E. Storm Water Runoff Management Measure

Boat hull maintenance areas can be designed so that all maintenance activities that are significant potential sources of pollution can be accomplished over dry land and under roofs (where practical), allowing the collection and proper disposal of debris, residues, solvents, spills, and storm water runoff. Boat hull maintenance areas can be specified with signs, and hull maintenance should not be allowed to occur outside these areas. The use of impervious surfaces (e.g., cement) in hull maintenance areas will greatly enhance the collection of sandings, paint chips, etc. by vacuuming or sweeping.

Source control practices prevent pollutants from coming into contact with runoff. Sanders with vacuum attachments are effective at collecting hull paint sandings (Schlomann, 1992). Encouraging the use of such sanders can be accomplished by including the price of their rental in boat haul-out and storage fees, in effect making their use by marina patrons free. Vacuuming impervious areas can be effective in preventing pollutants from entering runoff. A schedule (e.g., twice per week during the boating season) should be set and adhered to. Commercial vacuums are available for approximately $765 to $1065 (Dickerson, 1992), and approximately one machine is needed at a marina of 250 slips or smaller. Tarpaulins may be placed on the ground prior to placement of a boat in a cradle or stand and subsequent sanding/painting. The tarpaulins will collect paint chips, sanding, and paint drippings and should be disposed of in a manner consistent with State policy.

Sand filters (also known as filtration basins) consist of layers of sand of varying grain size (grading from coarse sand to fine sands or peat), with an underlying gravel bed for infiltration or perforated underdrains for discharge of treated water. Figure 5-2 shows a conceptual design of a sand filter system. Pollutant removal is primarily achieved by "straining" pollutants through the filtering media and by settling on top of the sand bed and/or a pretreatment pool. Detention time is typically 4 to 6 hours (City of Austin, 1990), although increased detention time will increase effectiveness (Schueler et al., 1992). Sand filters may be used for drainage areas from 3 to 80 acres (City of Austin, 1990). Sand filters may be used on sites with impermeable soils since the runoff filters through filter media, not native soils. The main factors that influence removal rates are the storage volume, filter media, and detention time. Three different designs may be appropriate for marina sites: off-line sedimentation/filtration basins, on-line sand/sod filtration basins, and on-line sand basins. Performance monitoring of these designs produced average removal rates of 85 percent for sediment, 35 percent for nitrogen, 40 percent for dissolved phosphorous, 40 percent for fecal coliform, and 50 percent to 70 percent for trace metals (Schueler et al., 1992).

Sand filters become clogged with particulates over time. In general, clogging occurs near the runoff input to the sand filter. Frequent manual maintenance is required of sand filters, primarily raking, surface sediment removal, and removal of trash, debris, and leaf litter. Sand filters appear to have excellent longevity because of their off-line design and the high porosity of sand as a filtering medium (Schueler et al., 1992). Construction costs have been estimated at $1.30 to $10.50 per cubic foot of runoff treated (Tull, 1990). Significant economies of scale exist as sand filter size increases (Schueler et al., 1992). Maintenance costs are estimated to be approximately 5 percent of construction cost per year (Austin DPW, 1991, in Schueler et al., 1992).

Wet ponds are basins designed to maintain a permanent pool of water and temporary storage capacity for storm water runoff (see Figure 5-3). The permanent pool enhances pollutant removal by promoting the settling of particulates, chemical coagulation and precipitation, and biological uptake of pollutants and is normally 1/2 to 1 inch in depth per impervious acre. Wet ponds are typically not used for drainage areas less than 10 acres (Schueler, 1987). Pond liners are required if the native soils are permeable or if the bedrock is fractured. Design parameters of concern include geometry, wet pond depth, area ratio, volume ratio, and flood pool drawdown time. Ponds may be designed to include shallow wetlands, thereby enhancing pollutant removal. Removal rates of greater than 80 percent for total suspended solids were achieved in many studies (Schueler et al., 1992). Pollutant removal is primarily a function of the ratio of pond volume to watershed size (USEPA, 1986).

A low level of routine maintenance, including tasks such as mowing of side slopes, inspections, and clearing of debris from outlets, is required. Wet ponds can be expected to lose approximately 1 percent of their runoff storage capacity per year as a result of sediment accumulation. To maintain the pollutant removal capacity of the pond, periodic removal of sediment is necessary. A recommended sediment cleanout cycle is every 10 to 20 years (British Columbia Research Corp., 1991). With proper maintenance and replacement of inlet and outlet structures every 25 to 50 years, wet ponds should last in excess of 50 years (Schueler, 1987). A review of capital costs for wet ponds revealed costs of $349 to $823 per acre treated and annual maintenance costs of 3 percent to 5 percent of the capital cost (Schueler, 1987).

A complete discussion of created wetlands can be found in Chapter 7.

Infiltration practices suitable for storm water treatment include basins and trenches. Figures 5-4 and 5-5 show examples of infiltration basins and trenches. Like porous pavement, infiltration practices reduce runoff by increasing ground-water recharge. Prior to infiltration, runoff is stored temporarily at the surface, in the case of infiltration basins, or in subsurface stone-filled trenches.

Infiltration devices should drain within 72 hours of a storm event and should be dry at other times. The maximum contributing drainage area should not exceed 5 acres for an individual infiltration trench and should range from 2 to 15 acres for an infiltration basin (Schueler et al., 1992).

Pretreatment to remove coarse sediments and PAHs is necessary to prevent clogging and diminished infiltration capacity over time. The application of infiltration devices is severely restricted by soils, water table, slope, and contributing area conditions. The sediment load from marina hull maintenance areas may limit the applicability of infiltration devices in these areas. Infiltration devices are not practical in soils with field-verified infiltration rates of less than 1/2 inch per hour (Schueler et al., 1992). Soil borings should be taken well below the proposed bottom of the trench to identify any restricting layers and the depth of the water table. Removal of soluble pollutants in infiltration devices relies heavily on soil adsorption, and removal efficiencies are lowered in sandy soils with limited binding capacity. Schueler (1987) reported a sediment removal efficiency of 95 percent, 60 percent to 75 percent removal of nutrients, and 95 percent to 99 percent removal of metals using a 2-year design storm.

Infiltration basins and trenches have had high failure rates in the past (Schueler et al., 1992). A geotechnical investigation and design of a sound and redundant pretreatment system should be required before construction approval. Routine maintenance requirements include inspecting the basin after every major storm for the first few months after construction and annually thereafter to determine whether scouring or excessive sedimentation is reducing infiltration. Infiltration basins must be mowed twice annually to prevent woody growth. Tilling may be required in late summer to maintain infiltration capacities in marginal soils (Schueler, 1987). Field studies indicate that regular maintenance is not done on most infiltration trenches/basins, and 60 percent to 70 percent were found to require maintenance. Based on longevity studies, replacement or rehabilitation may be required every 10 years (Schueler et al., 1992). Proper maintenance of pretreatment structures may result in increased longevity. Reported costs for infiltration devices varied considerably based on runoff storage volume. Annual maintenance costs varied from 3 percent to 5 percent of capital cost for infiltration basins and from 5 percent to 10 percent for infiltration trenches.

Chemical treatment of wastewater is the addition of certain chemicals that causes small solid particles to adhere together to form larger particles that settle out or can be filtered. Filtration systems remove suspended solids by forcing the liquid through a medium, such as folded paper in a cartridge filter (METRO, 1992b). A recent study showed that such treatment systems can remove in excess of 90 percent of the suspended solids and 80 percent of most toxic metals associated with hull pressure-washing wastewater (METRO, 1992a). The degree of treatment necessary may be dependent on whether the effluent can be discharged to a sewage treatment system. The cost of a homemade system for a small boatyard to treat 100 gallons a day was estimated at $1,560. The cost of larger commercial systems capable of treating up to 10,000 gallons a day was estimated at $3000 to $50,000 plus site preparation. The solid waste generated by these treatment systems may be considered hazardous waste and may be subject to disposal restrictions.

A complete discussion of vegetated filter strips can be found in Chapter 7.

Grassed swales are low-gradient conveyance channels that may be used in marinas in place of buried storm drains. To effectively remove pollutants, the swales should have relatively low slope and adequate length and should be planted with erosion-resistant vegetation. Swales are not practical on very flat grades or steep slopes or in wet or poorly drained soils (SWRPC, 1991). Grassed swales can be applied in areas where maximum flow rates are not expected to exceed 1.5 feet per second (Horner et al., 1988). The main factors influencing removal efficiency are vegetation type, soil infiltration rate, flow depth, and flow travel time. Properly designed and functioning grassed swales provide pollutant removal through filtering by vegetation of particulate pollutants, biological uptake of nutrients, and infiltration of runoff. Schueler (1987) suggests the use of check dams in swales to slow the water velocity and provide a greater opportunity for settling and infiltration. Swales are designed to deal with concentrated flow under most conditions, resulting in low pollutant removal rates (SWRPC, 1991). Removal rates are most likely higher under low-flow conditions when sheet flow occurs. This may help to explain that the reported percent removal for TSS varied from 0 to greater than 90 percent (W-C, 1991). Wanielista and Yousef (1986) stated that swales are a useful component in a storm water management system and removal efficiencies can be improved by designing swales to infiltrate and retain runoff. Swales should be used only as part of a storm water management system and may be used with the other practices listed under this management measure.

Maintenance requirements for grassed swales include mowing and periodic sediment cleanout. Surveys by Horner et al. (1988) and in the Washington area indicate that the vast majority of swales operate as designed with relatively minor maintenance. The primary maintenance problem was the gradual build-up of soil and grass adjacent to roads, which prevents the entry of runoff into swales. The cost of a grassed swale will vary depending on the geometry of the swale (height and width) and the method of establishing the vegetation. Construction costs for grassed swales are typically less than those for curb-and-gutter systems. Regular maintenance costs for conventional swales are minimal. Cleanout of sediments trapped behind check dams and spot vegetation repair may be required (Schueler et al., 1992).

Porous pavement has a layer of porous top course covering an additional layer of gravel. A crushed stone-filled ground-water recharge bed is typically installed beneath these top layers. The runoff infiltrates through the porous asphalt layer and into the underground recharge bed. The runoff then exfiltrates out of the recharge bed into the underlying soils or into a perforated pipe system (see Figure 5-6). When operating properly, porous pavement can replicate predevelopment hydrology, increase ground-water recharge, and provide excellent pollutant removal (up to 80 percent of sediment, trace metals, and organic matter). The use of porous pavement is highly constrained and requires deep and permeable soils, restricted traffic, and suitable adjacent land uses. Pretreatment of runoff is necessary to remove coarse particulates and prevent clogging and diminished infiltration capacity.

The major advantages of porous pavement are (1) it may be used for parking areas and therefore does not use additional site space and (2) when operating properly, it provides high long-term removal of solids and other pollutants. However, significant problems exist in the use of porous pavement. Porous pavement sites have a high failure rate (75 percent) (Schueler et al., 1992). High sediment loads and oil result in clogging and eventual failure of the system. Therefore, porous pavement is not recommended for treatment of runoff from hull cleaning/ maintenance areas. Porous pavement is appropriate for low-intensity parking areas where restrictions on use (no heavy trucks) and maintenance (no deicing chemicals, sand, or improper resurfacing) can be enforced. Quarterly vacuum sweeping and/or jet hosing is needed to maintain porosity. Field data, however, indicate that this routine maintenance practice is not frequently followed (Schueler et al., 1992).

The cost of porous pavement should be measured as the incremental cost, or the cost beyond that required for conventional asphalt pavement (up to 50 percent more). To determine the full value of porous pavement, however, the savings from reducing land consumption and eliminating storm systems such as curbs, inlets, and pipes should be considered (Cahill Associates, 1991). Also, the additional cost of directing pervious area runoff around porous pavement should be considered. Maintenance of porous pavement consists of quarterly vacuum sweeping and may be 1 percent to 2 percent of the original construction costs (Schueler et al., 1992). Other maintenance costs include rehabilitation of clogged systems. In a Maryland study, 75 percent of the porous pavement systems surveyed had partially or totally clogged within 5 years. Failure was attributed to inadequate construction techniques, low permeable soils and/or restricting layers, heavy vehicular traffic, and resurfacing with nonporous pavement materials (Schueler et al., 1992).

Oil-grit separators (see Figure 5-7) may be used to treat water from small areas where other measures are infeasible and are applicable where activities contribute large loads of grease, oil, mud, sand, and trash to runoff (Steel and McGhee, 1979). Oil-grit separators are mainly suitable for oil droplets 150 microns in diameter or larger. Little is known regarding the oil droplet size in storm water; however, droplets less than 150 microns in diameter may be more representative of storm water (Romano, 1990). Basic design criteria include providing 200-400 cubic feet of oil storage per acre of area directed to the structure. The depth of the oil storage should be approximately 3-4 feet, and the depth of grit storage should be approximately 1.5-2.5 feet minimum under the oil storage. Application is limited to highly impervious catchments that are 2 acres or smaller.

Actual pollutant removal occurs only when the chambers are cleaned out. Re-suspension limits long-term removal efficiency if the structure is not cleaned out. Periodic inspections and maintenance of the structure should be done at least twice a year (Schueler, 1987). With proper maintenance, the oil/grit separator should have at least a 50-year life span.

Simply put, holding tanks act as underground detention basins that capture and hold storm water until it can receive treatment. There are generally two classes of tanks: first flush tanks and settling tanks (WPCF, 1989). First flush tanks are used when the time of concentration of the impervious area is 15 minutes or less. The contents of the tank are transported via pumpout or gravity to another location for treatment. Excess runoff is discharged via the upstream overflow outlet when the tank is filled. Settling tanks are used when a pronounced first flush is not expected. A settling tank is similar to a primary settling tank in that only treated flow is discharged. The load to the clarifier overflow is usually restricted to about 0.2 ft3/sec/ac of impervious area. If the inflow exceeds this, upstream overflows are activated. Settling tanks require periodic cleaning.

A swirl concentrator is a small, compact solids separation device with no moving parts. During wet weather the unit's outflow is throttled, causing the unit to fill and to self-induce a swirling vortex. Secondary flow currents rapidly separate first flush settleable grit and floatable matter (WPCF, 1989). The pollutant matter is concentrated for treatment, while the cleaner, treated flow discharges to receiving waters. Swirl concentrators are intended to operate under high-flow regimes and may be used in conjunction with settling tanks. EPA published a design manual for swirl and helical bend pollution control devices (USEPA, 1982). However, monitoring data reveal that swirls built in accordance with this manual should be operated at lesser flows than the design indicates to achieve the desired efficiency (Pisano, 1989). Total suspended solids and BOD concentration removal efficiencies in excess of 60 percent have been reported, particularly under first flush conditions (WPCF, 1989). In another report removal effectiveness of total suspended solids from current U.S. swirls varied from a low of 5.2 percent to a high of 36.7 percent excluding first flush, 32.6 percent to 80.6 percent for first flush only, and 16.4 percent to 33.1 percent for entire storm events (Pisano, 1989). Removal efficiencies are dependent on the initial concentrations of pollutants, flow rate, size of structure, when the sumps in the catchments were cleaned, and other parameters (WPCF, 1989; and Pisano, 1989).

Catch basins with flow restrictors may be used to prevent large pulses of storm water from entering surface waters at one time. They provide some settling capacity because the bottom of the structure is typically lowered 2 to 4 feet below the outlet pipe. Above- and below-ground storage is used to hold runoff until the receiving pipe can handle the flow. Temporary surface ponding may be used to induce infiltration and reduce direct discharge. Overland flow can be induced from sensitive areas to either sink discharge points or other storage locations. Catch basins with flow restrictors are not very effective at pollutant removal by themselves (WPCF, 1989) and should be used in conjunction with other practices. Removal efficiencies for larger particles and debris are high and make catch basins attractive as pretreatment systems for other practices. The traps of catch basins require periodic cleaning and maintenance. Cleaning catch basins can result in large pulses of pollutants in the first subsequent storm if the method of cleaning results in the disturbance and breaking up of residual matter and some material is left in the catch basin (Richards et al., 1981). With proper maintenance, a catch basin should have at least a 50-year life span (Schueler et al., 1992).

A catch basin with sand filter consists of a sedimentation chamber and a chamber filled with sand. The sedimentation chamber removes coarse particles, helps to prevent clogging of the filter medium, and provides sheet flow into the filtration chamber. The sand chamber filters smaller-sized pollutants. Catch basins with sand filters are effective in highly impervious areas, where other practices have limited usefulness. The effectiveness of the sediment chamber for removal of the different particles depends on the particles' settling velocity and the chamber's length and depth. The effectiveness of the filtration medium depends on its depth.

Catch basins with sand filters should be inspected at least annually, and periodically the top layer of sand with deposition of sediment should be removed and replaced. In addition, the accumulated sediment in the sediment chamber should be removed periodically (Shaver, 1991). With proper maintenance and replacement of the sand, a catch basin with sand filter should have at least a 50-year life span (Schueler et al., 1992).

While there is some tendency for oil and grease to sorb to trapped particles, oil and grease will not ordinarily be captured by catch basins, holding tanks, or swirl concentrators. Adsorbent material placed in these structures in a manner that will allow sufficient contact between the adsorbent and the storm water will remove much of the oil and grease load of runoff (Silverman and Stenstrom, 1989). In addition, the performance of oil-grit separators could be enhanced through the use of adsorbents. An adsorbent/catch basin system that treats the majority of the grease and oil in storm water runoff could be designed, and annual replacement of the adsorbent would be sufficient to maintain the system in most cases (Silverman et al., 1989). Manufacturers report that their products are able to sorb 10 to 25 times their weight in oil (Industrial Products, 1991; Lab Safety, 1991). The cost of 10 pillows, 24 inches by 14 inches by 5 inches (total weight 24 pounds), is approximately $85 to $93 (Lab Safety, 1991).