ASSESSMENT OF PARTICULATE AND DISSOLVED DIOXIN CONCENTRATIONS IN STORM WATER RUNOFF TO HOUSTON-AREA WATERWAYS USING HIGH VOLUME SAMPLING

Polychlorinated dibenzo-p-dioxins and dibenzofurans (“dioxins”) are extremely persistent in the environment, and can affect human health at low concentrations. As a result of dioxin found in seafood organism tissue in the Houston Ship Channel (HSC) and upper Galveston Bay, a seafood consumption advisory was issued by the Texas Department of Health in September 1990. Subsequently, the HSC system was placed on the state's Section 303(d) list and a Total Maximum Daily Load (TMDL) study was initiated.Runoff sampling was needed to measure the dioxin load in storm water runoff from drainage areas in the impaired watersheds. Since dioxin levels in storm water runoff are extremely low and are typically lower than the analytical detection limit for dioxin, storm water sampling was conducted using a high-volume technique. In this technique a large volume of runoff water is slowly pumped through a stainless steel and Teflon® apparatus to allow particulate bound dioxins to be trapped on a glass fiber filter and dissolved dioxins to be adsorbed to a polymeric resin. After pumping, dioxins on the filter and resin are extracted using a non-polar organic solvent. The extraction liquid is then analyzed using EPA Method 1613B to quantify total dioxin levels on both media. The reported congener mass was then divided by the total volume of runoff water pumped through the sampling apparatus to obtain the dioxin concentration in runoff.In this study, runoff volumes of 700 liters were typically obtained and the effective analytical detection level was reduced from the method reporting limit of 10–50 picograms per liter (pg/L) to 0.03 pg/L. To minimize adsorption column breakthrough of dioxin and to avoid unreasonably long staffing durations, pumping rates were limited to 1.6 liters/min resulting in a pumping cycle of 7.3 hours for each runoff event.Storm water runoff site selection was challenging due to Houston's flat topography and extensive network of wastewater treatment facilities. The challenge was in identifying watersheds that were large enough to have runoff hydrographs that persisted long enough for sampling, but still did not have wastewater discharge. The project team used mapping information available from local agencies to identify 25 potential sites on drainage facilities that did not receive wastewater flows. These 25 sites were assessed for safety, access, and configuration and were ground-truthed by site visit and photography. Data sheets were created for all sample sites that utilized a quantitative scoring method for the aforementioned categories. This scoring system provided a quantitative selection process used to identify the most feasible and representative sample sites. Ultimately, ten sampling sites were selected from subwatersheds ranging from 3 to 5 square miles in size.Another complicating factor involves precipitation patterns in the Houston area. Rain events are infrequently associated with large, slow moving frontal systems. Instead, Houston receives rain from localized storm systems that move in from the Gulf of Mexico and from other weather patterns from the west. These rain patterns necessitated the development of a storm tracking and mobilization Decision Flow Chart.A total of ten sites were sampled. The results show total Toxicity Equivalency (TEQ) concentrations between 0.026 to 0.567 pg/L. Eight of the ten results exceeded the Texas human health-based water quality criteria for dioxin of 0.093 pg/L.This paper will provide details regarding the site selection process, sampling activities and challenges, flow measurement results, observed suspended sediment levels in runoff flows, observed partitioning of dioxin between solid and dissolved phases, and dioxin concentrations.