Improving Stakeholder Engagement and Regulatory Decision-Making Through Enhanced Visualization and Animation of Water Quality Model Output

The City of Richmond, Virginia has relied on an EFDC bacteria (E. coli) water quality model of the James River to support its combined sewer system planning efforts, with the model serving as a critical tool in evaluating proposed alternatives and reporting in-stream bacteria results to stakeholders and regulators. Like many bacteria water quality models, Richmond's model is technical and complex. It represents and captures complicated physical and biological properties like variable and multiple flow and bacteria sources, tidal influence, and bacterial decay. The resulting in-stream bacteria concentrations change over time and location and are highly dependent on weather and flow conditions. Because of this complexity, it can be challenging to effectively and clearly communicate model results to stakeholders and regulators. Model results are typically conveyed using static timeseries plots and tables that summarize total bacteria concentrations at specific locations and for specified time periods. Summarizing results in this fashion can be helpful to assess whether a river meets bacteria water quality standards at a defined location and period, but it does not show where the bacteria originate (i.e: from the CSO, WWTP, MS4/tributaries, or upstream sources) or how concentrations vary over time and space. To convey this information to stakeholders and regulators, LimnoTech used bacteria source components analyses to track and present individual bacteria sources independently, and created dynamic mapping and animations of model output that shows bacteria concentrations for each model grid cell. Richmond's development of its Final Combined Sewer System Plan will fulfill the requirements of a State Senate Bill to meet the special consent order issued by the State Water Control Board. Projects designed to reduce the City's bacteria load from CSOs are evaluated by the water quality model to determine relative impacts of combinations of projects to the in-stream bacteria concentrations and attainment of the bacteria water quality standards. Projects that were evaluated include high-cost, low-feasibility approaches such as full CSO capture and treatment, and lower-cost alternatives that include some combination of storage, high-rate-disinfection, and partial sewer separation. Static timeseries, box-and-whisker, and cumulative frequency distribution plots of model results remain necessary to document and communicate high-level model results and predictions of project impacts. These plots capture model output (and sampling data, when available) at a handful of key locations, in the form of E. coli concentrations. Plots can also be expressed in the language of applicable Virginia bacteria water quality standards for recreational waters, conveying bacteria concentrations as 90-day geometric means and 90-day statistical threshold value exceedance percentages, and comparing these against the E. coli water quality criteria. These types of plots do not convey the spatial aspect of changes to bacteria concentrations, nor do they inform stakeholders of the relative magnitudes of the sources (CSO, WWTP, MS4/tributaries, upstream) of the bacteria, or the movement of bacteria loads resulting from tidal forces. LimnoTech used the component-tracking module of the EFDC water quality model to track bacteria concentrations for each model grid cell by original source. At any grid cell, the total bacterial concentration can then be plotted alongside the source components; the portion of that bacteria load from upstream, CSO, MS4/tributaries, and WWTP sources can be determined. This enables stakeholders to see that the impact of CSO projects may be limited due to high bacteria loads originating from non-CSO sources. Similarly, calculations of model input loads by source — independent of model simulations and results — can help to inform stakeholders of the relative scale of CSO bacteria loads as a whole, compared with other persistent and significant sources of bacteria such as loads introduced upstream of the City. For a more holistic view of the dynamics of temporal and spatial bacteria load changes, animations of model results can provide stakeholders with a valuable tool for alternatives evaluation. This is especially true in an estuarine environment like the James River at Richmond, where tidal influences on flows are significant; this influence is not apparent in a static timeseries plot. LimnoTech used GIS-based water quality model results output, river flow timeseries, and collection system model output, and combined those to develop animations of the model grid that included 'picture-in-picture' insets featuring synchronized plots of river flows and CSO volumes. The animations showed stakeholders the baseline condition and the potential changes to bacteria concentrations from proposed CSS projects, with a focus on high-concentration wet weather events and periods of elevated river flow. LimnoTech used the source components analysis capabilities of EFDC and the dynamic model output animations to present a comprehensive story of the various bacteria sources in the river and to better convey the impact that future CSS projects will have on receiving water quality. This approach has helped stakeholders and regulators in understanding what contributes to bacteria impairments in the river and how the City's investments in CSS projects will reduce those impairments.