Woods Hole Coastal and Marine Science Center

next up previous contents
Next: Simulation of chlorination Up: No Title Previous: Statistical Comparison of

Effluent Modeling


Since the hydrodynamic model represents the stratification and statistics of low-frequency currents in the western part of Massachusetts Bay rather well, it is appropriate to use the model to investigate the transport of effluent from the existing and future outfalls over timescales of weeks to months. To model the effluent, a conservative tracer was added to the code, modeled the same as temperature and salinity, but with no dynamical effect. The effluent concentration was set to 100 at the Nut and Deer Island diffusers for the existing outfall run. For the future outfall run, the discharge from Nut and Deer Islands was added together and introduced at the future outfall, again with a concentration of 100. Along the open boundary, far from the sources, the effluent concentration was set to zero. Since the loading concentration is arbitrary, the effluent concentrations were converted to effluent dilutions for the purposes of presentation.

Since ECOM-si is a hydrostatic model with grid spacing of about 1 km in the vicinity of the future outfall, it cannot represent the physics of the turbulent entrainment process that occurs when the effluent is discharged from the diffusers. By designing the grid so that the size of the diffuser grid cell is comparable to the area of zone of initial dilution, however, it is possible to dilute the effluent by approximately the same amount as predicted by near-field models such as EPA's ULINE. This allows ECOM-si to effectively model the height to which the plume rises (Blumberg et al., 1994). Ph.D work by Xue-Yong Zhang (in preparation) at MIT has shown that with the relatively strong pycnocline observed in Massachusetts Bay there is considerable latitude in choosing the size of the diffuser grid cell to produce approximately the right trap height.

Effluent was discharged and tracked from the existing and future outfalls from October 1989 to September 1991, and the results have been presented in a variety of forms. Some animation clips comparing the two outfall locations during winter and summer periods can be seen on the World Wide Web at http://crusty.er.usgs.gov. The 200:1 dilution level is about the level at which the nutrient signal from the outfall should become lost in background variability, and thus is highlighted in presentations.

During the winter, highest effluent concentrations from both the existing and future outfalls occur at the surface, since the initially diluted effluent is lighter than the well-mixed, dense seawater (Figure 4.1).

 Diagrams of effluent comapirson in Winter 1990

The existing outfall, due to the shallow depth and confines of Boston Harbor, results in a much larger region of poorly diluted effluent. There is very little difference in the size of the 200:1 dilution region between 1990 and 1991, although the size of the 800:1 dilution region changes considerably (Figures 4.1 and 4.2).

 Diagrams of effluent comparison in winter 1991

During the summer, the highest effluent concentrations from the existing outfall are again found at the surface, since the effluent is effectively discharge directly into the surface layer due to strong tidal mixing in Boston Harbor. The 200:1 dilution region is very similar to the winter region (Figure 4.3).

 Diagrams of effluent comparison in summer 1990

At the future outfall, however, the effluent plume becomes trapped below the pynocline since the seawater near the surface has warmed and freshened to the point where it is lighter than the initially diluted effluent. At 16 m, the depth of maximum areal extent of the 200:1 effluent surface, the region affected by the future outfall is still considerably smaller than the surface extent of the 200:1 region from the existing outfall (Figure 4.4).

 Diagrams of effluent comparison in summer 1990

Average dilution comparisons for other seasons are presented in Appendix C.

While the dispersion of effluent can be characterized by the plots of the average effluent dilution, a residence time for effluent in Massachusetts Bay and Boston Harbor can be determined by dividing the total amount of effluent in the bay or harbor by the rate of loading. This calculation of residence time assumes that both the input effluent and export from the system are relatively steady. The boundary of Massachusetts Bay was specified as a straight line between Cape Ann and Provincetown, and the boundary of Boston Harbor was specified as a straight line across the harbor mouth. The residence times for Massachusetts Bay are very similar for the existing and future outfall scenarios, and show a distinct seasonal cycle (Figure 4.5).

 Plots of residence tie for effluent

Overall, the residence time for effluent in Massachusetts Bay is from 40--100 days, while the residence time in Boston Harbor is 5--9 days. The effluent has longer residence times in the winter and shorter residence times in the summer in both the harbor and the bay. Since the effluent discharge does not have a clear seasonal cycle, the change in residence time is presumably due to increased flushing. For the bay, the residence time calculated via this method changes over time scales that are comparable to or less than the residence time itself, indicating that an interpretation as total amount of effluent in the bay is more appropriate (since the loading is relatively constant). Effluent levels in the bay build during the fall and winter, reach a peak in March, then fall precipitously in April as the bay is flushed by energetic springtime currents. The lowest amounts of total effluent are found in summer, and as the water column destratifies, the effluent levels begin to climb again.

next up previous contents
Next: Simulation of chlorination Up: No Title Previous: Statistical Comparison of