Comparison of tracer techniques for monitoring sewer losses

Abstract

Two different experimental techniques employing conservative tracers to determine exfiltration losses in a sewer system are reported. Both techniques compare the in-sewer responses measured along designated indicator and reference sections to assess the level of exfiltration although with different degrees of reliability. The experimental and interpretation errors associated with a pulse dosing technique, in which Rhodamine WT is used as the single tracer, are described. The estimation of exfiltration requires the comparison of indicator and reference peaks which need to be well resolved and reproducible in order to provide realistic results. A more consistent technique involves the continuous dosing of two different tracers with the objective of producing simultaneous downstream peak tracer concentrations. Any errors due to flow variations are eliminated and the longer measurement period allows both peak and background concentrations to be more accurately determined. Rhodamine WT, Li and Br were used in different combinations over the approximately 20 min duration of each continuous dosing experiment and were found to demonstrate conservative behaviours with measured recoveries of 97.6-100.6%, 97.3-101.4% and 108.4-109.2%, respectively. The analytical problems associated with the determination of the different tracers in sewage are described and are shown to be responsible for the reduced accuracy of the results associated with the use of Br. The preferred combination of tracers is Rhodamine WT as the indicator and Li as the reference and these are shown to be capable of estimating sewer losses to an accuracy of +/-1.0% using the continuous dosing approach.

Water level and discharge measurements

“…This provides no full formal validation. Different tracer combinations were used: LiCl, NaBr (Prigiobbe & Giulianelli, 2011;Rieckermann et al, 2007) and Rhodamine WT (Revitt et al, 2006).…”

Use of sewer on-line total solids data in wastewater treatment plant modelling

“…This can be done through a measurement series using a marking substance (Revitt et al 2006) or a clamp-on flow meter.…”