Site-specific water quality standards (WQS) are developed in order to reach regulatory criteria appropriate for individual bodies of water. These WQS are required as the nationally suggested water quality criteria (WQC) for seawater was derived with clean coastal seawater that does not include the natural ingredients that buffer the toxic effects of contaminants. As such, federal WQC could be overprotective, enforcing effluent characteristics that are very difficult and expensive to attain. The regulatory community overcame this problem with the development of total maximum daily loads (TMDL) and water effects ratios (WER), approaches that require long-term demanding and expensive studies. In an effort to speed up the development of WQS for copper (Cu), in 2007 the U.S. Environmental Protection Agency (USEPA) incorporated the Biotic Ligand Model (BLM) into a freshwater WQC. This model takes into account the natural characteristics of each body of water to derive a site-specific WQS. In a similar effort, the regulatory community is supporting the development of a seawater BLM for application in marine waters.


The objective of this project was to demonstrate an integrated modeling system, developed under SERDP project ER-1156, that provides an improved methodology for achieving compliance for Cu in Department of Defense (DoD) harbors in a manner consistent with the current regulatory framework for freshwater systems. This system also provides a management tool for optimizing source control efforts, as it is robust enough to forecast their effects on ambient Cu concentrations and the potential for toxicity in DoD harbors. The integrated model consists of the Curvilinear Hydrodynamics in Three Dimensions (CH3D) and a seawater Cu toxicity model (seawater-BLM), for simultaneous evaluation of fate and transport (F&T) and potential effects of Cu on a harbor-wide scale.

Demonstration Results

Demonstration of the integrated model in San Diego Bay and Pearl Harbor fulfilled most of the performance objectives. These objectives included a reliability parameter of explaining ≥60% of the variability in the field data for the prediction of total Cu concentration (Cutot) and dissolved Cu concentration (Cudiss). Predicted Cutot explains 61 to 94% of the variability of the measured values in both DoD harbors. In the case of Cudiss, the predictive capability of the integrated model was affected by a required minimal gradient in concentration (ΔC). In the cases where there was a gradient in concentration of 0.22 μg/L-1 or greater, the predicted values explain 68 to 92% of the variability. In contrast, in those cases where the range in Cudiss was minimal (ΔC 0.009 μg/L-1), making Cudiss essentially a constant value, a case where linear regression is not applicable, the objective was not fulfilled in spite of the great similarity between the values. The performance objectives for the prediction of free Cu ion (Cu2+) were adjusted for the lack of gradient. In San Diego Bay, field-measured Cu2+ was extremely stable and constant (i.e., small ΔC) and neglected the use of linear regression. The objective was therefore modified to prediction of values within an order of magnitude of measured values. This objective was fulfilled for most of San Diego Bay, excluding the area by the mouth of the Bay. There are no measurements of Cu2+ in Pearl Harbor; therefore, there was no procedure available to evaluate the predictive capability of the integrated model there.

Regulatory use of the integrated model will be mainly on the prediction of toxicity and WQS for the entire bay. Toxicity predictions are within the expected performance criteria in both harbors, as 87% of the values predicted for both calibration and validation are within a factor of two of the measured values. Two advantages of applying the integrated model over the current approach of developing toxicity and WER studies are the spatial resolution of the predicted values and the extreme reduction in effort. The integrated model provided high-resolution (≈100 m) spatial distributions of toxicity and WER, which can only be developed by the inclusion of a significant number of samples when following the recommended WER approach.

Application of the integrated model for the development of WQS results in significant relief, while maintaining the intended level of environmental health. WER predicted by the integrated model for San Diego Bay and Pearl Harbor are comparable to those previously measured, as 80% and 98% of the cases for both calibration and validation are within a factor of two of the corresponding measured values, respectively. A geometric mean WER of 1.48 and 1.17 were predicted for San Diego Bay and Pearl Harbor, respectively, which are within the range previously reported. Application of a mean WER for each area in San Diego Bay results in significant relief, with an average WQS of 5.0 μg/L-1 for the entire bay.

Implementation Issues

Implementation of the integrated model in a new harbor will result in lower costs than those required for existing processes. The costs for this demonstration of the integrated model were compared to the costs expected from the individual implementation of a WER and an F&T model in a harbor of similar dimensions and characteristics as San Diego Bay. While this comparison is justified by the fact that both processes are required to provide information similar to that generated by the integrated model, the costs predicted for implementation of these processes was simplified to some degree. Moreover, a significant increase in effort should be expected in order for these processes to provide the same quality of spatial information and capability for forecasting effects. The cost of the demonstration in San Diego Bay was $580,000, which is $250,894 more than the costs estimated for implementation of a WER and a CH3D ($329,106). However, implementation of the integrated model in a new harbor is estimated at $189,368. This will provide better temporal and spatial resolution and forecasting capability of source controls.

This demonstration contributes to the transition of this technology to the user community by providing a clear example of implementation at real-world DoD sites. Critical aspects included development and refinement of the BLM for sensitive saltwater toxicity endpoints and implementation of USEPA guidance for TMDL and site-specific WQS within a rigorous numerical modeling framework for Cu and eventually other metals.