DEVELOPMENT OF AN URBAN INDICATOR BACTERIA TMDL: MODELING AND ALLOCATIONS

One of the most frequent problems plaguing our nation's waters is pathogen impairment (US EPA, 2000). In Houston area alone, there are 23 bayous, rivers and bays listed on the State of Texas 2000 303(d) List for pathogens. Two of these Houston bayous, Whiteoak and Buffalo Bayou, experience some of the most frequent water quality exceedances for contact recreation in Texas.The watersheds for Buffalo and Whiteoak Bayou are highly developed and urbanized. Highdensity commercial and residential land uses dominate the majority of the land area. Historical data show that fecal contamination of the bayous is extensive, with all water quality monitoring stations having geometric means significantly above the contact recreation criterion of 400 colony forming units (cfu)/100 mL during both wet and dry weather conditions. This urban bacterial contamination is unique due to the extreme nature of the urban contamination as well as the frequency of dry weather exceedances.Identified sources of the fecal contamination in the bayou include point sources (waste water effluent) and dry-weather storm water flows, in addition to sediment, wildlife and urban nonpoint sources and in Buffalo Bayou, upstream discharges from the Addicks and Barker Reservoirs. Sampling was conducted in 2001 to quantify source contributions from point sources, dry-weather storm water flows and bayou sediment.Monitoring of wastewater effluent from 76 plants in Buffalo and Whiteoak Bayous indicated that less than 10% of the facilities were exceeding the water quality standard for E. coli. Dry weather storm sewer discharges sampling indicated that over half of the discharging sewers had relatively high E. coli levels, in excess of the 394 cfu/100 mL standard. Model Development and Results Using the sampling data that were collected in 2001 in conjunction with historical data, a water quality model has been developed to simulate the fate and transport of E. coli in Whiteoak Bayou. The Hydrological Simulation Program – FORTRAN (HSPF) was selected to model the various bacterial point and non-point sources from April 1999 through September 2001. HSPF is a continuous, one-dimensional water quality model that was selected because of its capabilities to model build-up and wash-off as well as sediment-associated pollutants (Bicknell, Imhoff et al. 1996).Previous studies have concluded that simply modeling dissolved (in-stream) bacteria does not fully explain the in-stream concentrations (Socolofsky 1997, MapTech Inc. 2001). Additionally, findings from this study and the general literature strongly suggest that bayou sediment is a reservoir for bacteria and thus may significantly contribute to the observed water quality in the bayous (Baudart, et al 2000; Crabill, 1999; Buckley, 1998). For this reason, the developed model simulates both dissolved and sediment-bound indicator bacteria.This paper will present the model and its predictions as well as detailed sensitivity analyses on a watershed, and subwatershed specific, basis. Model parameters that have been determined to affect bacterial concentrations include bacteria surface accumulation parameters as well as sediment transport and soil detachment parameters. Allocation Scenarios Allocation scenarios have also been developed to examine the bacterial sources and their contributions to the load in Buffalo and Whiteoak Bayous. These scenarios were developed following the identification of major bacteria sources, i.e. the wastewater discharge, dry weather storm sewer discharge, upstream input from Addicks and Barker reservoirs and non-point sources. The dry weather storm sewer discharges were treated as a point source parameter, since they are an “end-of-pipe” discharge and thus are not considered in the non-point source scenarios. The non-point source (NPS) scenarios were simulated by reducing the bacterial accumulation on the pervious and impervious land. Although other nonpoint sources exist in the model, such as sediment build up and runoff, and sediment resuspension, these are not currently considered in the non-point source reduction scenarios. Preliminary results indicate that nonpoint sources contribute a large portion of the in-stream loading observed in Buffalo and Whiteoak Bayous.To address variability in flow conditions, both low and high flow situations are examined in the context of the allocation. The mechanisms maintaining elevated low flow E. coli concentrations are different than the mechanisms acting under high flow conditions. These mechanisms will be evaluated and presented in this paper, along with the implications for the TMDL allocation.