Dealing With the Deluge
The effect of phosphorus on the environment has been a hot topic of conversation among environmentalists for some time. The Clean Water Act (1970) established a two-phase approach for requiring communities -- more than 100,000 inhabitants being phase one and less than 100,000 being phase two -- to regulate stormwater runoff, including the amount of phosphate runoff. According to the U.S. Environmental Protection Agency (EPA), stormwater runoff from agricultural lands and urban areas is the primary source of pollutants in our nation's waterways. Although the main source of runoff has been identified, few counties have settled on what constitutes safe levels of phosphate and other chemicals for various rivers, lakes and ponds.
We are looking at what the chemicals do to the environment and are witnessing the depletion of the oxygen levels in the rivers, lakes and ponds by phosphate. Studies recorded by Dunne and Leopold (1978) illustrate the human effects of phosphorus on plants. According to this study, natural levels of orthophosphate usually range from 0.005 to 0.05 milligrams per liter (mg/l), but outside sources (i.e., stormwater) promote the rampant growth of algae by supplying additional phosphorus above background levels. While plants need phosphorus to grow, an abundance of phosphorus promotes excessive algae growth, resulting in depleted, dissolved oxygen levels within the water body during the algae's nocturnal cycle. The depleted oxygen, in turn, is detrimental to the freshwater fish.
Non-point source runoff (stormwater) delivers phosphorus to lakes and streams by dislodging sediment and organic matter from surfaces (pervious and impervious) and eventually delivering the phosphorus to the receiving water body.
Stormwater runoff combined with excessive use of fertilizers, decaying yard debris and animal waste delivers levels of dissolved phosphorus much higher than the natural systems need to sustain their biological cycle.
Test the waters
While treatment ponds can be a solution for biologically treating phosphorus found in stormwater runoff, the natural processing of the nutrient can take as many as 30 days depending on plant and soil conditions.
Stormwater Management Inc., Portland, Ore., began testing by using iron to absolve the hazardous levels of phosphates. Iron, an abundant resource, carries a +3 charge that associates with the -3 charge of the dissolved phosphorus. Other natural resources, such as tephra and surface modified zeolite were considered for removal of the phosphate, but the results were less effective than iron.
At the start of testing, iron-infused resin was evaluated for the removal of dissolved phosphorus from stormwater through isotherm sorption studies and using both a horizontal flow column and a full-scale filtration cartridge. The iron-infused media was screened to produce a filtration media with a grain size between two and 12.5 millimeters (mm) (0.08 and 0.5 inches).
Stormwater Management set a benchmark with the sorption of phosphate on the iron-infused media. Sorption studies produced a maximum uptake of dissolved phosphorus at 840 mg-P/l over 24 hours. Using small portions of media and subjecting it to various concentrations of dissolved phosphorus (ranging from 0.5 mg-P/l to 100 mg-P/l) determined the sorption maximum. Two stock solutions of dissolved phosphorus were created using sodium phosphate (NaH_2PO_4) and deionized water. The stock solutions had a concentration of 50 mg-P/l and 100 mg-P/l. Approximately 2.5 grams of iron-infused media were weighed for each 100 milliliters (ml) reaction vessel. By diluting the stock solutions with deionized water, 200 ml of each concentration was created. Half of each dilution was added to the reaction vessel with the remaining 100 ml being analyzed for the initial phosphorus concentration. The reaction vessels were turned, end over end, every hour for the first eight hours of the reaction. The reaction was left overnight (1
2 hours) and then turned every hour the remaining four hours of the reaction (24 hour total reaction time). The supernatant was then taken and analyzed for residual phosphorus remaining in the water column.
Other testing was available on a smaller scale. For example the horizontal flow column tests showed dissolved phosphorus removal stabilized at approximately 20 percent mass removal during the testing.
Full scale cartridge testing involved the determination of the media's effectiveness when used under flowing conditions within a cartridge. Testing was performed by filling a cartridge with an internal ring of iron-infused media and an external ring of perlite. The iron-infused media had a thickness of 3.5 inches and a total mass of 9,768 grams (volume = 27.6 liters). Perlite surrounded this media with a thickness of 3.5 inches and a mass of 5,126 grams (volume = 41.9 liters).
The cartridge was assembled and placed inside an acrylic tank. The tank was fitted with an under drain manifold where the cartridge was attached, and two separate reservoirs were used for influent and effluent samples.
The influent reservoir was filled to 110 gallons and received sediment and sodium phosphate from a 1-liter stock solution. (The stock solutions were created such that a 1-liter addition to 110 gallons would create concentrations of Total Suspended Solids (TSS) and phosphorus normally observed in the field. TSS = 100 mg/l, 0.3 mg-P/l)
The influent reservoir contained a sump pump that delivered the phosphorus containing water to a flume via flexible hosing. A flume is an artificial trough that carries water. The flume allowed the water to cascade across an energy dissipater and directly into the acrylic tank containing the cartridge.
Finally, samples were taken from the influent reservoir prior to initializing the pump. These were analyzed for TSS, total-P (phosphorus), which contains both dissolved and undissolved phosphorus, dissolved-P, pH and occasionally biochemical oxygen demand (BOD_5) -- the oxygen demand on microorganisms to breakdown contaminants. The water was then pumped into the cartridge tank to mimic a storm such that water begins to flow, reaches a flow peak and then tapers. The filling process peaked after approximately seven minutes with a flow rate of 15 gallons per minute (gpm). The entire influent reservoir was passed through the cartridge and discharged into an effluent reservoir (approximately 100 gallons due to water that cannot be removed by the pump). Samples of the effluent were taken to match those of the influent as well as treated water volume measurements to determine mass removal.
Conclusion
The ability of iron-infused media to remove dissolved-P (Phosphorus) from stormwater has been determined to be very effective. The sorption maximum for dissolved-P of 840 mg-P/kg-iron shows the high affinity of the media. The flow rates used during the test are also high (15 gpm) and only allow approximately 30-seconds of contact time with the iron-infused media. Removal rates associated with this flow rate tend to level at approximately 18.6 percent dissolved-P reduction.
Using the isotherm data and a saturation point of 840 mg-P/kg iron, estimates can be made as to the amount of dissolved-P potentially removed by a single cartridge. Since 9.77 kg of iron-infused media is contained within each cartridge (surrounded by perlite), 8,205 mg of dissolved-P can be removed. Although phosphorus concentrations vary considerably with each storm event, a single cartridge could effectively treat 82,500 liters (21,794 gallons, 0.80 acre-inches/cartridge) of stormwater at a dissolved-P concentration of 0.1 milligrams of phosphorus per liter (mg-P/l)(This calculation assumes a 100 percent removal). If the removal of 18.6 percent dissolved-P is used for this calculation, approximately 412,500 liters (108,970 gallons, 4.01acre-inches/cartridge) could be treated. Another consideration is sediment loading on the cartridge -- the primary indication for filter maintenance. Using the same scenario of 82,500 liters of treatment and assuming a TSS concentration of 100 milligrams of total suspended solids per liter (mg-TSS/l), the sediment loading on the cartridge would be approximately 18 pounds. Previous testing by Stormwater Management has shown that the cartridges start experiencing a decrease in flow rate (from 15 gpm) at approximately 16 pounds.
Although the iron-infused media has proven to be very effective for dissolved-P removal in the laboratory, research is underway to determine the effectiveness of the media in the field. Current applications of the media include a 13-home, residential neighborhood with a filtration system containing nine iron-infused and perlite cartridges.
The iron-infused media also has potential for structural changes to the resin. These changes include the addition of more iron particles, the addition of finer iron particles (to increase surface area) and a further expansion of the resin to form a thinner cell wall to expose more iron at the surface of the media. Mixing the iron media with perlite is also an option to effectively increase the surface area by spreading the iron throughout the cartridge rather than only using an internal ring surrounded by perlite.
Bryan Wigginton is a research chemist at Stormwater Management Inc., Portland, Ore. He can be reached via e-mail at bryanw@stormwatermgt.com.
This article originally appeared in the 11/01/2000 issue of Environmental Protection.