Environ. Sci. Technol. 2008, 42, 5740–5745
Landfill Disposal of CCA-Treated Wood with Construction and Demolition (C&D) Debris: Arsenic, Chromium, and Copper Concentrations in Leachate J E N N A R . J A M B E C K , †,‡ T I M O T H Y G . T O W N S E N D , * ,† A N D HELENA M. SOLO-GABRIELE§ Department of Environmental Engineering Sciences, University of Florida, Gainesville, Florida 32611-6450, and Department of Civil, Architectural, and Environmental Engineering, University of Miami, P.O. Box 248294, Coral Gables, Florida 33124-0630
Received February 5, 2008. Revised manuscript received April 29, 2008. Accepted May 8, 2008.
Although phased out of many residential uses in the United States, the disposal of CCA-treated wood remains a concern because significant quantities have yet to be taken out of service, and it is commonly disposed in landfills. Catastrophic events have also led to the concentrated disposal of CCA-treated wood, often in unlined landfills. The goal of this research was to simulate the complex chemical and biological activity of a construction and demolition (C&D) debris landfill containing a realistic quantity of CCA-treated wood (10% by mass), produce leachate, and then evaluate the arsenic, copper, and chromium concentrations in the leachate as an indication of what may occur in a landfill setting. Copper concentrations were not significantly elevated in the control or experimental simulated landfill setting (R ) 0.05). However, the concentrations of arsenic and chromium were significantly higher in the experimental simulated landfill leachate compared to the control simulated landfill leachate (R ) 0.05, p < 0.001). This indicates that disposal of CCA-treated wood with C&D debris can impact leachate quality which, in turn could affect leachate management practices or aquifers below unlined landfills.
Introduction Chromated copper arsenate (CCA)-treated wood was a common residential construction material. While CCAtreated wood is no longer manufactured for residential uses in the United States, projections forecast annual CCA-treated wood disposal quantities to vary between 6 and 10 million m3 in the U.S. through 2030 (1). Even though the CCA-treated wood may be structurally sound for 25 years, research has shown that aesthetic reasons play a large role in when CCAtreated wood structures are removed from service and disposed of (2, 3). CCA-treated wood is contained in * Corresponding author phone: 352-392-0846; fax: 352-392-3076; e-mail:
[email protected]. † University of Florida. ‡ University of Miami. § Present address: Environmental Research Group, Department of Civil Engineering, University of New Hampshire, Durham, NH 03824-3534. 5740
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 15, 2008
construction and demolition (C&D) debris, which is primarily disposed of in landfills in the U.S. C&D debris may also be targeted for recycling, which typically excludes CCA-treated wood. CCA-treated wood is sometimes inadvertently mixed with C&D debris and contaminates recycled wood mulch (4, 5). On a life-cycle basis, combustion of CCA-treated wood with energy recovery is favorable with proper air pollution controls (1); however, the ash contains high metal concentrations (6, 7). Although research has progressed on new and effective air pollution controls and ash stabilization technologies (8), public and regulatory acceptance of any treated wood combustion remains low worldwide. Significant events creating disaster debris, such as hurricanes (e.g., Katrina), concentrate landfill disposal of CCA-treated wood in one location with unknown future impacts, especially if the landfills are unlined (9). Historically, C&D debris was considered inert, without putrescible materials like those found in municipal solid waste (MSW) (e.g., food waste). However, research has shown that C&D debris goes through active processes of biological activity that affect both leachate and gas concentrations (10). These active processes can impact the timing and quantity of metals released from any metal-containing treated wood. Twenty-seven states in the United States do not require C&D debris landfills to have bottom liners (11). When CCA-treated wood is managed in the C&D debris waste stream and is disposed in unlined C&D debris landfills, there is concern for the quantity of arsenic, chromium, and copper that may be released in leachate, which could subsequently impact groundwater. Previous research on CCA-treated wood metal leachability has focused on terrestrial environments and in-service use (12–17) and batch leaching studies (18–20). Studies targeting C&D debris leachate in general have included CCA-treated wood, and the applicable groundwater limit for arsenic (prior to 2006 was 50 µg/L) was exceeded by C&D debris with only 0.5% of CCA-treated wood by mass (10, 21). CCA-treated wood has also been leached in a simulated monofill disposal setting, illustrating that arsenic, chromium, and copper leach differently in a simulated landfill setting than batch leaching studies (22). In a companion paper, Khan et al. (17) included some characterization of the leachate collected as part of the study presented here. However, the Khan et al. work focused specifically on the speciation of arsenic released from CCAtreated wood, as well as prediction of arsenic releases from various landfill scenarios in Florida. Chromium and copper results were not evaluated, and Khan et al. did not emphasize other physiochemical measurements that affect metal leachability and solubility in disposal environments. The experiment presented here differs from prior work in that an emphasis is placed on evaluation of the major ion chemistry, including the role of sulfate/sulfides, in influencing the releases of arsenic, chromium, and copper. Specifically, the objective of this research was to simulate the complex chemical and biological activity of a C&D debris landfill containing a realistic quantity of CCA-treated wood (10% by mass), produce leachate, and then evaluate the arsenic, copper, and chromium concentrations in the leachate as an indication of what may occur in a landfill setting.
Materials and Methods The experiment consisted of the design and construction of simulated landfill environments (leaching columns) followed by the operation of these columns. Both leachate and gas 10.1021/es800364n CCC: $40.75
2008 American Chemical Society
Published on Web 06/26/2008
samples were collected and analyzed from the columns. A description of the columns, sample collection, and analytical methods are contained in this section. Leaching Columns. Two 6.7-m high and 0.3-m (1-ft) diameter leaching columns, also called lysimeters, were constructed at the Alachua County Solid Waste Landfill, located in North Central Florida, U.S. The lysimeters were constructed (from the bottom to the top) with: 15.2 cm of washed gravel, a stainless steel screen, 15.2 cm of washed gravel, 6.1 m of simulated C&D debris, a cap with a water distribution system, and a catchment basin for rainwater. Details on the construction of the lysimeters are contained in Jambeck (23). One lysimeter was a control lysimeter (containing 157 kg of C&D debris and no CCA-treated wood), and one lysimeter was an experimental lysimeter (containing 150 kg of C&D debris with CCA-treated wood). The simulated C&D debris was composed of typical components including wood (33%), concrete (29%), asphalt roofing shingles (14%), gypsum drywall (12%), and smaller amounts of metals (e.g., copper, steel and aluminum) and other materials (e.g., insulation) by dry weight. The experimental lysimeter contained 10% CCA-treated wood by mass (a portion of the total wood fraction). This 10% was composed of 50% new CCA-treated wood and 50% CCA-treated wood that had been in service approximately 10 years. Further details on the composition and generation of the C&D debris used in this experiment are provided in the Supporting Information. Natural precipitation was allowed to infiltrate the lysimeters (1 cm of precipitation on the lysimeters is equal to 0.73 L of water addition). The lysimeters were exposed to 212 cm of natural precipitation and were supplemented with deionized water during the dry season (September 2002 to February 2003) for a total of 321 cm of additional water (234 L) to each lysimeter over the 786 day experimental time period. The lysimeters were located outside and exposed to ambient temperature variations. Thermocouple wires (type T) were placed at three separate depths (6.1, 4.6, and 1.5 m) within the lysimeters to obtain temperature readings. Sample Collection and Analytical Methods. Temperature readings (Omega Model HH21 Microprocessor thermometer) were taken weekly. Gas readings were taken with a GEM-500 meter, which characterizes the percentage of methane, carbon dioxide, and oxygen in the lysimeter air space. Leachate samples were collected from the lysimeters one to two times per month (a total of 26 occasions) in 20-L containers to homogenize the sample before it was split into proper containers for preservation and analysis. Samples were analyzed for general water quality parameters and metals through the end of the experiment. General water quality parameters were measured in the field each time leachate was collected. These measurements included pH and oxidation reduction potential (ORP) (Accumet, Model AP62), dissolved oxygen (DO), temperature (YSI Inc., Model 55/12 FT), and conductivity (Hanna Instruments, Model HI 9033). Samples were stored at 4 °C prior to analysis. Samples for dissolved ions (preserved below pH of 2 for cation analysis) were filtered through 0.45-µm membrane filters and analyzed using ion chromatography (Dionex DX 500). Chemical oxygen demand (COD) was analyzed using a spectrometer (Hach, DR/4000) using method 2720 (24). Analyses for total dissolved solids and alkalinity were performed using standard methods (Methods 2540 C and 2320 B) (25). For metals analysis, leachate samples were not filtered, preserved below pH of 2 with nitric acid, and digested following U.S. EPA Method 3010A (26). The digestates were analyzed by inductively coupled argon plasma (ICP) (Thermo Jarrell Ash, Model 61E). Arsenic speciation of the leachate from this study (through day 380 of 786) was completed on unpreserved samples by high-performance liquid chromatography coupled with inductively coupled plasma-mass
spectrometry (HPLC ICP-MS). Arsenic species examined in the leachate included arsenate (As(V)), arsenite (As(III)), monomethylarsonic acid (MMAA), and dimethylarsinic acid (DMAA). Details and methods can be found in Khan et al. (17). Statistics lysimeters were completed on the metals results from both of the lysimeters. The t test for independent samples of unequal variance was used because the total number of sample results are greater than 30 and the variances, which are not equal, are taken into account (27). Raw data for all analyses completed on the leachate may be found in Jambeck (23). Quality assurance and quality control (QA/QC) included the collection and analysis of lysimeter leachate field blanks, duplicate analyses, and analysis of matrix spiked blank and leachate samples with acceptable results. Specific details on QA/QC procedures and results are contained in Jambeck (23).
Results and Discussion Gas Composition. The gas in the lysimeters did not contain methane but did contain carbon dioxide with corresponding low amounts of oxygen. Hydrogen sulfide was not quantified, but when leachate was sampled, a rotten egg odor characteristic of hydrogen sulfide was observed. The C&D debris in the lysimeters contained gypsum drywall (CaSO4 · 2H2O and paper), which releases sulfate in solution with the infiltration of water. A sulfate-rich C&D debris landfill environment promotes the growth of sulfate-reducing bacteria (28). Sulfate-reducing bacteria (SRB) convert sulfate to sulfide (which in the lower pH range exists as hydrogen sulfide) by utilizing sulfate as an electron acceptor and in the process create carbon dioxide. The rate at which sulfide is generated by SRBs depends on the amount of organic matter (which in this case could come from paper, cardboard, and wood), the concentration of dissolved oxygen in the leaching solution, the temperature, and the pH. The maximum carbon dioxide percentage observed was 22% (with oxygen at 1%) in July (the balance gas is likely primarily nitrogen with hydrogen sulfide). In the cooler winter months, the carbon dioxide gas percentage decreased to approximately 8%, while the oxygen increased to approximately 11%, indicating the SRBs were less active during the winter because of cooler temperatures. Air intrusion also likely occurred although the columns were completely sealed and a water trap was designed to keep gases in while still allowing precipitation to infiltrate. Further discussion of the temperature effects on the microbial activity in the lysimeters follows in the next section. Physiochemical Parameters. Physiochemical parameters were monitored throughout the experiment to evaluate whether typical C&D landfill conditions were simulated. The results obtained were consistent with C&D landfill conditions and degradation processes. A discussion of the most relevant physiochemical parameters is provided here, with additional data and a detailed comparison of the results to other studies provided in the Supporting Information. The pH, specific conductance, and ORP versus cumulative volume of leachate drained from the lysimeters followed similar patterns for both the control and experimental lysimeter (Figure 1). The pH of the lysimeter leachate remained consistent throughout the experiment in the range of 6.5-7, which is typical for C&D debris (10, 21, 29, 30). The specific conductance for the experimental lysimeter was greater at the start of the experiment than that measured for the control lysimeter (5.5 mS/cm versus 3.5 mS/cm). Because CCA-treated wood is impregnated with salts, it contributes more ions to solution than untreated wood; a similar observation was noted by Gifford et al. (31) in lysimeters containing soil with untreated wood and soil with CCAtreated wood. The specific conductance of both lysimeters VOL. 42, NO. 15, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5741
FIGURE 1. pH, specific conductance, and ORP in the lysimeter leachate. decreased over time as the cumulative volume increased, a common observation in dynamic leaching studies (10, 30). Both lysimeters exhibited reducing conditions throughout the experiment with an ORP in the range of -300 to -600 mV (Figure 1). Dissolved oxygen (DO), sulfides, COD, and alkalinity were all affected by the ambient seasonal temperature changes. DO concentrations were initially 3-5 mg/L in both lysimeters, decreasing to near zero in the warmer summer months, coinciding with the increase in carbon dioxide in the gas. The DO and ORP levels are indicative of the anaerobic/sulfate reducing phase of activity in a C&D debris landfill (10). During cooler months (affecting microbial activity), DO increased to 2 mg/L and then decreased again to near zero during the subsequent warmer months (Figure 2). Because of its effect on microbial activity, temperature and the concentration of sulfides in the leachate had similar trends (not a direct correlation as there is lag time in microbial activity adjustments). Initially both sulfide and sulfate concentrations increased (sulfate from dissolution of the drywall and sulfide from the SRB activity) and then sulfides decreased during the cooler time period, while sulfate remained high. After microbial activity returned in the warmer months, sulfides again increased while the sulfate concentration remained variable (Figure 2). 5742
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 15, 2008
FIGURE 2. Temperature/DO, sulfate/sulfide, and COD/alkalinity concentrations in the lysimeter leachate. COD can result from both organic and inorganic compounds. Because sulfide can exhibit an oxygen demand, COD also had a similar trend with sulfide production in both lysimeters. In addition, in the neutral pH range, sulfide exists as HS-, which can also exhibit alkalinity by accepting a proton. The alkalinity trend in both the control and experimental lysimeter were similar to each other and remained between 500 and 2500 mg/L as CaCO3. Alkalinity decreased during cooler temperatures, potentially because of less hydrogen sulfide and carbon dioxide from microbial activity produced during this time, and it rebounded quickly when temperatures increased. The overall trend for alkalinity did not decrease significantly over time, an indication that the source of bicarbonate (e.g., CO2) was not depleted (Figure 2). Overall the ORP, sulfide, COD, and alkalinity concentration trends indicate that microbial activity was still occurring at the termination of the experiment. Arsenic, Copper, and Chromium Concentrations. The physiochemical environment inside the lsyimeters impacted the leached concentrations of arsenic, chromium and copper in variety of ways. The total arsenic, copper and chromium content of the CCA-treated wood and the fraction of these
TABLE 1. Metal Content and Leachability of 10% CCA-Treated Wood by Mass in C&D Debris (leaching column) arsenic metal content in new CCA-treated wooda (mg/kg) metal content in used CCA-treated wooda (mg/kg) minimum concentration (mg/L) maximum concentration (mg/L) overall concentrationb (mg/L) percent leached
chromium
1390 ( 20.0 814 ( 52.4
copper 1450 ( 68.3
1960 ( 27.7 1340 ( 54.0 2550 ( 48.0 1.09
0.3
