Did We Get It Right? Post Sewer-Separation Analysis in Boston, MA

This paper evaluates the effectiveness of both the sewer separation in the Stony Brook area of Boston, Massachusetts and the pre-separation GIS and modeling analysis used to plan the effort. The goal of this separation was to reduce CSO discharges to the Charles River from twenty-two to two or fewer overflows during a typical year as per the facilities plan developed by the Massachusetts Water Resources Authority and agreed upon by the EPA and Massachusetts DEP (MWRA, 1997).During preliminary design, extensive field survey data was linked with GIS analysis and collection system modeling to develop a sewer separation plan (BWSC, 1999 and Tsay et al, 2003). The sewer separation plan developed during the preliminary design in 1999 was implemented, with construction completed in 2006. To evaluate the effectiveness of the sewer separation, flow monitoring equipment was installed at seven regulators within the separated area.Comparisons of pre- and post-monitoring data indicate that sewer separation in the Stony Brook area performed as predicted by the planning model. Observed reductions in inflow were similar to the reductions predicted by the preliminary design modeling. The planning analysis was also successful in identifying the level of separation required in specific areas. Results of the planning analysis indicated that areas upstream of regulator RE-046-50 would require some disconnection of the more difficult to disconnect roof leaders. The cost prohibitive nature of these disconnections prevented this from taking place. Post-monitoring results indicated that RE-046-50 activated during the post-monitoring period. Modeling of post-separation conditions indicated that the activations were consistent with the level of separation actually achieved during construction. In lieu of additional separation, the Boston Water and Sewer Commission modified the regulator to reduce activations in this area to zero for a typical year.This study also demonstrates the importance of post-monitoring in sewer separation efforts even with successful planning. Analysis of data recorded at regulator RE-046-30 indicate that there was a decrease in inflow at the regulator consistent with values predicted during preliminary design. However a change in the flow regime downstream of the monitoring location was resulting in overflows at this regulator where there were previously none. Further investigation revealed that this regulator was affected by a large grease deposit which was eventually removed.