to surface. Site Complications
A serious problem exists with the quantity and quality of available information. The major obstacle results from the exceptionally high costs of sampling due to the local hazards. As a result, the extent of contamination is difficult to determine. Further, its distribution within the groundwater and the degree of inter-dependence with complex soil lithology adds to the uncertainty. These all create major complications that render interpretive methods typically used to extrapolate the spatial continuity of the contaminant as conflictive and of suspicious credibility. Application of Computerized 3D Geoscience Modeling
The Lynx modeling technology was the investigator's choice to best characterize subsurface. It is designed specifically to deal with difficult spatial problems. Lynx integrates spatial data management, geostatistical techniques, 3D modeling, volumetrics, engineering and 3D visualization in one facility. It becomes possible to use established analytical techniques to create detailed representations of highly complex subsurface problems, using methods to maximize data utility. In addition, it provides the ability to measure the degree of characterization and sampling uncertainty. Site Information and Basic Objectives
Information available for the study included a set of 68 boreholes, 30 with geological information. Two analytes were available, TRCLE and TOX reflecting the concentration of most serious contamination. Geology samples were available in the form of the USGS standard codes thus allowing codes to be grouped into the major lithological units of Silt/Clay, Sand, and Other. In addition, various DXF map information was available showing location of wells, study area, water table and various interpreted sections.
The main objective was to assemble the various sources of information into one spatial repository in order that a comprehensive preliminary characterization of subsurface be performed. This would lead to a better understanding of the contaminant volume, its location and provide a better visual appreciation of the nature of the problem. It would also provide a platform for a further analysis and investigation, guiding an optimized sampling program based on computed levels of uncertainty. Such a process could provide an objective, defensible model and set the initial guidelines for remediation planning. Geostatistical and Geological Analysis
Initial geostatistical analysis performed on the data showed some sample continuity up to 100 feet horizontally. Beyond this the uncertainty of estimations becomes very high. This provided the basis for developing a range of influence for initial estimation of contaminant distributions. In addition, lithology was modeled on a few sections, along with the top of the aquifer. By then using enhanced visualization, it was possible to cut planes through isosurfaces at various thresholds, overlay geology and water table models to show the plume inter-relationship to subsurface physics.
Finally, the development of a visual 3D model also provided an explicit visual picture of extent, degree and uncertainty, a vital tool that facilitated better analysis and understanding of the site - an expedient necessity where faster cleanup is the objective. Characterization Summary
Three dimensional geoscience modeling was used effectively to characterize and visualize a complex situation. Although not performed here, the next phase would be to display critical areas of uncertainty, and the effect on volumes. The objective would be to plan an optimized, cost-effective sampling program. A complete soil geology model and saturated/unsaturated zones would be used to define the local influences on contaminant migration. These would form critical input to the development of remediation volumes on the basis of pre-defined uncertainty criteria. About the Author:
Ed Rychkun has a degree in applied mathematics with special training in geology and operations research. He is president and CEO of Lynx Geosystems Inc. in Vancouver, B.C. Back |
