Representative Projects

Los Alamos National Laboratory. Senior consulting risk assessor providing risk assessment and statistical support for human health and ecological assessments for the North Ancho Canyon and S-Site Aggregate Area Investigation Reports and Potrillo/Fence Canyon Aggregate Area Supplemental Investigation Reports.

US Department of Energy. Senior consulting risk assessor for human health and ecological assessments for chemicals and radionuclides in soil and rock via multiple scenarios, including vapor intrusion modeling at the Miamisburg, Ohio closure project.

Rocky Mountain Oilfield Testing Center (RMOTC). Provided data validation support for landfill groundwater sampling program.

Gary/Chicago International Airport. Developed a sampling plan to assess pollution on airport property and form plans for cleanup activities in support of the upcoming airport expansion project.

NuWest North Maybe Mines. Calculated bioaccumulation factors for plant tissue in support of ecological risk assessments at East Mill Dump and North and South Open pit mines. Estimated Tier1 and Tier2 risk values. Developed a sampling plan to define the distribution of concentrations representative of background for the area and specified methods for development of site-specific background threshold values and comparison of background concentrations to site concentrations.

Magnablend, Inc., Mills, WY site. Related to the presence of arsenic and selenium in soil and selenium in groundwater in excess of default criteria, developed a background sampling design that utilizes use of background soil data collected previously as part of a remedial investigation process at the nearby BP/Amoco Former Refinery North Properties. BP North Properties soil boring logs were reviewed to identify locations with comparable lithology Magnablend site. The resulting data set consisted of selenium and arsenic soil samples considered for developing background threshold values, which were compared to site data using several non-parametric two-sample hypotheses testing approaches.

Lower Passaic River Study Area. Estimated the distribution exposure durations for residents in the LPRSA, based on the probability of a resident moving out of the study area during a 5-year age group and he probability that a person will die during that period using U.S Census Bureau data and N.J. Department of Health (NJDH), Center for Health Statistics Bergen, Essex, Hudson and Passaic county mortality data for 2000-09.

NAVFAC Bangor.  Performed a cluster analysis in support of the sampling plan for the Nearshore Functional Assessment Tool calibration area in Puget Sound. A total of 4770 potential sampling locations were selected based on 14 Biology, 14 engineering and 29 hydrological clustering variables. A Gaussian finite mixture EEE (ellipsoidal, equal volume, shape and orientation) model fitted by the Expectation-Maximization algorithm resulted in the identification of 22 clusters.

Arrow Energy.  Developed a sampling plan to divide the priority matrix sources into categories including facilities, equipment and components identified by equipment type, mode of operation and type of emission and incorporating measurements at valves, flanges, connectors, compressor seals, pressure relief valves, compressor open-ended line and at flares. Initially measured data was subjected to a multivariate statistical analysis to determine precision, bias, accuracy and uncertainty. Comparison of total measured emissions extrapolated across Arrows gas field/ processing operations against current calculated loss determinations was performed.

FE Warren AFB Landfill-C501/Landfill 8 long-term monitoring. The goals of the statistical evaluation were to determine the presence of a statistically significant concentration increase (S51) for any of the laboratory-analyzed constituents. The statistical evaluations included both inter-well and intra-well analyses; the inter-well analyses compare the groundwater concentrations between background well and downgradient wells and the intra-well screening comparing median concentrations to the UPL. To further evaluate the potential for an SSI, a Non-Parametric Prediction Limit Statistical Test for Future Median was performed at a 97.5% level of confidence.