Alternative Procedures for Setting Reference TDS Endpoints in TMDLs

Abstract

Total dissolved solids (TDS) is an aggregate measurement of chemical salts dissolved in water. Elevated levels of TDS are typically associated with mining operations. State water quality standards are often used, both to trigger the assessment of water quality impairments in a stream, and to serve as the target total maximum daily load (TMDL) endpoint. In the case of biological impairments, where the causative pollutant(s) are identified through an analysis of all available data, pollutant(s) may be identified for which there are no state standards. This is the case with TDS in Virginia where there is neither a state water quality standard for TDS, nor NPDES mining permit limits. As a result, an alternative procedure must be used to identify an acceptable level of TDS in the stream. Three alternative procedures were explored during the development of a TMDL to address a biological impairment in Bull Creek in southwestern Virginia. These procedures included the use of two different models (GWLF and HSPF) and two different types of endpoints (concentrations and loads). The first two procedures used a “reference watershed approach”, where the average annual TDS load from a non-impaired, comparable reference watershed was used to set the load-based endpoint. The first procedure simulated loads simulated with the GWLF model for both the Bull Creek and an area-adjusted reference watershed (Upper Dismal Creek), while the second procedure used the HSPF model. The third procedure used available in-stream TDS monitoring data from a comparable, non-impaired reference watershed to set a concentration endpoint equal to the 90th percentile of the monitoring data (369 mg/L). For this third procedure, HSPF was used to simulate daily TDS concentrations in Bull Creek. Pros and cons of each procedure will be discussed, together with implications of the required reductions and the relative strength of justifications.