During 2004, BaySaver Technologies, Inc. began a thorough series of laboratory tests with the University of Minnesota’s St. Anthony Falls Laboratory (SAFL). SAFL is an internationally known hydraulics laboratory that has extensive experience in academic-industrial partnerships. The project was conducted by Dr. Omid Mohseni, the laboratory’s Associate Director of Applied Research.

SAFL researchers began testing the standard BaySaver system using an F-95 sediment gradation in August, 2004. At the same time, researchers created an empirical model of the system based on experimental data. This model was used to quantify the flow rates through the different system components under varying flow conditions. After the model and initial testing were completed, research was focused on optimizing the design. After two years of work with SAFL, BaySaver is introducing the BaySeparator™ System

The BaySeparator™ system is based on the same principles and protected by the same patent as the original BaySaver Separation System. However, modifications to the separator unit have improved both the flow capacities and the sediment removal efficiencies of the system. The system has been extensively modeled and tested in the laboratory, and this research program has resulted in a superior product.

A 24″ system was constructed in the laboratory. This system comprised the 24″ separator unit as well as two fiberglass manholes. The system was tested with both 48″ and 60″ manholes. Tests were run at varying flow rates to establish the efficiency under a range of operating conditions. Once flow began, the system was run until steady state conditions (verified with a salt tracer) were established. After steady state was reached, sediment was introduced into the inlet pipe by a metered sediment feeder. The target influent concentration was 200 mg/l, and this concentration was confirmed by grab samples taken from the influent water. The system was allowed to run for a given length of time before the flow was cut off. Following the test run, the manholes were dewatered and the mass of collected sediment was measured. This mass was compared to the total influent sediment load to calculate removal efficiency.

F-95 sediment is a commercially available mix that contains sediments ranging in size from 53 microns to 425 microns. The bulk of the sediment (87%) is between 75 microns and 212 microns in diameter. Table 1 shows the sediment grain size distribution for F-95 mix used during the tests. The F-95 sediment gradation has a d50 of 125 microns.

A number of tests were run on the 24″ laboratory installation. The first of these series of tests was run on the 24″ BaySeparator™ system with two 72″ manholes. Six tests were conducted on this configuration: two tests at 100% of the unit’s maximum treatment rate (MTR); two tests at 50% MTR; and two tests at 25% MTR. MTR is defined as the maximum flow the unit can treat without bypassing any water during high intensity storm events. The influent concentration of all tests was set at about 200mg/l with the F-95 gradation.

The second series of tests featured the same 24″ Separator Unit and 72″ Storage Manhole, but with a 48” Primary Manhole. Four tests were conducted in this configuration, two at 100% MTR and two at 15% MTR. Each test again had an influent concentration of approximately 200 mg/l of F-95 sediment gradation.

For each test run, three removal values were calculated: the fraction of sediment removed by the Primary Manhole; the fraction of sediment removed by the Storage Manhole; and the overall removal efficiency of the system. The fraction of sediment removed in each manhole is calculated by dividing the total mass of sediment introduced by the mass of sediment retained in each manhole. The overall efficiency of the system is calculated by dividing the total mass of sediment introduced by the total mass of sediment collected in *both* manholes. A brief summary of the test results can be found in Table 2.

Calculating these numbers using mass balances rather than grab samples or composite samples provides a much more robust and accurate dataset and reduces to a large extent the potential for sampling errors common in stormwater sampling projects.

SAFL researchers established a relationship between the sediment removal in each manhole and the Peclet Number in that structure. The Peclet Number is a dimensionless characteristic number of fluid flow that represents the ratio of advection to diffusion within a fluid system. In the case of the BaySeparator™ system, advection is the settling of sediment particles, while diffusion is measured with a turbulence factor 1. The Peclet Number for a manhole is a function of the manhole dimensions (depth and diameter), the settling velocity of the target sediment particle, and the flow rate through the manhole. Note that, for a given flow rate, each manhole in the BaySeparator™ system will have a different Peclet Number.

Separate sediment removal functions were developed for each manhole. The sediment removal in each manhole is expressed as a function of the Peclet Number, which is in turn a function of the flow rate through the manhole. These functions can be combined with the hydraulic model developed by SAFL to determine the removal efficiency of a given system over a range of flow rates. Because of the variability of manhole sizes and flow rates, each configuration has a slightly different flow rate vs. efficiency function. However, all of the functions are of the form shown in Equation 1 and Figure 2 below.

**Figure 2: Typical BaySeparator**™** Function**

In Equation 1, *E* is the removal efficiency of the system, *Q* is the flow rate through the system, *MTR* is the maximum treatment rate of the BaySeparator™ unit, and *m* and *b* are constants that depend on the configuration of the BaySeparator™ system. The value of *m* varies between -0.261 and -0.386 while *b* falls between -0.105 and 0.825. For each BaySeparator™ configuration, this function describes the performance of the system over the range of design flows. A typical function is shown above in Figure 2.

As expected, the function indicates that the BaySeparator™ system’s sediment removal efficiency increases as the flow rate through the system decreases. Low flow rates typically correspond to the more frequent, low intensity storms on the site. As the flow rate through the system increases, the system’s performance decreases. At the same time, low intensity storms represent 90% or more of the storm events on a site. To quantify the rainfall patterns on a site, BaySaver uses precipitation databases going back more than 45 years. These databases have been reviewed for integrity and consistency by BaySaver Technologies’ engineers. This distribution of storm events is the basis for BaySaver Technologies’ recommended Annual Aggregate Removal Efficiency sizing methodology.

Cost-effective BaySeparator™ systems can be designed for most sites by taking advantage of the frequency of low-intensity storms. In most jurisdictions, BaySeparator™ systems are designed to remove 80% of the suspended sediment load on an annual aggregate basis. In addition to the 80% annual aggregate removal, the system must also be capable of conveying the peak design flow rate during bypass, and the head loss through the system must be low enough to avoid backing up the flow upstream.

The peak design capacity of the BaySeparator™ determines the minimum separator size. Each separator unit has a maximum treatment rate (MTR) associated with it as well. Using the Rational Method, this MTR flow can be translated into rainfall intensity on the design site. The Rational Method, show below in Equation 2, is a hydrologic computation used to relate

runoff flow rate to rainfall intensity and the characteristics of the site. In Equation 2, *Q* is the runoff flow rate; *c* is the runoff coefficient (a constant between 0 and 1 that represents the fraction of total precipitation that runs off the site); *i* is the rainfall intensity on the site, and *A* is the drainage area of the site. Given *Q* (the MTR of the selected BaySeparator™), *c*, and *A*, we can rearrange Equation 2 and solve for i, as shown in Example 1.

On a typical site, the vast majority of precipitation comes at intensities far below the calculated intensity of 1.01 inches per hour. Figure 3, for example, shows the precipitation distribution for Nashville, Tennessee. As that plot demonstrates, approximately 90% of the total precipitation in Nashville falls at an hourly intensity below 0.71 inches per hour.

To include the distribution of precipitation in the sizing methodology, it is necessary to determine the fraction of precipitation falling at incremental intensities between 0 and the intensity associated with the MTR of the BaySeparator™. Example 2 shows this calculation, using the rainfall data from Nashville shown in Figure 3. The total amount of precipitation falling on the site is divided into 10 intensity increments. The lowest intensity increment, which corresponds to rainfalls between 0.01 and 0.10 inches per hour, contains more than 30% of the total precipitation that falls on the site. The second increment, rainfalls between 0.11 and 0.20 inches per hour, contains over 20% of the total precipitation, and subsequent increments contain less. For each increment, the fraction of total precipitation falling at that intensity is determined from the rainfall record.

The removal efficiency of the system is determined for the flow rate associated with each particular increment, and the percent of the sediment load for that increment is calculated by multiplying the fraction of precipitation by the incremental removal efficiency. In Example 2, 23.2% of the total precipitation falls within the intensity range between 0.01 and 0.10 inches per hour. According to the efficiency function for a BaySeparator SA30457.0 system, runoff generated by precipitation in this intensity range is treated at an efficiency of 99%. Therefore,

22.9% of the total sediment load (23.2% * 99%) is removed from these flows. The annual

aggregate removal efficiency of the system is calculated by adding together the ten incremental load reductions.

For sites in ecologically sensitive areas or those with particular runoff concerns, the BaySeparator™ system may be designed to remove a given fraction of the sediment load at a specified flow rate. This methodology is usually reserved for sites that discharge into wetland watersheds, fish spawning areas, or other critically sensitive drainages.

*Dhamotharan, S., Gulliver, J., Stephan, H., Unsteady One-Dimensional Settling of Suspended Sediment, Water Resources Research, Vol. 17 (4), pp 1125-1132 (1981).*