Abstract
This case study investigates the feasibility of capturing urban runoff from a primarily residential drainage area and conveying it to the nearby E.A. Fairbairn Water Treatment Plant (WTP) in Sacramento, CA. The intent is to use the treatment plant as an alternative stormwater treatment strategy. Runoff would be conveyed, stored, and blended with the existing river water source at a ratio known as the stormwater fraction (SWF). A feasible SWF is one for which the stormwater contaminants can be treated to meet drinking water primary maximum contaminant levels (MCLs). Treatment costs were compared to those of retrofitting the urban catchment with Low Impact Development (LID) Best Management Practices (BMPs) to meet stormwater regulations and permits. Blending calculations using local stormwater and river water monitoring data were performed for thirty-one (31) contaminants with primary MCLs. Thirteen (13) contaminants including turbidity, Escherichia coli (E. coli), fecal coliform, copper, lead, and eight (8) polycyclic aromatic hydrocarbons (PAHs) did not meet applicable MCLs for one or more SWFs. After conducting treatment calculations utilizing operational data from the WTP and applying treatment efficiencies from literature sources, it was found that MCLs could not be met for only lead and four (4) PAHs at one or more SWFs. Lead was the limiting contaminant and the allowable SWF was forty percent (40%). The WTP staff was concerned, however, that any SWF over ten percent (10%) would exceed existing total organic carbon (TOC) concentrations and could result in excessive production of disinfection byproducts (DBPs). Consequently, three scenarios based on SWFs of ten (10%), forty (40%), and a hundred percent (100%) were selected for additional hydrologic calculations and cost analysis. Stormwater hydrology calculations were conducted to determine the urban runoff flow, storage volume, drainage area, and annual treatment volume for the selected SWF scenarios. The operating strategy was to capture and store water from a twenty-four (24) hour rainfall event and then process the water through the treatment plant during the twenty-four (24) hour period. The eighty-fifth (85th) percentile twenty-four (24) hour rainfall event was used to determine the required drainage area and annual runoff volume. The Basin Sizer software tool was also used to calculate the annual runoff volume based on continuous simulation. Present worth costs for the three scenarios were calculated from capital costs and the present value of the annual operating costs. No new treatment processes were added to the WTP so the capital costs included only a storage basin and a major conveyance pipe. Chemical requirements (alum, caustic soda, polymer, and chlorine) and sludge production for blended flows were estimated from correlation functions developed from historical plant data. The life-cycle cost for implementing LID in an existing neighborhood was adopted from an Urban Greening grant application in a residential neighborhood in Davis, CA. Each scenario was compared to the retrofit of a residential neighborhood in Davis, California with annual operating costs estimated at seven percent (7%) of the capital cost. With a wet weather capacity averaging thirty (30) million gallons per day (MGD), the drainage areas need to serve the WTP at three SWFs ranged from five hundred twenty-two (522) to five thousand two hundred twenty-one (5,221). The volume of stormwater treated ranged from sixty-seven (67) to six hundred sixty-nine (669) million gallons per year (MG/year). Present worth costs for the three scenarios ranged from $32,800 to $34,700 per acre of catchment area, compared to $245,000 per acre of catchment area for the LID retrofit. The intent of this case study is to be a starting point for further analysis and consideration of the proposed use of urban runoff. Capture and use of urban runoff in a drinking water treatment plant may be economically attractive depending on local conditions. Key areas of needed research are DBP formation from chlorinating urban runoff (or adopting an alternative disinfectant), chemical requirements for a blended influent, and removal efficiencies for metals and organics in conventional water treatment processes.