At Gemini Elementary School in Melbourne Beach, FL, there is a history of repeated flooding on the school grounds and in adjacent properties. In 1999, Creech Engineers Inc. (CEI) was chosen by Brevard County Regional Stormwater Utility Department to design drainage improvements to alleviate these flooding conditions, as well as to provide for stormwater treatment within this 49.6-ac. drainage basin. The project was divided into two phases: Phase 1 made improvements in order to accelerate initial flood-control measures for homes downstream of the school; Phase 2 involved the design of more extensive flood and water-quality control measures along Oak Street for further protection of school property and roadway flooding at a nearby church property. This article highlights the political challenges of retrofitting stormwater systems in developed areas and demonstrates a methodology for performing a nonpoint-source pollutant loading analysis. Figure 1. Oak Street Proposed Improvements
Existing Conditions Gemini Elementary School is located on a 19.8-ac., triangle-shaped property along the south side of Oak Street, a two-lane collector road, about a half mile from the Atlantic Ocean (see Figure 1). Residential properties lie downstream of the school, along its southeast and southwest borders. The 21-ac. Doug Flutie Park is on the north side of Oak Street. A soccer club uses the park and school grounds on a daily basis. There was no stormwater system at the park, along Oak Street, or on the school site. Stormwater flowed southward from Doug Flutie Park, across Oak Street, through the school site, and into the yards and homes south of the school. These yards, and the roads downstream of them, are very flat and only a few feet above sea level. Once water stages high enough in the yards, it gradually sheet flows down the adjacent roads a few hundred yards to the Indian River. The affected homeowners naturally blamed the school for allowing its water to flood them.West of the school, a few hundred yards along Oak Street, was a low point in the road–where water ponded and flooded the road–and an adjacent churchyard. Because a thin clay lens at 10 in. deep caused a perched water table, water stood in the road for several days after even a nominal rainfall. This drainage basin was almost completely built out, with no easy path for developing outfalls to relieve flooding.This section of the Indian River is a Class 2 water body, with a shellfish harvesting classification bringing intense scrutiny from the St. Johns River Water Management District. Because of seagrasses near the shoreline, a permit from the United States Army Corp of Engineers (USACE) is required for new outfalls in the area.The park, the school, and Oak Street lie in unincorporated Brevard County. The church and properties west of the school are in Melbourne Beach. Because Oak Street is a collector road, all of the utility companies have major transmission lines in the right of way.This challenging project involved Brevard County, Melbourne Beach, the school board, the Brevard County Parks and Recreation Department, the Brevard County Road and Bridge Department, the Brevard County Regional Stormwater Utility Department, a church, three different homeowners associations, a soccer club, the St. Johns River Water Management District, USACE, and several utility companies. Stakeholder involvement and partnerships were going to be critical to weave a solution through the many players involved.Proposed Improvements
Dry pond wrapped around playground and soccer field
Dry pond, in baseball outfield, and control structureThe first priority was to alleviate flooding in the homes adjacent to the school. As an interim measure, a berm was designed and constructed by county personnel along the south property lines of the school, with a swale behind the berm directing water to the southernmost point of the school property. At that location, an inlet and an 18-in. outfall pipe were constructed in a utility easement through two heavily landscaped and fenced yards to Pompano Street, where it was tied to an existing storm-drain pipe.A short time later, heavy rains overflowed the berm and swale and again flooded homes adjacent to the school. CEI was engaged at that point to provide more effective drainage improvements.Fortunately, Gemini Elementary School had a significant area of vacant land on its site. The school entered into agreements with Brevard County, allowing the construction of three dry retention ponds totaling 7.3 ac. to reduce flows leaving the school site and to provide stormwater treatment where none existed. These dry ponds were wound around several soccer and baseball fields. The soccer fields’ locations had to remain in place due to previous agreements with the school and the Brevard County Parks and Recreation Department. The ponds were only 10-16 in. deep and sodded, allowing the soccer teams to use the pond areas as practice fields when dry. When the ponds were excavated, the confining clay layer was removed to allow for infiltration through the beach sand at the site. Construction was scheduled during the summer when school was out.A control structure was designed at the outfall-pipe location to provide protection for a 25-year storm. The temporary connection to the existing downstream pipe had overloaded the downstream system in a heavy-rain event, so a new outfall to the Indian River was designed through a park adjacent to the river. The park was owned by a homeowners association, which reluctantly gave a drainage easement through the park. The county agreed to make several improvements to the park and its boat ramp in exchange for the easement. USACE was concerned that the new outfall pipe discharges would impact the nearby seagrasses, so the new discharge pipe was not permitted to be constructed in the Indian River. A bubble-up box was designed 10 ft. back from the shoreline, and rock riprap was placed between the bubble-up box and the mean high water line to prevent erosion. As mitigation for disturbing the shoreline, spartina and other plants were planted among the rocks to further buffer the shoreline from the stormwater discharges.
Phase 1 of improvements was finished in September 2000 at a cost of $124,000. The improvements proved successful in preventing any flooding of adjacent homes in several large rainfalls in 2001.Phase 2 of the project addressed stormwater quantity and quality concerns along 1 mi. of Oak Street, from Highway A1A to Cherry Street. To provide further flood protection at Gemini Elementary School, retention swales were designed along both sides of Oak Street, and 684 yd. of storm-drain pipe was designed to intercept runoff and prevent it from crossing the road onto school property. The piping also provided an outfall for the low spot in the road by the church. This new pipe system discharged into a residential canal system, which was used by many of the adjacent residents for boating to the Indian River Lagoon. These canals were very politically sensitive because they needed dredging, and the Town of Melbourne Beach does not dredge canals. The residents were concerned that the new stormwater system would lead to further sedimentation of the canals. The first alternative for treatment was to use land at the church site for a pond for the road runoff. The church was willing to donate the land where its septic-tank fields were located if the county would provide a sewer connection. This scenario was designed, but when it came time for the church to give easements to the county, the church balked, and it was back to the drawing board.St. Johns River Water Management District criteria require stormwater treatment for improvements that increase discharge rates, pollutant loadings, or impervious areas. With this project, no new increased impervious areas were proposed, but there would be additional water flowing to the residential canals from the extension of the pipe system to the flood-prone areas. These new flows create the potential for increased pollutant loadings to the canal. Normal design methods would have used treatment ponds to offset these potential impacts, but due to a lack of available land for ponds, alternative treatment methods were proposed for this project. The district will consider alternative treatment methods if it can be demonstrated that all other possible alternatives have been exhausted. It would not be possible politically to use more school or park area for treatment ponds. For this project, CEI showed that the only alternatives were to tear down houses for ponds or use alternate treatment technologies. The treatment strategy involved maximizing treatment methods within the project basin with alternative best management practices (BMPs), as well as retrofitting two adjacent watersheds as additional mitigation. A total of 1.67 ac.-ft. of retention storage was provided in Phase 2 in the roadside swales and small ponds. This was equivalent to 0.032 in. of retention from the drainage areas flowing to the retention areas. A treatment train along Oak Street was designed by using nine grated inlet skimmer boxes from Suntree Technologies Inc. in the new inlets to trap debris entering the inlets, constructing berms to slow runoff from the ball fields, and installing one baffle box at the downstream end of the new pipe system along Oak Street. Baffle boxes are inline stormwater treatment devices that trap sediment, trash, and debris. Brevard County has used them successfully for the last nine years. In offsite Basin 4, which had only one existing baffle box to provide sediment removal, 16 curb inlet skimmer boxes were installed in all of the existing inlets to provide nutrient removal by trapping grass clippings, leaves, and yard debris. Nutrients were a concern in the canals because the nutrients promote algae blooms and low dissolved oxygen, which in turn increase muck buildup in the canals. In offsite drainage Basin 5, there are three existing pipes that discharge directly to the canals. Three baffle boxes and six curb inlet skimmer boxes were designed to provide sediment and nutrient treatment for this drainage basin. The Brevard County Regional Stormwater Utility Department will implement this project and be responsible for all maintenance of the improvements. The baffle boxes will be inspected twice a year and cleaned as needed. The inlet traps will be cleaned twice a year. Brevard County has a vacuum truck dedicated to cleaning stormwater BMPs.The numerous BMPs used on this project provided a high degree of treatment for the new piping system along Oak Street and provided treatment for two offsite basins where little treatment existed. The retrofitting of the offsite areas was, in effect, mitigation for the new discharges to the canal. See Figure 1 for a map of the improvements.CalculationsIn Phase 1 of the project, the dry ponds and outfall pipes were modeled hydraulically using the Interconnected Pond Routing program. Because the dry ponds in the Phase 2 project area were too small to provide effective attenuation, the predevelopment and postdevelopment runoff calculations were made using Hydraflow and the rational method. The only available storm-drain pipe for Phase 2 was a 36-in. pipe in offsite Basin 4. The new piping along Oak Street was connected to the existing 36-in. pipe, and the piping downstream of the connection was upgraded to a 42-in. pipe. The pipes were designed for a 25-year storm. Basins 1, 2, and 3 were much farther from the outfall than was Basin 4. As a result of different times of concentration, the peak flows from Basin 4 passed sooner than did Basins 1, 2, and 3, giving only a slight increase in peak discharge despite adding 30.3 ac. to the area flowing to the existing outfall.The potential for increased pollutant loadings in the canal system was a concern of local residents. These canals had a history of dredging operations every eight to 10 years, and the residents did not want to increase the frequency of costly dredging. The main pollutants of concern leading to muck deposition in the canals were total suspended solids (TSS), total nitrogen (TN), and total phosphorus (TP). Sediment buildup at the end of the pipes was common. Most of the material dredged from residential canals is typically muck. To address this concern, a pollutant-loading analysis of the existing and proposed stormwater discharges was performed. In the existing conditions, the only stormwater treatment for the canal system was a baffle box along Cherry Street for offsite Basin 4 of 59.9 ac. A total of seven outfall pipes discharged into the canal system.In Phase 1 of this project, stormwater treatment was provided for 19.8 ac. of the school grounds with three dry detention ponds. The discharge from these ponds was to the Indian River rather than the canal system, so these pollutant loads were not included in the pollutant-load analysis for the canal outfall.The existing pollutant load to the canal came only from the drainage Basins 4 and 5, totaling 77.1 ac. The runoff from Oak Street did not drain to the canal in existing conditions, but only in the postdevelopment conditions. The strategy for the pollutant analysis was to calculate the pollutant loads in the existing conditions and then calculate the pollutant loads after the new pipes were added to the system and offsite areas were retrofitted for stormwater treatment. The pollutants used in this analysis were TSS, TP, and TN.Each drainage basin was categorized by land use. Areal, annual, mass loading rates from Stormwater Loading Rate Parameters for Central and South Florida (Harper, 1994) were multiplied by each basin’s area to give existing and potential annual pollutant loadings (see Table 1).The next step was to calculate the pollutant removal rates for the different BMPs. Individual BMP removal efficiencies were take from Guide for Best Management Practice (BMP) Selection in Urban Developed Areas (American Society of Civil Engineers, 2001). What was challenging with this analysis was the use of multiple BMPs in series for the treatment train. Each BMP receives cleaner and cleaner water as the water moves down the train. At each BMP, the removal efficiency for each constituent was multiplied by the remaining percentage of the initial loading to give a weighted, cumulative removal efficiency for each constituent (see Table 2). These calculated removal efficiencies were then multiplied by the total calculated pollutant loads to give the reduced pollutant loadings after the BMPs were installed (see Table 3). Table 4 shows that the total loads to the canal were reduced as a result of the retrofitting of onsite and offsite basins.The pollutant-loading analysis in Table 1 demonstrates that, as a result of the numerous BMPs proposed, the total pollutant loadings entering the canals after project completion will actually be significantly reduced from the existing pollutant loadings entering the canals. The key to overall pollutant reduction is to provide additional treatment in offsite drainage basins. This will result in a net benefit of reduced pollutants entering the canals and a reduction of the severe flooding often seen along Oak Street.Table 4. Net Pollutant RemovalsTSS (kg/yr.) TP (kg/yr.) TN (kg/yr.) Predevelopment 3015.78 35.13 380.83 Postdevelopment 630.97 21.95 289.15 Net Reduction 2384.81 (79%) 13.18 (37.52%) 91.68 (24.07%) SummaryThe days of solving flooding problems in communities with simple ditch and pipe solutions have disappeared. Environmental concerns now dictate that stormwater treatment techniques be integrated into these flood-relief projects. By adding water-quality components to water-quantity projects, communities can help achieve pollution remediation goals being established for National Pollutant Discharge Elimination System, total maximum daily load, and pollutant load reduction goal programs. Retrofitting existing stormwater systems to provide water-quality treatment is more complicated, expensive, and time-consuming than are traditional stormwater designs for new development. The scarcity of available land and the numerous existing utilities in older, built-out areas will tax an engineer’s imagination to provide innovative BMPs in these locations. A carefully planned treatment train was designed consisting of swales, ponds, berms, baffle boxes, and inlet traps to provide overall stormwater pollution reduction.To address stormwater pollution concerns, treatment mitigation was designed in offsite drainage basins. The pollutant loadings and removals were calculated using a simple but effective spreadsheet analysis incorporating the latest in BMP efficiency studies. While complicated stormwater modeling software can be used for pollutant analysis, this type of modeling is more cost-effective on large basin studies than on small basins and individual projects. The pollutant-removal calculations showed an annual net reduction of 79% for TSS, 37% for TP, and 24% for TN in the Oak Street basin despite the creation of a new storm drain system for a landlocked area.As this project demonstrates, there are typically numerous stakeholders who need to be brought into the project early in the process and kept in the process throughout the life of the project. Many meetings were held with city, county, and state officials, homeowners associations, schools, soccer clubs, churches, and utility companies. All it takes is one uncooperative stakeholder to set back or kill a project, as was demonstrated with the church backing out of the land acquisition process after many verbal indications of approval. Using creative partnerships with other entities and agencies allowed the development of a unique strategy to solve flooding at several locations in the project area.
