September 19, 2007
A Storm Water Runoff Model For Open Windrow Composting Sites
By Kalaba, Ljubisa Wilson, Bruce G; Haralampides, Katy
Precipitation that falls on compost sites picks up organic material from the windrows and the composting pad. The resulting runoff can contain high levels of nutrients, suspended solids, and organic matter, making it unsuitable for direct release into a receiving water body. Many jurisdictions require that the runoff from these sites be collected in a detention pond. Unfortunately, some of the recommended or required procedures for quantifying the volume of runoff from these sites are based on archaic or inappropriate hydrologic models. The development of better hydrologic models for open composting operations has been hampered by a lack of basic information regarding rainfall/runoff relationships at windrow composting sites. In this paper, a standard hydrologic model - the unit hydrograph method - is used to model the hydrology of a small, paved composting site. The model results compare well with field data collected at the site over a six month period. The volume of runoff predicted by the model was within 5% of the measured runoff volume for each of seventeen runoff events observed at the site over the study period. The results suggest that other industry standard hydrologic models can be adapted for use at open composting sites to account for the presence of large quantities of organic material on the site. Introduction
Most open windrow composting operations are situated on a prepared surface of asphalt, concrete or gravel that is graded slightly to encourage good drainage of the composting pad. It is usually recommended that the site be set on a grade of between 1 and 2% and that the windrows are placed parallel to the direction of the slope of the composting pad to ensure that storm water drains into a detention pond (NRAES 1992; AMRC 1994; USEPA 1994). Most sites are also designed to ensure that only water from the compost pad enters the pond and that rainfall that does not land on the pad is directed away from the pond. The direct result of these design measures is that compost pads often resemble large parking lots connected directly to a storm water detention pond.
These design features are intended to ensure that precipitation drains from the composting pad quickly. However, even when good water management practices are followed, rain falling on the site can wash small amounts of compost from the sides of windrows and from the aisles between windrows. Precipitation can also leach though the windrows and drain from the toe of the pile, as illustrated in Figure 1, which shows water leaching from a compost pile approximately 24 hours after a heavy rain. In laboratory experiments, Wilson et al. (2004) found that water sprayed on compost windrows leached out over a period of 48 hours or more.
There is considerable evidence that the storm water runoff from these sites is usually high in suspended solids, biochemical oxygen demand (BOD), chemical oxygen demand (COD) and nutrients and cannot be released to surface water or groundwater without prior treatment (Richard and Chadsey 1990; Cole 1994; Krogmann and Woyczechowski 2000; Peigne and Girardin 2004). In response, most regulatory agencies require that runoff from open composting facilities be treated prior to discharge. In some cases this can be accomplished by discharging the runoff to a sanitary sewer, however, since many composting facilities are located in remote areas, the most common approach is to collect runoff in a storm water management pond for treatment prior to release (USEPA 1994).
FIGURE 1. Water leaching from a compost windrow onto an asphalt pad.
Although the quality of runoff from composting sites has been well documented (USEPA 1994), there is very little information available regarding the quantity of runoff generated at open windrow sites. Krogmann and Woyczechowski (2000) reported that leachate was generated at a rate of 25 L/Mg of compost in a covered facility in Germany. They also reported that approximately 15 to 20% of rainwater reaching an uncovered compost pile could be released as surface runoff. Marques and Hogland (2001) estimated volumetric runoff coefficients and hydrologic abstractions for several different activity areas at a waste management park in Sweden and reported a runoff coefficient of 0.44 for the gravel based composting pad at the site. However, this coefficient was based only on the characteristics of the gravel base and did not account for the compost on the pad. Wilson, et al. (2004) reported that approximately 68% of the rainfall volume falling on a saturated compost windrow located on an asphalt pad would eventual end up as runoff.
Despite a lack of scientific information regarding rainfall/ runoff relationships at composting sites, specific regulatory requirements exist in many jurisdictions and many others offer specific guidance regarding the size of storm water detention ponds. For example, New York State requires that site operators provide "calculations of surface water runoff that must be handled at the site, based on a rainfall intensity of one hour duration and a 10 year return period" (NYCRR 1999). The State of Georgia requires storm water detention ponds be sized to handle the expected runoff from a 24 hour rainfall volume with a 25 year return period (Governo 2001). A guideline from the British Columbia Ministry of Agriculture and Food recommends sizing a pond to hold the total volume of rainfall from the worst 6 months in a 25 year period (BCMAFF 1996).
Unfortunately, there is no agreement in the literature about what constitutes an appropriate design event for a compost facility, leading to considerable uncertainty regarding the size of the detention pond required. For example, application of the New York regulations to a 54 m x 32 m asphalt paved composting pad in Fredericton, New Brunswick (assuming C = 0.8, t^sub c^ = 1 hr, and i = 26 mm/hr) results in a required detention pond volume of 36 m^sup 3^. Application of the Georgia requirements to the same pad (assuming C = 0.8, t^sub c^ = 24 hr, and i = 5 mm/hr) results in a required pond volume of 166 m^sup 3^, while the B.C. criteria would require a pond with a volume of more than 1000 m^sup 3^. Clearly, each of these guidelines implies a different level of protection from overtopping of the pond and a proportionate capital cost.
In addition to a lack of consensus regarding an appropriate design event, there are several other problems with applying the Rational Method to the design of storm water detention ponds for compost facilities, including the following:
1. The time of concentration for the paved portions of a compost site is likely to be measured in minutes, not hours. Conversely, the time of concentration of precipitation passing through a compost windrow is likely to be measured in days;
2. Depending on the shape of the windrow and on antecedent conditions, compost windrows can shed precipitation, absorb precipitation, or act as a reservoir and detain precipitation. In addition, the water balance in the compost changes with time as water and carbon dioxide are generated during degradation of organic matter (Haug 1980). The Rational Method does not account for these processes;
3. Use of a bulk runoff coefficient may not be appropriate for compost. Kalaba (2006) showed that the ratio of runoff to precipitation from compost windrows varied with rainfall intensity;
4. The Rational Method is intended to be used to estimate runoff from a single precipitation event. It cannot be used to evaluate the performance of a storm water treatment pond over an extended period of wet weather, which is precisely when most composting sites need to collect and store significant quantities of leachate and runoff;
5. Specification of the runoff volume only addresses half of the problem. Unless runoff is treated and released from the storage pond, it will continue to accumulate until the pond is overtopped. Therefore, it is necessary to model the rate at which runoff will be treated and released from the site. The Rational Method needs to be modified if the intent is to model both inflows and outflows (Mays 2005); and
6. It is impossible to use the Rational Method to quantify the risk that a detention pond will be overtopped. The Rational Method approach suggests that the calculated runoff volume has a return period equal to the return period of the rainfall intensity, however, this is not generally true. Only on rare occasions will the product of rainfall intensity and duration produce a volume with the same return period as the IDF curve (Adams and Howard 1986; Adams and Papa 2000).
For these reasons, we conclude that use of the Rational Method for sizing storm water detention facilities at open windrow composting sites is not a technically sound approach. It is likely that much better results could be obtained by modeling these sites using more rigorous hydrologic models such as the Unit Hydrograph Method (Mays 2005), continuous simulation models such as the USEPA storm water Management Model (SWMM) (Rossman 2005) or derived probability models (Adams and Papa 2000).
In this paper, our goal is to report on the development of an effective storm water model for use at open composting sites, with the specific objectives to: a) develop a conceptual water balance for open windrow composting sites; b) collect field data on the quantities of precipitation and runoff at an open windrow composting facility; and c) model the hydrology of the site using the unit hydrograph technique. In the latter part of this paper, we describe the development of a conceptual water balance for an open windrow composting site, provide a summary of field work conducted to collect precipitation and runoff data at a composting site, and report the results of modeling the field data using the unit hydrograph method. Materials and Methods
Development of a Conceptual Water Balance For Open Composting Sites
A hydrologic system is defined as a structure or volume in space, surrounded by a boundary that accepts water and others inputs, operates on them internally, and produces them as outputs (Chow et al. 1988). This paper will treat outdoor composting sites as a hydrologic system. The basic components of the system include precipitation, evaporation, infiltration, surface runoff and ground water flow. Storm water runoff from a composting site depends on the volume of rainfall incident on the composting pad, the quality and quantity of the composting material, specific site management practices, and the amount of water generated during the composting process. If we think of the site as a hydrologic system, it is possible to develop a water balance for such as site, as shown schematically in Figure 2.
FIGURE 2. Conceptual Water Balance for a Windrow Composting Site
The figure illustrates that there are generally three sources of water at open windrow composting sites: a) moisture in the incoming waste material and bulking agents, b) precipitation, and c) water generated as a result of the biodegradation of organic materials. Occasionally, during dry periods or in warmer climates, site operators will add supplemental process water to compost that has dried out. As shown in the figure, supplemental water can be drawn from the detention pond or it may be imported onto the site.
Water can be lost from the system in several ways (Figure 2). The major water output components are: a) direct release or removal of water from the detention pond, b) infiltration through the composting pad, c) infiltration through the bottom of the detention pond, d) evaporation from the composting piles, and e) evaporation from the runoff collection pond. Water is also removed from the site in the form of moisture in the finished compost. The windrow itself can also act as a reservoir, detaining precipitation for several hours, before releasing it as leachate from the toe of the compost pile.
Description of the Composting Site
Field measurements of rainfall and runoff volumes were collected at a small experimental compost site located at the Agriculture Canada Potato Research Station in Fredericton, New Brunswick. The facility composts small amounts of a wide range of materials including vegetative wastes, potato culls, animal manure, poultry manure, wood chips and chipped brush, leaves, sawdust and paper mill sludge. The composting pad is also used to cure and store finished compost.
Composting occurs on a 54 m by 32 m compost pad paved with 75 mm of asphalt. The site is sloped slightly towards a catchbasin located at the edge of the composting pad. The catchbasin is connected to a dry-well, which acts as a covered detention pond, by a short section of 300 mm diameter concrete pipe. The drywell was excavated from native soil, loosely filled with 20 mm crushed rock, and then covered with 0.5 m of clay soil. Water in the drywell infiltrates through the bottom and sides of the drywell into the surrounding groundwater table. The total volume of the drywell is 80.4 m^sup 3^, however, because of the presence of the crushed rock, it is estimated that the drywell can hold a maximum volume of 16.4 m^sup 3^ of runoff. At the beginning of the field work, the catchbasin was pumped out and a good hydraulic connection between the catchbasin and the drywell was observed.
At the beginning of the field work, in May, 2004, a total volume of 250 m of organic material was on the site. Most of this material was in a well mixed windrow of composted vegetative waste which was nearing the curing stage. In total, approximately 40% of available surface area of the composting pad was covered with organic material. Moisture content samples were taken from all of the different materials on site at the beginning and at the end of the field work. No new material was received at the site until October 25, 2004 when approximately 5 tonnes of potato culls and 15 straw bundles were added to the site. With the addition of this material, approximately 60% of available asphalt surface of the composting pad was covered with organic material.
Meteorological data were collected using a HOBO Model H21-001 remote weather station equipped with an S-RGB-M006 Tipping bucket rain gauge (with an accuracy of +-1.0% at up to 20 mm/hour) and an S- THA-M006 Temperature/Relative Humidity probe (with an accuracy of +- 0.7 [degrees]C at 25 [degrees]C). The weather station was located directly adjacent to the study site. Water levels were recorded using a Solinst Model 3001 Levelogger installed in the catchbasin on the site, below the level of water in the drywell. The water level recorder also recorded the temperature of the water in the catchbasin. Water levels were compensated for barometric pressure using a Solinst Model 3001 Barologger suspended above the water, but still inside the catchbasin.
The remote weather station was on site from May 30, 2004 to November 15, 2004. Data collection ended in November because it was expected that precipitation at the site would soon turn to snow, rendering measurements from the tipping bucket gauge inaccurate. Measurements of rainfall, temperature, barometric pressure and relative humidity were made at ten minute intervals for the period May 30 to October 24, 2004 and at 15 minute intervals for the period October 25 to November 15, 2004.
Data from the weather station were verified using data from a permanent Agriculture Canada weather station located approximately 1 km from the compost pad. In addition to rainfall, temperature, barometric pressure and relative humidity measurements, this facility also collected pan evaporation rates for the period May 2004 to October 2004.
Modeling the Hydrology of the Site
Using this conceptual water balance as a model, the authors sought to collect precipitation and runoff data from an operating compost site and model it using unit hydrograph methods.
A time series of precipitation volumes (mm) as measured by the tipping bucket rain gauge and the corresponding water level in the catchbasin (m) above an arbitrary datum are shown in Figure 3. (The figure shows only a portion of the field data when several rainfall events occurred in succession over a period of 45 days. Similar results were obtained throughout the study period; however, rainfall events were less frequent during other time periods and a time series plot of the entire data set makes it very difficult to illustrate the response of the system to an individual rainfall event.) The figure demonstrates that the water level in the catchbasin peaks shortly after the end of the rainfall event and then recedes slowly as the runoff is conveyed to the drywell and ultimately infiltrates into the soil. During some events, the catchbasin filled to capacity and runoff backed up onto the compost pad. However, this runoff was contained on the pad and eventually drained through the catchbasin to the drywell as the water levels receded. Water levels in the catchbasin were assumed to represent the water level in the drywell because there was a very good hydraulic connection between the two.
FIGURE 3. Precipitation (mm) and water level in the catchbasin above datum (m) from June 30 to August 12, 2004
FIGURE 4. Precipitation and water temperature from June 30 to August 12, 2004
For analysis, the data were used to identify a number of distinct rainfall /runoff events. An event was defined to begin at the time of the first measurable rise in the pond level and lasted until the water level in the catchbasin returned to its pre-event level. In total, seventeen distinct rainfall /runoff events were identified from the data. Some events consisted of a short, intense rainfall which resulted in a sharp rise in the water level in the drywell followed by a gradual decline in the pond level over the next several hours. Other events consisted of prolonged periods of rain of varying intensity lasting several hours. For the events analyzed, rainfall volumes ranged from 0.75 mm to 29.23 mm, average rainfall intensities ranged from 0.67 mm/hr to 7.31 mm/hr, and rainfall durations ranged from 45 minutes to 520 minutes.
The Levelogger in the catchbasin also recorded water temperatures at 15 minute intervals. This data is presented as a time series in Figure 4. The data showed that the temperature of the water in the catchbasin increased substantially after some rainfall events. During one event (03 August 2004), the temperature of the runoff from the pad was more than 7 [degrees]C warmer than the water in the drywell. This rise in temperature was to be expected since precipitation incident on the asphalt composting pad became heated as it ran over the pad. Release of very warm runoff to a receiving water could result in direct biological impacts or changes in water quality in that receiving water (VanBuren et al. 2000), providing yet another reason for controlling runoff from composting facilities.
Evaporation rates were not measured directly at the site, however pan evaporation rates were measured at the nearby Agriculture Canada weather station. That data showed that the average pan evaporation rate was 5.5 mm per day, with a maximum reported rate of 18.7 mm per day. Evaporation rates were generally very low during the 17 precipitation events analyzed. As a result, evaporation was ignored in the event analysis below. This does not mean that evaporation is not an important hydrologic process at open composting sites. In fact, much of the moisture lost from the site was lost through evaporation between precipitation events and should be accounted for in long term analysis of the hydrology of these sites. However, during individual events, evaporation rates were low enough to be ignored since they did not affect the volume of runoff generated during the event. Prediction of Runoff Volume
The purpose of the modeling phase of this research was to predict the volume of runoff from the study site using the precipitation data as an input. This predicted volume was then compared to the observed runoff volume to validate the model. A synthetic unit hydrograph approach was selected as an appropriate model for this work because the unit hydrograph method is well known and documented, it produces good results for small catchments, and it is specifically designed to generate the hydrograph from any volume of excess rainfall. A unit hydrograph is the hydrograph produced by a unit depth of runoff (usually 1 inch or 1 cm) distributed uniformly over a basin for a defined period of time (Mays 2005).
Equation (4) requires an estimate of the runoff coefficient for the site. Because of significant differences between asphalt and compost material characteristics it is necessary to calculate a weighted average runoff coefficient for the entire site. Runoff coefficients for asphalt are reported in many textbooks and generally range between 0.70 and 0.95 (Mays 2005; Viessman and Hammer 2005). Unfortunately, as noted above there are very few references in the literature to runoff coefficients for compost. Wilson et al. (2004) report a runoff coefficient of C = 0.68 for saturated compost based on laboratory experiments. This likely represents an upper bound on true field values. Drier compost, subject to summer temperatures and lighter rain intensity would likely result in a lower measured runoff coefficient. Therefore, a runoff coefficient for compost material in the range between 0.50 and 0.70 is likely more representative of the conditions at the site.
The volume of organic material on the site was relatively constant for two separate time periods during data collection. From May to October 24, 2004 (period 1), approximately 40% of the compost pad was covered with organic material. From October 25 to November 2004 (period 2), approximately 60% of the pad was covered with organic material. Therefore, two different weighted average runoff coefficients were calculated for the site. For period 1, a value of C^sub 1^ = 0.75 was used (based on C^sub asphalt^ = 0.85 and C^sub compost^ = 0.6) while a runoff coefficient or C^sub 2^ = 0.65 was used for period 2 (based on C^sub asphalt^ = 0.85 and C^sub compost^ = 0.5).
Using these weighted average runoff coefficients in equation (4), the time of concentration for the field site was estimated to be approximately 8 minutes for period 1 and approximately 11 minutes for period 2. Sensitivity analysis determined that the unit hydrograph method is very sensitive to the time of concentration, especially for such a small catchment. The two times of concentration were different enough to require that different unit hydrographs be developed for each period.
Using equation (5), it was determined that a 1-minute synthetic unit hydrograph, with a time to peak of 5 minutes would be appropriate for this site during period 1. This unit hydrograph is shown in Figure 5. Given the longer time of concentration in period 2, a 2-minute synthetic unit hydrograph, with a time to peak of 10 minutes was used for that time period. This unit hydrograph is also presented in Figure 5.
FIGURE 5. Synthetic Unit Hydrographs
Using the direct runoff hydrograph and an outflow hydrograph based on measured rates of infiltration from the drywell into the groundwater table, the variation in the water level in the drywell with time could be estimated using equation (3) (Kalaba 2006).
The result of this analysis for one event (on 28-29 July 2004) is shown in Figure 6. This event consisted of 3.5 mm of rainfall over 2.5 hours. The unit hydrograph technique was able to predict the water level measured in the catchbasin to within +-6.6 cm. More importantly, the volume of runoff estimated using the unit hydrograph (4.56 m^sup 3^) compares very well with estimates of the volume of water which infiltrated from the drywell (4.45 m^sup 3^). This difference of less than 2.5% suggests that the unit hydrograph method provides a reasonable estimate of the runoff for this rainfall event.
FIGURE 6. Predicted and observed water levels in the catchbasin for July 28-29, 2004.
FIGURE 8: Predicted and measured runoff volumes for 17 runoff events
Figure 7 shows the results of a more complicated rainfall event which occurred on 7-8 August, 2004. For this event, 10.58 mm of rain fell on the site. Most of the precipitation occurred over a 9 hour period, although there were some small amounts later in the day. In this case, the unit hydrograph technique only predicted the water level in the pond to within +-30 cm. However, the volume of runoff estimated using the unit hydrograph (12.23 m^sup 3^) compares very well with the volume of water lost through infiltration (12.40 m^sup 3^), a difference of less than 1.3%.
FIGURE 7. Predicted and observed water levels in the catchbasin for August 7-8, 2004.
Similar results were found for each of the other 15 runoff events. In all cases, the difference between the runoff volume predicted by the unit hydrograph method and the volume of water infiltrated from the drywell was less than 5%. Figure 8 shows that the unit hydrograph method accurately predicted total runoff volumes over a range of rainfall/runoff events. Errors in prediction of the exact water level in the pond are most likely due to incomplete knowledge of how the volume of the drywell varies with depth due to the presence of crushed rock fill. Other possible sources of error include neglecting evaporation during events and small changes in the moisture content of the compost piles after events.
The unit hydrograph method was used to model the water level in a storm water detention pond located adjacent to a paved composting pad. Seventeen separate rainfall/runoff events were modeled and in all cases, the runoff model results were in good agreement with field measurements. The unit hydrograph technique is a standard hydrologic technique and good agreement between this model and the field data suggests that other standard hydrologic techniques can also be used to model rainfall/runoffrelationships at open composting sites. In particular, continuous simulation models, such as the USEPA Storm Water Management Model (SWMM) (Rossman 2005) should produce good results. Preliminary research into the use of SWMM for this purpose is promising (Kalaba and Wilson 2006).
When applying hydrologic models to composting sites, care must be taken in the calculation of basin response time. The model employed in this study was very sensitive to the time of concentration. However, an empirical formula for the time of concentration developed by the Federal Aviation Agency for use with airport runways provided good results. Both this formula and observations at the site suggest that the time of concentration varies depending on the amount of compost material on the site.
The presence of large amounts of organic material on a composting site complicates the hydrology of the site. As a result, a weighted runoff coefficient for the entire compost pad should be used. In this study, a weighted runoff coefficient calculated using an assumed constant runoff coefficient for bare asphalt of 0.85 and a runoff coefficient for compost material in the range between 0.5 and 0.6 provided good results. However, there are very few references in the literature for runoff coefficients for compost material. In addition, runoff coefficients would likely be different for different compost feedstocks at different stages of the composting process. Additional field research in this area is needed.
Water temperature data from the site showed that the temperature of runoff from the composting pad can be affected by the temperature of the asphalt pad. Rainfall landing on hot asphalt surfaces can be heated significantly and for this reason (as well as those associated with the quality of the runoff) runoff should not be directly released into receiving waters. Additional research into temperature changes in runoff from composting pads is warranted.
This study focused on the analysis of a number of individual rainfall /runoff events and evaporation from the site during those events was considered negligible. However, evaporation from composting sites is a significant process between rainfall events. Continuous simulation of the hydrology of composting sites will need to address evaporation rates. Evaporation from windrow piles depends on many parameters including: ambient temperature, wind speed, humidity, temperature and moisture content of the compost mass, and turning frequency of compost mass. Field data on evaporation from compost piles and the development of predictive models of evaporation from compost piles are both needed.
Finally, since this project has demonstrated that the hydrology of composting sites can be modeled effectively with unit hydrographs, and since the Rational Method is not an appropriate method for sizing or modeling the performance of storm water detention ponds, the authors suggest that use of the Rational Method for estimating runoff volumes from composting sites be discontinued.
The authors wish to thank Sherif Fahmy and Lien Chow from Agriculture Canada for their assistance and cooperation on this project. Financial support from the Natural Sciences and Engineering Research Council of Canada and the University of New Brunswick is gratefully acknowledged. References
Adams, B.J. and F. Papa. 2000. Urban Storm Water management planning with analytical probabilistic models, J. Wiley, New York.
Adams, B.J. and C.D.D. Howard. 1986. Design storm pathology. Canadian Water Resources, 11(3):49-55.
AMRC. 1994. Leaf and yard material composting manual, Association of Municipal Recycling Coordinators, Guelph, Ontario.
BCMAFF. 1996. Composting factsheet, British Columbia Ministry of Agriculture, Food and Fisheries, Order Number 382.500-6, [accessed online at www.agf.gov.bc.ca/resmgmt/publist/300series/382500-6.pdf]
Chow, V.T., D.R. Maidment, and L.W. Mays. 1988. Applied hydrology. McGraw-Hill, New York, New York.
Cole, M.A. 1994. Assessing the impact of composting yard trimmings. BioCycle 35(4):92-96.
Federal Aviation Administration. 1970. Advisory circular: Airport drainage. AC No. 150/5320-5B. Federal Aviation Administration, Department of Transportation, July 1, 1970, Washington, DC.
Governo, J. 2001. Modeling a compost facility. BioCycle 42(8):55- 58.
Haug, R.T. 1980. Compost engineering: Principles and practice. Technomic Pub. Co., Lancaster, Pennsylvania.
Kalaba, L. 2006. Modelling Storm Water Runoff from Open Windrow Composting Sites. M.Sc.E. Thesis, Department of Civil Engineering, University of New Brunswick, Fredericton, New Brunswick, Canada.
Kalaba, L. and B.G. Wilson. 2006. Modelling compost site hydrology using USEPA SWMM, Proc. of 21st International Conference on Solid Waste Technology and Management, Philadelphia, PA U.S.A., March 26 - 29, 2006, pages 77-86.
Krogmann, U. and H. Woyczechowski. 2000. Selected characteristics of leachate, condensate and runoff released during composting of biogenic waste. Waste Management & Research 18(3): 235-248.
Mays, L.W. 2005. Water resources engineering. John Wiley & Sons, Hoboken, New Jersey.
Marques, M. and W. Hogland. 2001. Storm Water run-off and pollutant transport related to the activities carried out in a modern waste management park. Waste Management & Research 19(1): 20- 34.
NRAES. 1992. On-farm composting handbook. NRAES-54, Northeast Regional Agricultural Engineering Service, Ithica, New York.
NYCRR. 1999. Environmental conservation rules and regulations. Title 6 of the official compilation of codes, rules and regulations of the State of New York, 6 NYCRR, Part 360-5, Section 360-5.2 (b)(6)(xii).
Peigne, J. and P. Girardin. 2004. Environmental impacts of farm- scale composting practices. Water, Air, and Soil Pollution, 153(1): 45-68.
Richard, T. and M. Chadsey. 1990. Environmental impact of yard waste composting. BioCycle 31(4): 42-46.
Rossman, L.A. 2005. Storm water management model user's manual - Version 5.0. Report EPA/600/R-05/040, U.S. Environmental Protection Agency, Cincinnati, Ohio.
USEPA. 1994. Composting yard trimmings and municipal solid waste. Report No. EPA530-R-94-003. United States Environmental Protection Agency, Office of Solid Waste and Emergency Response.
Van Buren, M. A., W. E. Watt, J. Marsalek, and B. C. Anderson. 2000. Thermal enhancement of Storm Water runoff by paved surfaces. Water Research, Volume 34(4):1359-1371.
Viessman, W. and M.J. Hammer. 2005. Water supply and pollution control, 7th ed., Pearson-Prentice Hall, Upper Saddle, New Jersey.
Wilson, B.G., K. Haralampides, and S. Levesque. 2004. Storm Water runoff from open windrow composting facilities. Journal of Environmental Engineering and Science 3: 537-540.
Ljubisa Kalaba1, Bruce G. Wilson2 and Katy Haralampides2
1. Gemtech Limited, Fredericton, New Brunswick, Canada
2. Department of Civil Engineering, University of New Brunswick, Fredericton, New Brunswick, Canada
Copyright J.G. Press Inc. Summer 2007
(c) 2007 Compost Science & Utilization. Provided by ProQuest Information and Learning. All rights Reserved.