Environ. Sci. Technol. 2007, 41, 277-283
Field Application of Partitioning Gas Tracer Test for Measuring Water in a Bioreactor Landfill BYUNGHYUN HAN,† P A U L T . I M H O F F , * ,† A N D RAMIN YAZDANI‡ Department of Civil and Environmental Engineering, University of Delaware, 360 DuPont Hall, Newark, Delaware 19716, and Division of Integrated Waste Management, Planning, Resources, and Public Works Department, Yolo County, 292 West Beamer Street, Woodland, California 95695
Two field-scale partitioning gas tracer tests (PGTTs) were performed to evaluate the utility of the PGTT method for measuring water saturation and moisture content in a fullscale bioreactor landfill, where waste biodegradation resulted in elevated temperatures and significant landfill gas production. The average water saturation and moisture content were measured for waste volumes of approximately 20 m3 and results were compared to gravimetric measurement of moisture content made on samples collected from the landfill. In the center of the landfill, the moisture content estimated from the PGTT was Mc ) 0.26 ( 0.03, which was nearly identical to the gravimetric measurement of waste samples taken from the same region (Mc ) 0.28). PGTT-estimated moisture contents in a dry area of the landfill were much smaller (Mc ) 0.10 ( 0.01) and consistent with available gravimetric measurements. Biodegradation of tracers and temporal variations in landfill gas production were minimal and did not influence the tests. These field experiments demonstrate the utility of the PGTT method for measuring water saturation and estimating moisture content in bioreactor landfills with active waste degradation and generation of landfill gases. However, use of the PGTT to estimate the in situ moisture content requires estimates of the refuse porosity, dry bulk density, and temperature, which might limit its application.
1. Introduction “Bioreactor” landfilling entails controlling and optimizing biological conditions in landfilled waste to allow more rapid and complete waste decomposition. A key component of operating landfills as bioreactors, either aerobically or anaerobically, is maintaining sufficient moisture so that waste biodegradation is not inhibited. Typically, this is achieved by adding water or recirculating leachate to maintain optimal moisture conditions. To determine how much water to add and where to add it, measures of moisture content (mass of water/total mass of waste) within the landfill are needed. The most common method of measuring moisture content is based on gravimetric measurement of core samples * Corresponding author phone: (302)831-0541; fax: (302)831-3640; e-mail:
[email protected]. † University of Delaware. ‡ Yolo County. 10.1021/es061233e CCC: $37.00 Published on Web 11/22/2006
2007 American Chemical Society
removed from the landfill. While this method produces reliable results, these measurements provide information only at the point where each sample is taken. Because of the heterogeneous nature of solid waste and the fact that water is known to move preferentially within landfills, point measurements may be inadequate. In addition, drilling and core removal are expensive and destructive, altering landfill properties at each drilling location. For this reason, coring and gravimetric measurement may not be a viable option for monitoring moisture content over the operational lifetime of a bioreactor landfill. Because of these problems, various alternative methods have been proposed and tested for measuring moisture in solid waste (e.g., 1, 2). While many of these techniques can be used to track the movement of moisture fronts during liquid addition, they have not been able to accurately quantify the degree of water saturation or moisture content of the waste (3). It is for this reason that new technologies are needed for measuring the moisture content of waste, since the maintenance of optimal moisture conditions is the most important operational parameter for bioreactor landfilling. The partitioning gas tracer test (PGTT) is a promising technology recently developed by hydrologists for in situ measurement of soil water saturation (4, 5). A previous laboratory study demonstrated the potential of PGTT through a comprehensive set of column experiments involving solid waste samples of different composition with varying amounts of water (6). While these tests demonstrated that the PGTT could be used to measure water in refuse, experiments were conducted in a closed system, at ambient temperatures, and with negligible biological activity and biogas production. More recently, seven pilot-scale PGTTs were conducted to assess the accuracy and reproducibility of the PGTT method for measuring water in a 17-year-old municipal solid waste landfill located at the Central Solid Waste Management Center in Sandtown, Delaware (7). Tests were periodically conducted with pulse input of tracers in a small near-surface region of the landfill (refuse volume ∼1.5 m3) over a 12-month period, and measured moisture conditions ranged from possible dry waste to refuse with a moisture content of 24.7%. This final moisture content of 24.7% was in reasonable agreement with gravimetric measurements of excavated refuse, where the moisture content was 26.5 ( 6.0 CI % (CI ) 95% confidence interval). While results from the previous laboratory and field tests are encouraging, these tests were conducted in relatively small waste volumes that were not undergoing significant biological degradation, with the concomitant production of landfill gases and heat generation, which elevates temperatures. There is a particular need to evaluate the utility of the PGTT method in bioreactor landfills, where (1) biological activity is great, potentially resulting in tracer degradation; (2) in situ gas production is significant, which may alter tracer transport and confound data interpretation; (3) temperatures are elevated and spatially variable, resulting in nonuniform partitioning of tracer into water; and (4) liquid is continually or intermittently injected often resulting in elevated water saturations in at least localized regions. Each of these factors may complicate the PGTT method. Finally, application of the PGTT method to existing landfills would likely require injecting and/or extracting tracers through existing gas collection wells for refuse volumes significantly larger than 1.5 m3 sampled in previous field tests. Field tests reported herein address each of these factors, which represent the unique contributions from this work. VOL. 41, NO. 1, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
277
In this study the PGTT method was evaluated for measuring water in a 3-year-old bioreactor landfill in Yolo County, California. In this landfill water was periodically added and leachate was continuously recirculated in an attempt to maintain optimal moisture conditions. The volumes of refuse sampled were ∼20 m3, ten times larger than the Sandtown, Delaware field test. In addition, biological activity was significant, resulting in elevated landfill temperatures and significant production of landfill gases. In support of the field tests, laboratory experiments were also conducted to quantify tracer partitioning as a function of temperature and to assess the influence of tracer concentration on partitioning into water.
2. Background The basic principle underlying the PGTT is the same as that for chromatographic separation of chemicals and is described in detail elsewhere (5, 6). Based on the mean travel times of a conservative and partitioning tracer, the fraction of the pore space occupied by water in streamtubes carrying the gas tracers can be determined from
Sw )
(Rf - 1)KH 1 + (Rf - 1)KH
(1)
where Sw is the water saturation, Rf is the retardation factor and is the ratio of the mean travel time of the partitioning tracer to the conservative tracer, and KH is the dimensionless Henry’s law constant of the partitioning tracer. Thus, if tracer retardation is dominated by partitioning into water, eq 1 can be used with the measured retardation coefficient to estimate the average water saturation in the volume of waste sampled by the gas tracers. In previous work difluoromethane was identified as a potentially suitable compound for the partitioning tracer for PGTT in solid waste (6). Although KH was only measured at room temperature in that work, theoretical analyses suggested that KH should be corrected for temperature in field PGTTs, since temperatures may vary between 10 and 60 °C in landfills (6). Dissolved inorganic salts may also alter KH, but this effect will increase KH by less than 5-10% for a typical leachate (conductivity ≈ 1.7 × 104 µS/cm) (6). An additional factor not discussed in this earlier work is the influence of tracer concentration on KH. Based on thermodynamic considerations, high tracer concentrations can affect tracer partitioning in the liquid phase and thus the Henry’s law constant (8). In the work described below, the influence of temperature and tracer concentration on KH was evaluated for difluoromethane. While the PGTT provides a direct measurement of the water saturation, landfill operators are primarily interested in the moisture content, the mass of water divided by the total mass of wet waste. The relationship between the water saturation and the moisture content is
Mc )
SwnFw Fdb
+ SwnFw
(2)
where Mc is the moisture content (mass of water/total wet mass of waste), n is the porosity, Fdb is the dry bulk density of the waste, and Fw is the density of water (ML-3). Thus, to estimate Mc from a PGTT it is necessary to estimate both the porosity and bulk density of the waste. For young landfills, bulk densities can be determined from the volume of the landfill and the known mass of waste deposited, as long as appropriate corrections are used to account for variations with depth caused by the additional waste compaction from the overburden and the initial moisture content (9). For older landfills, the initial estimated bulk densities must be corrected 278
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 1, 2007
for the effects of waste degradation and settlement, which can be accomplished with settlement models (10). Porosity is much more difficult to estimate since few have measured it for compacted solid waste. Because a technique to measure the porosity over the region of a PGTT is unknown, the porosity was estimated from limited measurements in the literature to calculate the moisture content. This estimation results in some uncertainty in determination of Mc that is not present in the measurement of Sw.
3. Materials and Methods 3.1. Laboratory Tests. 3.1.1. Temperature Effect on KH. To determine the influence of temperature on the Henry’s law constant for difluoromethane (DFM), KH was determined at temperatures ranging from 4 to 50 °C. This range was selected based on temperatures observed in landfills in North America and the limitations of the constant temperature rooms at the University of Delaware. The measurement of KH at each temperature was conducted using 250 mL amber glass bottles. Five bottles were prepared for each designated temperature. Each bottle was filled with 200 mL of distilled water, and 2.5 mL of 1% DFM gas (balance 99% helium, He) was injected into each bottle. The bottle was shaken intensely by hand and stored at a controlled temperature in a constant temperature room or in a water bath for more than 20 h to equilibrate. Selected measurements of DFM in the gas phase through time confirmed that the 20-hour equilibration period was adequate. Once equilibrated, the headspace gas of each bottle was analyzed for He and DFM concentrations using gas chromatography. The concentrations of He were used to confirm injected tracer mass. The concentration of DFM in the water was determined from mass balance, using the known injected mass of DFM and the measured gas-phase concentration. The Henry’s law constant was then determined from the measured concentrations. 3.1.2. Influence of Tracer Concentration on KH. To investigate the influence of tracer concentration on KH, procedures similar to those used to evaluate the influence of temperature were followed. Five 250 mL amber glass bottles were prepared for each concentration. Three bottles were filled with 200 mL of distilled water to measure KH, and two bottles were empty and used as controls to verify the mass injected. Gastight syringes were used to inject Suva 410A (Dupont Fluorochemicals, Wilmington, DE), a mixture of 50% difluoromethane and 50% pentafluoroethane, into all five gas bottles. Injected volumes were selected to provide almost 2 orders of magnitude range in injected DFM mass. The prepared bottles were shaken intensely by hand and stored in a water bath at room temperature for more than 20 h to achieve equilibrium partitioning. After equilibration, the headspace gas in each bottle was analyzed using gas chromatography and the Henry’s law constant was computed. 3.2. Field Tests in Aerobic Bioreactor at Yolo County. 3.2.1. Site Description. The Planning, Resources, and Public Works Department of Yolo County, California and the Institute for Environmental Management, Inc. constructed full-scale anaerobic and aerobic bioreactors in Yolo County, California to evaluate the efficacy of the bioreactor landfill concept at scales (230,000+ tons) representative of typical landfill operations. The southeast aerobic cell was used in this study and is equipped with an air collection system for air supply, liquid addition lines for leachate recirculation, sampling tubes for monitoring gas pressures and internal gas composition within the landfill, temperature sensors, and electrical resistance moisture sensors for measuring water. Two PGTTs were conducted in this bioreactor in May 2004, and the results from these tests were compared with
FIGURE 1. Plan and cross-sectional view of Yolo County aerobic bioreactor cell. Section A-A shows instrument location 1-13-SE, although these instruments and the trench containing them is offset horizontally from the trench containing 2-A3-SE, 2-3-SE, and 2-13-SE (see plan view).
moisture contents determined from solid waste cores removed from the landfill. Because air was not injected for extensive periods before and after the PGTTs, the bioreactor was anaerobic during these field tests. A cross-section through the landfill and a schematic of the field tests is shown in Figure S-1 in the Supporting Information, while a plan and cross-section view showing the locations of the two tracer tests is provided in Figure 1. Test 1 was near the center of the bioreactor, while Test 2 was near the west side. Both tests sampled solid waste located 3.0-5.5 m below the topmost surface of the landfill. The size, waste mass, and a summary of instrumentation used in the southeast aerobic cell are given in Table S-1 in the Supporting Information. 3.2.2. Procedure. Based on their successful use in the laboratory study (6), He and DFM were selected as the conservative and partitioning tracers in the field tests. In Test 1 tracers were contained in individual gas cylinders and then mixed in a 270 mL mixing chamber before injection into the landfill. For Test 2 the gases were purchased as a mixture in a single gas cylinder (Scott Specialty Gases, Inc., South Plainfield, NJ). While it is more difficult to achieve and maintain desired concentrations with the mixing chamber, the use of individual gas cylinders rather than a premixed cylinder significantly decreased tracer costs. To verify the concentrations of injected tracers, gas samples taken from injection port were analyzed using gas chromatography. For field measurements sample analysis times were kept to 15 min, which permitted frequent sampling. While analysis procedures were sufficient to separate He and DFM from other landfill gases, hydrogen was eluted at times similar to those of He and affected He measurements at low concentrations, which is discussed further below. The tracer gases were injected at about 100-120 L/min into the landfill for approximately 1 h for each test. Tracer gases were injected into existing gas sampling tubes (1-13SE for Test 1 and 2-1-SE for Test 2, see Figure 1) that were installed in the landfill during construction. Each sampling tube consisted of 9.53-mm outer diameter linear low-density polyethylene tubing that was terminated at a different landfill location. All sampling tubes were also terminated in an equipment shed adjacent to the landfill, where the tracers were injected. The detailed experimental conditions for both tests are shown in Table S-2 in the Supporting Information. Before and after tracer gas injection, gas flow in the landfill was monitored to ensure that gas flow was constant. Flow in the horizontal gas collection wells was adjusted so that landfill gases were extracted from a single horizontal gas
collection well (2-A3-SE for Test 1 and 1-A2-SE for Test 2, see Figure 1) at a volumetric flow rate of ∼ 2100 L/min during each test. Operation of a single gas collection well ensured that a significant portion of the injected gas tracers would pass through the region of solid waste intended for moisture measurement. To evaluate the spatial variability of moisture within each test region, gas was also withdrawn intermittently from gas sampling tubes (2-13-SE and 2-3-SE for Test 1, and 1-5-SE and 1-21-SE for Test 2, see Figure 1) and analyzed for tracer concentrations. As shown in Figure 1, these sampling tubes were installed in the same trenches containing the gas collection wells. PGTT results for the refuse sampled by tracers collected at each sampling tube were used to evaluate the uniformity of moisture within each test region, which are shaded in Figure 1. At the initiation of tracer injection, gas samples were collected from the main header pipe of the gas collection system. Some samples were immediately injected into the gas chromatograph for field analysis, while others were injected into pre-evacuated (∼
