Model Grid of the PVSC CSO Long-Term Control Planning | Integrating Models Into Water Quality Solutions
Article

Integrating Models Into Water Quality Solutions

Water Quality Modeling 

What is water quality modeling? It can be defined as numerical modeling related to assessing and characterizing the biological, chemical, and physical properties of natural water systems for the purpose of providing decision makers with water quality-based information for assessing environmental regulations aimed at maintaining and protecting water quality, human health, and aquatic resources. A numerical model provides a quantitative tool to assess management decisions and their implications on improving water quality. That said, it is still only one tool in the toolbox, and environmental solutions require a broad-based view of all the factors affecting water quality restoration (environmental, economic, social).

An early definition of water quality modeling was the word potomology, as the science of rivers in Stream Sanitation (Phelps, 1944). An extended definition was later developed, as presented below, which uniquely describes the process of water quality modeling.

“Potomology – the science of rivers. Potomology applies knowledge from many areas of the physical and biological sciences and mathematics. The potomologist need not qualify as a specialist in each of these areas, but he must be able to integrate knowledge from all of them as he pursues his own specialty. The definition is extended to also include knowledge in areas of the social sciences and engineering, essential elements in effective pollution control and wise use of water resources. Furthermore, in application potomology becomes as much an art as a science, tempered by experience and professional judgment.” (Velz, 1984) 

These sentiments have been echoed by other noted engineers and scientists involved in the field of water quality modeling:

“In today’s high-tech, computer-oriented, hardware/software-focused world, the need to understand the basic concepts of mathematical models of water quality is particularly acute.” Principles of Surface Water Quality Modeling and Control (Thomann and Mueller, 1987) 

“On the positive side these trends could result in models contributing more effectively to improved water-quality management. On the down side widespread and easy use of models could lead to their being applied without insights as ‘black boxes’.” Surface Water Quality Modeling (Chapra, 1997)

Protecting and managing the quality of the world’s natural resources in a cost-effective manner presents many technical challenges. As one of the stakeholders who takes responsibility for maintaining and improving the water quality of our rivers, lakes and coastal systems, we at HDR believe it’s important to understand the unique variables and interrelationships of the elements impacting water quality, as well as the available alternatives for improved management.

We provide integrated water quality solutions for managing water resources for both human and ecological needs. Our mathematical models are instrumental in understanding the fate and transport of contaminants in the natural environment. The development and application of these models enables us to evaluate the effectiveness of management alternatives to protect receiving water quality and water supply for our clients. Today, we work with industry, utilities and government agencies to develop innovative programs to help identify cost-effective, sustainable solutions to water quality problems. Our modeling tools address the full spectrum of water quality management for regulatory compliance: watershed models for determining rainfall-driven runoff (quantity) and nonpoint source loads (quality); hydrodynamic or circulation models for determining the movement of water in fresh and marine environments (transport); and water quality models for determining water quality impacts due to point and nonpoint sources (fate). These water and sediment studies include assessment of conventional pollutants 
(nutrients, BOD, ammonia, dissolved oxygen, bacteria), toxic pollutants (PCB, PAH, dioxin, mercury, metals), and thermal discharges and bioavailability (metals, organics) for water quality management decision support (permitting, TMDLs, remediation).

Our water quality modeling experts routinely integrate multiple modeling types, as discussed in the prior sections, to address important water quality issues as highlighted in a few of our projects below.

New York City  Integrated Modeling

The Long-Term Control Plan for combined sewer overflows in New York City uses an integrated modeling approach to address pathogen control options. Some waters in and around the New York City area, particularly confined tributaries and embayments, do not currently meet the primary contact recreation water quality standards for fecal coliform. In an effort to meet these standards, the New York City Department of Environmental Protection is evaluating options to reduce impacts from CSOs. In order to help optimize the CSO controls, several numerical models are being used as part of the CSO control evaluations. The integrated modelingframeworks have been developed for 12 areas that include over 15 specific water bodies around New York City, ranging from highly urbanized tributaries to Jamaica Bay, to the open waters of New York Harbor and western Long Island Sound.

Fresh Creek water body and New York City Department of Environmental Protection Long-Term Control Plan Areas | Integrating Models  Into Water Quality Solutions
Figures 1 and 2: Fresh Creek water body and New York City Department of Environmental Protection Long-Term Control Plan Areas

The first model in the integrated modeling approach is the InfoWorks hydraulic model. InfoWorks is a hydraulic sewer system model that calculates the time-variable runoff from precipitation events, which can also estimate the fraction of sanitary and stormwater flow from CSOs. The New York City models are set up to provide time-variable flow from wastewater treatment plants, CSOs, storm sewers and direct drainage runoff. The flow information from InfoWorks is fed into the Estuarine and Coastal Ocean Model, which is a three-dimensional, time-variable, hydrodynamic model that calculates the tidal circulation of water through advection and dispersion along with water temperature and salinity. The output from InfoWorks and ECOM is fed into a third model, the Row-Column AESOP (RCA), which is a water quality model that can be used to calculate pathogen concentrations in the receiving water. RCA is an advanced water quality model developed at HDR (formerly HydroQual) from the original WASP water quality model. The original model was also developed by our staff, and includes advanced eutrophication kinetics for completing nutrient-algal-dissolved oxygen studies. Pathogen loads from CSOs, storm sewers, direct drainage and WWTPs are assigned to RCA, which in turn calculates receiving water pathogen concentrations (fecal coliform and enterococcus).

Based on data collected during the LTCP program, each model is calibrated so that it reproduces the measured conditions within the sewer system and receiving water bodies. Once the models are calibrated, they serve as useful tools to assess which controls can be applied to CSOs to meet the water quality standards in a cost-effective manner. The InfoWorks model is used to assess how to cost-effectively reduce CSO volume. Then RCA is used to assess how these CSO loading reductions affect water quality and attainment of the applicable water quality standards.

Typically, InfoWorks is run to simulate CSO reductions of 25, 50, 75 and 100 percent. The water quality model is then used to assess the changes to pathogen concentrations. Design engineers then look at alternatives that either reduce CSO volume, such as tanks, tunnels, WWTP optimization and green infrastructure, or other means of reducing CSO loading, such as disinfection. Based on the information from the models and design analysis, cost-benefit curves are developed in an effort to optimize the CSO controls. Additional model runs are then conducted to address more specific CSO control options.

New York City currently has more than 400 CSOs and is considering spending approximately $3 billion to reduce their impact. Even this amount of funding is not enough to address all of the CSOs. By using models in an integrated fashion, NYCDEP is optimizing its available funding by using the best tools available so that the waters surrounding New York City meet their designated uses and protect public health.

Lower Passaic River/Newark Bay Contaminated Sediment Modeling

The Diamond Alkali Superfund Site is located in and near Newark, NJ, with four operable units or remediation areas identified to date. Those OUs consist of: OU1 – the facility where Agent Orange and pesticides were produced near river mile 3; OU2 – the lower 8.3 miles of the Lower Passaic River; OU4 — the sediments of the upper 9 miles of the Lower Passaic River and the water column throughout the 17 miles of the Lower Passaic River; and OU3 — Newark Bay, downstream of the Passaic River. During its operation, the Diamond Alkali facility produced DDT and Agent Orange. A by-product of Agent Orange production was 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD), one of the most toxic chemicals ever made by man. The sediments of the Passaic River and Newark Bay have elevated concentrations of 2,3,7,8-TCDD as well as a number of other chemicals contributing to unacceptable levels of risk to both human health and ecology.

We are performing modeling on three projects covering OUs 2, 3 and 4 of the Diamond Alkali Site. On OU2, we performed modeling to support EPA’s Remedial Investigation/Focused Feasibility Study, Proposed Plan and Record of Decision and continue to provide modeling support for remedial design oversight. On OUs 3 and 4, we are providing oversight of potentially responsible party-led RI/FS studies. All three projects use variations of the same suite of models drawing from a wide range of expertise. The suite of models includes hydrodynamic; sediment transport; organic carbon production, fate and transport; and contaminant fate and transport models. Additionally, these models utilize inputs from CSO and stormwater modeling of the watersheds and sewersheds surrounding the study area. The project team’s intimate knowledge of the models allows us to adapt them to meet clients’ needs, including the representation of bathymetry changes; sediment composition; and the releases of solids, organic carbon, and contaminants associated with the simulation of potential remediation options.

Our modeling expertise across a broad range of disciplines provided a key component to EPA’s assessment of remedial options and final decision for OU2. In addition, our expertise has allowed EPA to provide insights to the PRP groups and make progress on both OU3 and OU4 RI/FS studies, while avoiding pitfalls of overly simple or complex revisions to the models being used for those projects.

Long Island Nutrient Modeling

We recently completed nutrient modeling as part of a dissolved oxygen TMDL in the Forge River and started hydrodynamic modeling of numerous water bodies in Suffolk County, Long Island, NY, for nitrogen management planning as part of the Long Island Nitrogen Action Plan and Suffolk County’s Reclaim Our Water Initiative. In these studies, hydrodynamic models (and a water quality (eutrophication) model for the Forge River) have been integrated with a groundwater model to provide freshwater flows and nitrogen loads to the receiving water models. Much of Suffolk County on Long Island is still served by private septic systems that contribute to widespread groundwater nitrogen contamination that impacts surface water bodies. These nitrogen impacts are manifested in the surface waters through harmful algal blooms (e.g., brown tide), low dissolved oxygen levels, fish kills, and suspended phytoplankton and macro-algal blooms (e.g., Ulva or sea lettuce).

Forge River Model | Integrating Models  Into Water Quality Solutions
Figures 3: Forge River model bathymetry area

The hydrodynamic modeling for Suffolk County’s Subwatersheds Wastewater Plan is providing flushing time calculations for over 150 tidal water bodies. These flushing time results will be used as part of a ranking matrix that includes baseline water quality and nitrogen loading from surface runoff and groundwater to rank water bodies and focus nitrogen management efforts in the future. Due to significant groundwater contribution to freshwater flow and nitrogen loads in Suffolk County, groundwater model output was coupled with the hydrodynamic model to properly represent freshwater flows in the tidal water bodies. In the Forge River TMDL eutrophication modeling, inclusion of the groundwater model nitrogen loads was just as important as the coupling of groundwater flows so that all nitrogen loads are included and proposed nitrogen reductions can be correlated to private septic systems control or new advanced WWTP sources that discharge to groundwater.

Long Island Model | Integrating Models Into Water Quality Solutions
Figure 4: Long Island hydrodynamic modeling area

Although the two modeling projects are still underway, the coupling of the groundwater model to the surface water models (hydrodynamic and water quality) represents a unique effort to treat freshwater inflows and nitrogen loads on a holistic basis — in a way that is not typically considered in watershed or TMDL studies used for investigating management alternatives.