Determination of time-of-travel, dispersion characteristics, and oxygen reaeration coefficients during low streamflows--Lower Tacony/Frankford Creek, Philadelphia, Pennsylvania

Abstract

Time-of-travel, dispersion characteristics, and oxygen reaeration coefficients were determined by use of dye and gas tracing for a 2-mile reach of Tacony/Frankford Creek in Philadelphia, southeastern Pennsylvania. The reach frequently has concentrations of dissolved oxygen (DO) below the water-quality standard of 4 milligrams per liter during warm months. Several large combined sewer overflows (CSOs), including one of the largest in Philadelphia (former Wingohocking Creek), discharge to the study reach in this urbanized watershed, affecting water quality and the timing and magnitude of storm peaks. In addition, a dam that commonly results in backwater conditions and reduced natural reaeration is present a few hundred feet from the end of the study reach. Time-of-travel and reaeration data were collected under base-flow conditions in August and September 2009 for three sub-reaches from Roosevelt Boulevard (U.S. Route 1) to Castor Avenue. Determination of traveltimes to the centroid of the dye cloud were needed for calculation of the reaeration coefficients. Results of the dye study in Tacony/Frankford Creek indicate that traveltimes were affected by the presence of man-made structures, such as the large scour hole and pool developed at the outfall of the T14 CSO and the dam, both of which reduce stream velocities. Mean stream velocities during the dye-tracer tests ranged from a maximum of 0.44 to 0.04 foot per second through a large pool. The dispersion efficiency of the stream was determined from relations between normalized unit concentrations to time to peak for use in water-quality modeling. Oxygen reaeration coefficients determined by a constant rate-injection method using propane as the tracer gas were as low as 0.04 unit per hour in a long pool affected by backwater conditions behind a dam. The highest reaeration coefficient was 2.29 units per hour for a steep-gradient reach with multiple winding channels through gravel deposits, just downstream of a large scour pool developed at the outlet of the T14 CSO. Reaeration coefficients determined from the field tracer-gas method were compared to values calculated by two other methods, one that is based on theoretical equations using physical properties of the stream as variables and the other that is based on equations using the timing of measured daily maximum DO concentrations in the stream. Reaeration coefficients from the two alternate methods were most similar to values determined from the field tracer-gas method for the upstream portion of the study reach, characterized by free-flowing riffle and pools. Values of reaeration coefficients determined by the tracer-gas method were 2 to 10 times higher than values determined by 2 alternate methods for most subreaches hydraulically affected by man-made structures. In addition to the tracer gas, propane, the gas analysis also included methane, ethane, and ethene, of which only methane was measured in concentrations above a few micrograms per liter. Methane, thought to occur naturally or because of ongoing processes in the stream, was measured in concentrations ranging from 6.6 to 78 micrograms per liter; the concentrations were greatest in sub-reaches dominated by pools.