Stormwater Best Management Practices in an Ultra-Urban Setting: Selection and Monitoring

Monitoring Case Study-Vertical Volume Recovery Structure (Vvrs), Orlando, Florida

This case study is based on an evaluation of a vertical volume recovery structure by Dyer, Riddle, Mills, & Precourt, Inc. (1996).

Due to chronic flooding problems within a highly urbanized section of the drainage basin of Lake Olive, Florida, the City of Orlando proposed to divert approximately 10 ha (25 ac) of the Lake Olive drainage basin to nearby Lake Lawsona. In addition to basin runoff diversion, it was requested that a system be designed that could provide both underground storage and treatment of the stormwater to reduce pollutants entering Lake Lawsona. The engineers at Dyer, Riddle, Mills, & Precourt, Inc. (DRMP) recommended investigating a sand filter structure known as the Vertical Volume Recovery Structure (VVRS). A VVRS, which is an underground stormwater runoff treatment and conveyance facility, combines in-pipe storage, a sump device (sediment settling), and a sand filter (fine sediment and pollutant removal) in its runoff treatment operation. The VVRS system for the City of Orlando was designed to treat 25.4 mm (1 in) of runoff from the diverted basin area, and was installed under Pine Street just west of Lake Lawsona.

Study Objectives

• Measure the effectiveness of the sump to remove larger sediment particles.

• Measure the effectiveness of the VVRS at removing a wide range of stormwater pollutants.

• Establish backwash requirements for the VVRS system designed for the City of Orlando.

Design and Operation

The VVRS system is located within the Pine Street right-of-way, between the intersection of Pine Street and Hyer Avenue, and just west of Lake Lawsona, in Orange County, Florida. The basic design of the system consists of a sump device, which acts as an initial settling basin, and a sand filled trench that acts as a filter to remove stormwater runoff pollutants. The system is designed to contain and treat the first 25.4 mm or the first inch of diverted runoff. Figure 61 shows a VVRS system prior to a rainfall event. The main weir crest acts as an overflow device for storm runoff occurrences greater than the design storm. The hypothetical flow path of a 25.4 mm ( in) or greater storm event through the VVRS system is shown in Figure 62. The sump is used to capture and settle larger sediments carried through the storm sewer (see Figure 61, station 1). The slope of the 600 mm (24 in) connector line (see Figure 61) prevents the transmission of floatables to the VVRS; however, in the event that floatables reach the filter box and collect within the sand pores, they can be removed by backwashing the filter using the effluent pipe that drains the VVRS. This was an important design feature of the VVRS. To accomplish this, a backflush system was constructed that forces potable water up through the sand medium to dislodge the accumulated particles. The sump area was designed to be cleaned using the opening through the main overflow weir structure. Figure 63 shows a schematic of the VVRS backwash operation.

Figure 61. Vertical Volume Recovery Structure system at rest prior to rainfall event (adapted from Dyer, Riddle, Mills, & Precourt, 1996)

Figure 62. Vertical Volume Recovery Structure full utilization of treatment volume (Dyer, Riddle, Mills, & Precourt, 1996)

Figure 63. Vertical Volume Recovery Structure backwash operation (Dyer, Riddle, Mills, & Precourt, 1996)

Two significant operational problems were observed during the monitoring phase of the VVRS study. The first problem was structural failure of the sump device. After a storm event of approximately 38 mm (1.5 in), the sump structure was observed to have a large crack in its foundation floor. The problem was corrected by pouring additional concrete onto the slab of the slump; however, this reduced the depth and consequently the storage volume of the sump. The second problem was the surface blinding of the filter medium. During the first several storm sampling events, the sand filter medium was found to have an unacceptable rate of drawdown. After careful analysis of the filter medium, it was determined that the surface of the medium was "blinding", preventing significant flow through the system. The manufacturer of the filter medium recommended that a larger grain size medium be placed over the existing sand. This would allow the larger particles that were blinding the fine grained sand to be trapped without reducing the flow rate. It was determined that anthracite coal would be used because it was less dense than the existing sand medium, and would remain on top during backwashing of the system. Three-tenths of a meter (1 ft) of the existing sand was removed and replaced with the coal. After the anthracite coal was installed, field inspection of a storm event found surface blinding was still a factor. The sampling program was continued by frequent probing of the filter during bleed down, and frequent backwashing after storm events to reduce the effects of surface blinding and improve the drawdown rate.

Monitoring Program

Three monitoring locations were established within the VVRS system to determine the pollutant removal efficiency. Station 1 is located in the inflow storm sewer pipe, and samples the stormwater runoff entering the system. Station 2 is located within the filter box, above the filter media. Station 3 is located in the outfall pipe, just beyond the filter media. All three sampling stations are schematically shown in Figure 61. Composite samples were collected from station 1 during six storm events, and stations 2 and 3 during 11 storm events. A storm event was defined as a minimum of 2.54 mm (0.10 in) of rainfall, separated from other storm events by at least 24 hours. The pollutant removal efficiency results of the monitoring study are given in Tables 36, 37, and 38. Table 37 shows the mean removal efficiency for three sampling events at the sump device. Only three samples from the six storm events at station 1 were used due to limitations on paired samples from stations 1 and 2. Table 37 gives the mean removal efficiency for 11 sampling events at the filter mechanism, and Table 38 shows the mean pollutant removal efficiency for the entire VVRS system. Mean removal efficiencies were calculated using a mass balance approach. The mean efficiency was determined as the sum of the individual removals divided by the total number of events.

Conclusions

• The sump device is somewhat effective at removing several heavy metals (cadmium, lead, and zinc), total nitrogen, total phosphorus, orthophosphate, and total suspended solids. The slow bleed down time associated with the blinding of the filter media may have contributed to this result.

• The filter media is almost completely ineffective at removing dissolved or suspended pollutants. The drawdown problems associated with the blinding of the filter media may be partly responsible for this result. When the stormwater sat in the filter box for extended periods, much of the settling and chemical action may have occurred prior to the water flowing through the filter.

• The use of larger grain size filter media would result in a lower filtering efficiency, but would likely eliminate the surface blinding problem.

• Pulsing, which is a very brief backwash event triggered automatically, can be used in lieu of using the larger grain size filter media.

References

Dyer, Riddle, Mills, & Precourt, Inc. 1996. Vertical Volume Recovery Structure. Final Report, Project No. WM436. Prepared for City of Orlando Stormwater Utility Bureau. Submitted to the Florida Department of Environmental Protection.