BioCycle June 2005, Vol. 46, No. 6, p. 47
Data collected at research plots show that the potential exists for biosolids stored at application sites to affect the nitrogen content of shallow groundwater. A mitigating factor is the relatively slow rate of leachate formation.
J.M. Peckenham, J.A. Nadeau and A. Amirbahman

CONCERNS about nitrate loading to groundwater near sites where class B biosolids had been field stacked prompted the Maine Department of Environmental Protection (MDEP) to impose stricter siting regulations effective in 2002. In order to substantiate these stricter regulations, the MDEP sponsored an experimental field study to measure the amount of nitrogen, and what form it is in (its speciation), that might be leaching from field-stacked Class B biosolids. Field stacking of biosolids prior to spreading is a common practice in Maine. It is necessary in northern climates because biosolids are produced continuously while farmers have only narrow windows of time to apply the material to fields or croplands. Field stacking also allows for efficiency of operation at the farm as several hundred cubic yards of material can be applied in one block of time. Field stacking is typically done in areas of a farm that are remote from public access and on the unprepared ground.
This research project was a collaboration with the Senator George J. Mitchell Center for Environmental and Watershed Research and the Department of Civil and Environmental Engineering at the University of Maine. The field experiment was conducted at Highmoor Farm in Monmouth, Maine, an experimental farm within the University of Maine system.
EXPERIMENTAL DESIGN
The study site location was in a remote field that is also used for composting demonstrations. The stockpile experiment consisted of two parallel studies to evaluate leachate formation and transport into soil. The primary study was the collection of leachate and surface runoff from a stockpile built on a plastic barrier, referred to as a lined bed. The objective of the lined bed was to quantify the volume and composition of all liquid leaving the stockpile. The secondary study was the measurement of leachate moving vertically through the soil. This was accomplished by building a stockpile over an array of pan lysimeters installed at various depths. At this site, the soils are thick silty loams derived from glacial till.
Two lined beds were constructed using different field-stacking geometries to determine if the way in which the biosolids are stored in the field influences the release of nitrogen species. One bed measured approximately 13 feet by 82 feet and is referred to as the wide pile. The other had approximate dimensions of 8.2 feet by 115 feet and is referred to as the long pile. Both the wide and long beds were designed to generate two discharge streams – surface flow across the pile (overflow) and liquid that flowed through the pile (leachate). Surface flow was sampled separately from the leachate.
To construct both beds, vegetation was removed and the ground surface was sloped inward in a shallow “V” shape. A flexible PVC liner was cut to size and placed over the scraped area. A slotted flexible plastic pipe was run down the centerline of each bed. The plastic liner and drain pipe were covered with clean road sand to an approximate depth of 6 inches. A runoff collection gutter was constructed along the edge of the bed using conventional rain gutters.
When the construction of the lined beds was completed, shallow trenches were dug at the low end of the beds to form sampling sumps. The sumps held 5-gallon plastic buckets modified with a constant overflow pipe so that the collected liquids would flow through the buckets. The overflow was directed to a small level-spreader constructed of perforated PVC pipe. Liquid accumulation in the buckets was monitored and collected with automated flow meters and samplers.
The wide bed had 72 cubic yards of Class B lime-stabilized biosolids unloaded directly onto the sand covering. Approximately 51 cubic yards were placed on the long pile bed. The biosolids used contained approximately 3 percent nitrogen at 20 percent total solids.
Rainfall data were collected daily by farm research personnel and used in the study. It was intended to sample within 24 hours after a rainfall event that generated 132 gallons of leachate for sampling. Leachate generation was not closely linked to rainfall. Therefore, sampling was conducted on regular intervals ranging from one to two weeks. Actual precipitation during the experiment was close to the average for the region.
LYSIMETER PLOT
The second portion of the experiment was to collect leachate samples moving from the field stacked biosolids, through the soil and into pan lysimeters. This was accomplished by constructing a lysimeter plot beneath an unlined stockpile, which measured approximately 20 feet by 50 feet. A total of 15 pan lysimeters were installed beneath this plot – five each at depths of 0.8 feet, 1.6 feet and 3.3 feet. The lysimeters were arrayed by location near to the edge or center of the stockpile. Accumulated liquid drained into half-gallon polyethylene collection bottles containing a small diameter polyethylene tube running to the surface for sampling.
Due to the extremely rocky and dense nature of the soils in the experimental field, slit trenches were dug and the pans were placed at depth and back filled with the soil that was removed. While the disrupted soil does not preserve important physical properties of the soil such as macropores, the leachate collected in the pans was subject to a variety of soil chemical interactions. Samples collected were analyzed for nitrogen species (total Kjeldahl nitrogen, nitrate, nitrite, and ammonia), chemical oxygen demand, and total organic carbon.
RESULTS
Lined Beds: Surprisingly, there was very little surface runoff generated by the stockpiles during this experiment. Runoff volumes could not be measured because they were below the calibrated range of the gauges. Water either was sorbed by the biosolids or pooled on the surface until it evaporated. Even during intense rain events the generation of runoff was limited. The biosolids were able to absorb moisture and expand in volume, however, with this additional water content, they also became less cohesive and slumped.
Leachate production was measurable and continuous during this experiment. Leachate flow was slow at first (< 0.04 gallons/hour), then gradually increased to a peak rate of 0.09 gallons/hour, followed by a decline back to the starting rates. The peak rate occurred about one month after the material was stockpiled (Figure 1). The pile geometry had almost no effect on flow rates; both the wide and long piles had similar leachate evolutions.
Although there was an a priori correspondence between rainfall and leachate generation, the flux of leachate was modulated by in-pile processes. The absence of a simple relationship is readily seen in the data plotted in Figure 1. In particular, during the first two weeks of stockpiling there was negligible flux followed by a short-lived spike in flux near days 20 to 30 (August 8 to 18). Since there was no rainfall event near these dates, it is likely that a physical transformation occurred in the biosolids, possibly the breakdown of the polymer used to increase solids content. Following this change, leachate flux appeared to show a closer relationship with rainfall events, rising after several events and declining during a dry period.
The results from the experiments indicate that concentrations of nitrogen in the leachate increased over time (Figure 1). In particular there was a lag period of 10 to 14 days with very little leachate, then a rapid rise in concentrations, followed by a plateau after 30 to 40 days. No marked decrease in nitrogen species concentrations was noted during the experiment. The chemical oxygen demand and total organic carbon values were slightly different in that they were detected in greater concentrations that did decline later in the experiment.
The initial concern about the presence of nitrate in the leachate was not substantiated by these results (Figure 2). The dominant form of nitrogen was ammonia and this form can convert to nitrate under atmospheric conditions. The high organic carbon and oxygen demands measured suggest that leachate is a strongly reducing liquid, that will enhance the stability of ammonia, i.e., the leachate consumes oxygen so rapidly that the ammonia can’t be converted to nitrate (or nitrite).
An important part of this experiment was to determine the flux of nitrogen from the stockpiles to the soil and groundwater. Even though leachate may contain elevated concentrations of nitrogen, loadings would be dependent upon leachate flow rates. The loading of nitrogen gradually increased during the first month of stockpiling and reached a relative maximum at six to eight weeks. Loadings decreased markedly after two months.
Lysimeter Results: Nitrogen in the leachate was transported down into the soil. The results from the lysimeter samples exhibited two important characteristics: 1) Leachate composition varied markedly within a given depth range; and 2) Nitrogen species concentrations were significantly attenuated below two feet. For example, the samples from the lysimeters were very dark brown in the shallow samples and almost clear in the deep samples. However, some deep samples were colored, reflecting differential flow paths from the surface. As in the leachate from the lined cells experiment, ammonium dominated the other nitrogen species. Compared to nitrate, ammonium is much less mobile in soil, and tends to be held by the soil, however too much ammonia is toxic to plants. In addition, ammonium, which has a positive charge, can exchange with other cations and deplete soil of calcium, magnesium and other minerals.
Our results make a strong case for preferential flow paths as an important mechanism for nitrogen movement through soil columns. It appears that biological activity in the pores of the soil matrix may retard flow at a depth greater than two feet. It is possible that with this increased downward resistance to fluid flow, lateral movement may increase. If leachate moves laterally, more would be available for plants and thus consumed. (Grass around the research plots was six to eight feet tall.) In general, if downward movement is slowed, groundwater quality is protected.
DISCUSSION
Experimental results confirm that stockpiling has a smaller area impact than field spreading but the concentration of some nitrogen species released are higher, which has a potentially negative impact on groundwater quality. The data show, however, that the volumes of concentrations are low and decrease over time.
Liquid Loss. Two types of pile morphologies were built and runoff-leachate relationships were similar for each. Although there was some flow across the surface, neither pile shape generated measurable runoff. Almost all precipitation was adsorbed or evaporated from the piles. The leachate generated was a combination of precipitation passing through the stockpile and dewatering of the biosolids.
The biosolids stockpiles were stable and generated negligible leachate for a period of two to three weeks after forming the stockpiles. Leachate generation increased as the biosolids dewatered and accumulated precipitation was processed. The discharge of leachate was nearly constant at less than 0.10 gallons/hour for four to six weeks, after which the flux declined to less than 0.02 gallons/hour. This decline in leachate flow is believed to be the result of biosolids dewatering and compaction. As liquid drained out, the biosolids were sufficiently plastic to allow macropores to close. This process may be dependent upon specific biosolids composition.
An important observation is that the dewatering process was modulated by the ability of the stockpiled biosolids to absorb and release moisture. The stockpiles would visibly swell after a heavy rain. This phenomenon caused discharge to be independent of precipitation.
Nutrient Loss. Nitrogenous compounds were transported from the biosolids stockpiles and discharged in the leachate. Concentrations of total nitrogen quickly increased to several thousand ppm, and then slowly increased by hundreds of ppm during the period of testing. Contrary to expectations that we would find nitrate, most (>95 percent) of the nitrogen was in the form of ammonia (due to the leachate consuming more oxygen than expected, thus blocking conversion to nitrate or nitrite). Nitrate was consistently much less than 1 ppm and nitrite was just detectable at less than 0.1 ppm.
The dominance of ammonia and occurrence of elevated total organic carbon and high chemical oxygen demand in the leachate are all indicative of strong reducing conditions in the stockpile. These reducing conditions persisted into the soil. The potential exists for stockpiles to affect the nitrogen content of shallow groundwater. A mitigating factor is the relatively slow rate of leachate formation. Overall, storage of biosolids on bare ground for more than two weeks may be a concern because of leachate concentrations. These findings may lead to changes in storage practices, such as storing in controlled structures, stockpiling on compacted surfaces that have minimal perc rates and/or adopting new delivery and spreading policies.
John Peckenham is Director of the Maine Water Resources Research Institute and Senior Research Scientist at the Senator George J. Mitchell Center for Watershed and Environmental Research, based at the University of Maine, Orono. James Nadeau, and Aria Amirbahman are graduate student and associate professor respectively, in the department of civil and environmental engineering at the University of Maine. This project was funded by the Maine Department of Environmental Protection and the U.S. Geological Survey through the Maine Water Resources Research Institute.