- Email: [email protected]
Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors
Waste Management xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors Wenjie Sun a,⇑, Mei Sun b, Morton A. Barlaz b a b
Department of Civil and Environmental Engineering, Southern Methodist University, PO Box 750340, Dallas, TX, United States Department of Civil, Construction, and Environmental Engineering, North Carolina State University, Campus Box 7908, Raleigh, NC, United States
a r t i c l e
i n f o
Article history: Received 17 October 2015 Revised 22 January 2016 Accepted 25 January 2016 Available online xxxx Keywords: Sulfur-containing waste Coal combustion residue Ash Hydrogen sulfide Landfill
a b s t r a c t Landfills that accept municipal solid waste (MSW) in the U.S. may also accept a number of sulfurcontaining wastes including residues from coal or MSW combustion, and construction and demolition (C&D) waste. Under anaerobic conditions that dominate landfills, microbially mediated processes can convert sulfate to hydrogen sulfide (H2S). The presence of H2S in landfill gas is problematic for several reasons including its low odor threshold, human toxicity, and corrosive nature. The objective of this study was to develop and demonstrate a laboratory-scale reactor method to measure the H2S production potential of a range of sulfur-containing wastes. The H2S production potential was measured in 8-L reactors that were filled with a mixture of the target waste, newsprint as a source of organic carbon required for microbial sulfate reduction, and leachate from decomposed residential MSW as an inoculum. Reactors were operated with and without N2 sparging through the reactors, which was designed to reduce H2S accumulation and toxicity. Both H2S and CH4 yields were consistently higher in reactors that were sparged with N2 although the magnitude of the effect varied. The laboratory-measured first order decay rate constants for H2S and CH4 production were used to estimate constants that were applicable in landfills. The estimated constants ranged from 0.11 yr1 for C&D fines to 0.38 yr1 for a mixed fly ash and bottom ash from MSW combustion. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction There are approximately 2000 landfills in the U.S that are permitted to receive municipal solid waste (MSW) (U.S. EPA, 2014). In addition to MSW, many of these landfills receive a variety of non-hazardous industrial wastes. Examples of such wastes include (1) construction and demolition (C&D) waste that contains gypsum wallboard (i.e., CaSO4), (2) the fines fraction from C&D recycling facilities that contains small pieces of wallboard, (3) flue gas desulfurization (FGD) residue that is generated from processes to remove SOx from combustion off-gas at both coal-fired power plants and MSW combustion facilities, and (4) fly ash that may or may not be mixed with FGD. A common feature of these wastes is that they contain sulfate, which, when co-disposed with MSW in landfills, can be biologically reduced to hydrogen sulfide (H2S). For example, while the composition of C&D waste is highly variable, the drywall (gypsum or calcium sulfate) content has been estimated at 17–27% by weight (Cochran et al., 2007; U.S. EPA, 1998).
⇑ Corresponding author. E-mail address: [email protected] (W. Sun).
The presence of H2S in landfill gas (LFG) has been reported to occur at landfills throughout the U.S. and globally at concentrations as high as 12,000 ppm (Lee et al., 2006; Eun et al., 2007; Ko et al., 2015). The presence of H2S in landfill gas is problematic for several reasons: (1) its low odor threshold, 0.02–0.13 ppm, may result in odors due to fugitive LFG emissions (Beauchamp et al., 1984; OSHA, 2005), (2) it is toxic to humans and presents challenges for occupational safety in enclosed areas at landfills such as subgrade elements of the leachate collection system (Selene and Chou, 2003; WHO, 2000), and (3) it is corrosive to LFG collection and control systems. In addition to these well-recognized problems, the toxicity of H2S to anaerobic microbial activity is often overlooked. The biological formation of H2S via sulfate reduction has been reported to inhibit the activity of both sulfatereducing (SRB) (Reis et al., 1992; O’Flaherty et al., 1998) and methanogenic microorganisms (McDonald and Parkin, 2009). A series of biochemical reactions is initiated when MSW is disposed in landfills with CH4 and CO2 as the major end products (Eq. (1)). When sulfate is present, it can be biologically reduced to H2S (Eq. (2)). In both reactions, MSW or MSW leachate is the source of organic matter. In ecosystems in which organic matter is limiting, there is competition for organic matter between
http://dx.doi.org/10.1016/j.wasman.2016.01.028 0956-053X/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
2
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
sulfate-reducing and methane-producing microorganisms. However, there is an excess of organic matter in MSW landfills such that sulfate reduction and methane production proceed concurrently.
Organic matter þ H2 O $ CO2 þ CH4
ð1Þ
Organic matter þ SO2 4 þ H2 O $ H2 S þ CO2
ð2Þ
While the microbial processes associated with H2S production are well understood, the ability to predict the rate and extent of H2S production from sulfur-containing wastes in landfills is poor. Predictability of H2S production would allow landfill operators to estimate H2S concentrations in LFG as a function of the quantity of sulfur-containing waste received and may inform operating guidelines on the upper limit of a sulfur-containing waste to be accepted. Unfortunately, field-scale data on sulfate-reduction are limited. Anderson et al., (2010) analyzed landfill gas data from four landfills in the northeastern U.S. that used C&D fines for daily cover and estimated an average decay rate of 0.7 yr1. Tolaymat et al. (2013) estimated a decay rate of 0.14 yr1 for sulfate in drywall in a 1250 cm3 lysimeter. However, this decay rate was measured under carbon-limited conditions representative of a landfill that only accepts C&D waste. In previous work, Sun and Barlaz (2015) presented a procedure to measure the chemically leachable sulfate content for a range of sulfur-containing wastes including residuals from the combustion of coal, wood and MSW, as well as fines from the processing for C&D waste. However, the relationship between sulfate content, sulfate mineral form, and the rate and extent of sulfate conversion in simulated landfill systems has not as yet been established. The overall objectives of this research were to (1) demonstrate a method to measure H2S production from a range of sulfurcontaining wastes in laboratory-scale reactors that are not inhibited by the accumulation of H2S as an endproduct; (2) characterize the effect of gaseous H2S on both H2S and CH4 production; and (3) estimate field-scale rate constants for H2S production by using laboratory reactor data on H2S and CH4 production. 2. Materials and methods 2.1. Experimental design Tests were conducted in 8-L reactors that were filled with a mixture of newsprint to serve as the source of organic matter required for microbial sulfate reduction, leachate as an inoculum, and a sulfur-containing waste. Newsprint was selected as a lignocellulosic waste representative of MSW. The sulfur-containing wastes that were tested are listed in Table 1. It was hypothesized that the accumulation of H2S in the reactor system could result in inhibition of methane production and sulfate reduction. To eval-
uate the potential for H2S inhibition, reactors were operated with and without continuous N2 sparging to reduce gaseous H2S concentrations in the reactor headspace. Duplicate reactors were operated for each treatment. Reactors were operated under conditions designed to maximize the rate and extent of decomposition. This included inoculation of each reactor with leachate from a 300-L reactor of decomposing residential MSW, leachate neutralization and recirculation, the use of shredded newsprint as a moderately degradable carbon source, and incubation in a room maintained at 37 °C. All reactors were monitored until the H2S concentration was near zero, though some reactors were terminated before methane production was complete (120–450 days) as described in the Results. Control reactors with newsprint and inoculum only (no sulfur-containing waste added) were operated to verify the activity of the inoculum, and to measure background H2S and CH4 production in the absence of sulfate. 2.2. Materials The sulfur-containing wastes tested in this study included fly ash, FGD residues in which either trona (a sodium carbonate mineral) or lime was used for SOx removal, MSW combustion ash, and C&D fines (Table 1). All sulfur-containing wastes were dried in an oven at 75 °C, ground in a Wiley Mill to pass a 1 mm screen as needed, and redried prior to use. The newsprint was collected from the North Carolina State University daily newspaper and shredded in a copy paper shredder to long narrow ( 0.5 cm) strips before use. 2.3. Experimental equipment and reactor operation Reactors were made from 8-L polypropylene mason jars sealed with a screw cap (U.S. Plastics Corp., Lima, OH, USA). Jars were modified for installation of a leachate inlet port, a leachate collection port, a gas sparging port, and a gas collection port. Each port accepted 6.35 mm I.D. tubing. Gas was collected in flex-foil gas bags (SKC Corp., Houston, TX, USA) fitted with a hose valve. Leachate was collected in 2-L intravenous bags (Baxter Healthcare, Deerfield, IL, USA). Assembled reactors together with gas bags, leachate bags and gas connections were tested for leaks using a vacuum pump. The reactors were operated as stationary solid phase beds with leachate recirculated in down flow mode. In control reactors, 500 dry gm of shredded newsprint was added to each reactor in the absence of a sulfur-containing waste. In the treatment reactors, 500 dry gm of shredded newsprint and 125 dry gm of ground sulfur-containing waste were added to each reactor. One liter of leachate was added to each reactor, including the controls, as an inoculum. Leachate was obtained from a 300-L reactor that contained actively decomposing residential MSW. The
Table 1 Sulfur-containing wastes tested in laboratory-scale reactors.
a b
Sulfurcontaining waste
NCSU ID
Waste generation process description
Sulfate content a (% SO2 4 –S)
Sulfide content (% S)
Theoretical H2S yield from stoichiometric sulfate reduction (mL H2S/gm dry waste)
Fly Ash C&D fines
12–307 11–200
Fly Ash only without flue gas desulfurization (FGD) residue C&D fines that contain gypsum (CaSO4)
0.38 (0.02) 7.99 (0.15)
Not detected Not detected
2.68 55.93
Trona b Ash A Trona b Ash B Lime Ash A Lime Ash B
12–280 12–393 12–392 13–03
Fly Ash residue from wet scrubbers, dry sorbent injection, or spray dryer treatment in which trona or lime was used to neutralize SOx
3.64 1.07 0.96 2.28
(0.74) (0.01) (0.01) (0.31)
7.48 (0.96) Not detected Not detected 6.94 (0.68)
77.89 7.49 6.70 64.53
MSW Ash A MSW Ash B
12–365 13–199
Mixed fly ash and bottom ash from MSW combustion
2.00 (0.02) 1.55 (0.01)
Not detected Not detected
14.00 10.85
The results were adopted from Sun and Barlaz, 2015. Standard deviations are given parenthetically. Trona is a natural sodium carbonate mineral.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
3
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
presence of sulfate-reducing activity in the leachate inoculum was verified in preliminary work. Sufficient deionized (DI) water was added initially to ensure the generation of leachate for recirculation and then added periodically to maintain 500–600 mL leachate. Leachate was neutralized to 7.4–7.6 as necessary and recirculated about three times a week throughout the incubation period. Duplicate reactors were continuously sparged with humidified N2 (i.e., bubbled through water); while there was no sparging in a parallel set of two reactors. The N2 flow rate was set at 5 mL min1 and controlled by using a borosilicate glass tube flow meter (VWR, Radnor, PA, USA). This rate was designed to flush each reactor 3 times per day based on an estimate of the reactor pore space. Gas was added to a diffuser stone placed at the bottom of each reactor and a layer of Ottawa sand was placed above the stone to distribute the gas. Ottawa sand was also added above the newsprint/ sulfurcontaining waste mixture in each reactor to reduce the overall reactor pore space and subsequently the required N2 flow rate. All gas produced or sparged through the reactors was collected in 40-L flex-foil gas bags, and both total volume and gas composition were measured once or twice a week as needed. 2.4. Analytical methods The sulfate content was measured by a 5-day acid extraction at room temperature as described previously (Sun and Barlaz, 2015). Briefly, 10 g of sample were extracted in 1.95-L of deionized water and 50-mL of 10 N HCl in a 2-L plastic bottle that was shaken at 2 100 rpm for 5 days. Sulfate (SO2 4 ) and sulfite (SO3 ) were analyzed by suppressed conductivity ion chromatography (IC) using a Dionex IC-2500 system (Sunnyvale, CA, USA) fitted with a Dionex IonPac AS19 analytical column (4 250 mm) and an AG18 guard column (4 50 mm). The eluent was 20 mM KOH. For reporting, sulfate and sulfite were combined and expressed as sulfate-S. Sulfide in the liquid phase was analyzed by iodometric titration (U.S. EPA, 1996). An acid extraction method modified from U. S. EPA method 9030B (U.S. EPA, 1996) was used to measure the solid phase sulfide content (Sun et al., 2015). The newsprint was characterized by measurement of its cellulose, hemicellulose, and lignin content as described previously (Wang et al., 2015). Briefly, the cellulose and hemicellulose were converted to their monomeric sugars using a two-stage sulfuric acid hydrolysis. The monomeric sugars were then quantified using a high performance liquid chromatograph (Shimadzu, Columbia, Maryland, USA) equipped with an electrochemical detector. Klason lignin was determined by loss on ignition (2 h, 550 °C) of the solids remaining after 72% sulfuric acid hydrolysis. The biochemical methane potential (BMP) of newsprint was measured by incubation of a known mass of ground sample at 37 °C in triplicate 160 mL serum bottles that included anaerobic growth medium and an anaerobic consortium maintained and acclimated to the conversion of MSW to methane (Wang et al., 1997). BMPs were corrected for methane attributable to the inoculum. The methods employed for analyses of CH4 gas concentration and volume have been presented previously (Wang et al., 2015). Gaseous H2S was measured by using a Jerome H2S analyzer (Model Jerome J645) manufactured by Arizona Instrument LLC. (Chandler, AZ, USA). The Jerome H2S analyzer is an industrially calibrated ambient air analyzer with a range of 3 ppb (ppb) to 5 ppm (ppm). Gaseous H2S samples were prepared by using serial dilution; 1–10 mL of gas was removed from the reactor headspace using a gas tight glass syringe, and diluted in a 1-L flexfoil gas bag (SKC Corp., Houston, TX, USA) filled with N2 to a total volume was 1 L. Further dilution was repeated as needed to achieve a final H2S concentration in the range of 3 ppb to 5 ppm. In preliminary work, the accuracy of the instrument was confirmed by preparing dilutions from an H2S standard gas. The ratio of the measured to
theoretical H2S concentration ranged from 91.2% to 115% for samples between 0.025 and 5.35 ppm (Table S1 of the Supporting Information). 2.5. Data analysis All gas data were corrected to dry gas at 0 °C and 1 atm (STP). The H2S yield was corrected for H2S measured in the control reactors and attributable to the inoculum. Data from the sparged reactors were used to calculate a decay rate constant for each sulfur-containing waste by adopting the method described in De la Cruz and Barlaz (2010). To begin, it was assumed that H2S production could be estimated by using a first order reaction. A first order equation, such as the U.S. EPA’s Landfill Gas Emissions Model (LandGEM), is commonly used to represent the biological production of methane. For this study, the same equation was used to describe H2S production in a batch reactor in which all waste is disposed in one increment (Eq. (3)) (U.S. EPA, 2005).
Q ¼ ks L0 M ekt
ð3Þ 3
where Q is the rate of gas production (H2S or CH4) (m /yr); k is the first-order decay rate constant (H2S or CH4) (yr1); L0 is ultimate gas production (H2S or CH4) (m3/Mg); M is the mass of sulfate or MSW disposed (Mg/yr); and t is time (yr). Note that k in Eq. (3) is equivalent to k in the IPCC landfill gas model (IPCC, 2006). Eq. (3) was converted to a linear form (Eq. (4)) to calculate laboratory-scale first order decay rate constants for both H2S (kslab) and CH4 (kCH4-lab).
lnðY 0 Y t Þ ¼ klab t þ lnðY 0 Þ
ð4Þ
where Y0 is the ultimate yield (L0 for CH4 and Ls0 for H2S), Yt is the yield (H2S or CH4) at any time t during the reactor monitoring period, and klab (kslab or kCH4-lab) is the lab-scale first order decay rate constant. The cumulative H2S yield (Ls0) was calculated from the measured H2S yield in the sparged reactors and includes sulfide from both sulfate reduction and that released from the initial solid sample (in samples where solid phase sulfide was present). Time course data between 1% and 99% of the measured ultimate yield were used for the regression in Eq. (4). The square of the correlation coefficient (r2) for all linear regressions ranged from 0.85 to 1.0, which suggests that representing the data with a first order decay model was appropriate. 3. Results and discussion 3.1. Materials characterization The sulfur content of the wastes tested is presented in Table 1. Two of eight tested samples were shown to have non-sulfate sulfur in previous work and as a result, these samples were analyzed for solid phase sulfide by acid extraction (Sun et al., 2015) (Table 1). The volatile solids, cellulose, hemicellulose, and lignin contents of the ground newsprint were 95.3, 46.5, 16.3, and 17.1% of dry weight, respectively. The BMP was 125.7 mL CH4/dry gm newsprint. 3.2. Hydrogen sulfide and methane production in landfill simulation reactors The ultimate H2S yield of each waste is summarized in Table 2, and the H2S concentration and yield with time are presented in Figs. 1 and 2, respectively. H2S yields in the sparged reactors were significantly higher (p < 0.05) than those in the non-sparged reactors for all eight materials tested, which indicates that sparging enhanced the conversion of sulfate to sulfide for all wastes. While
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
4
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Table 2 H2S yield and recovery in laboratory-scale reactors. Treatment
Fly Ash C&D Fines Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
H2S yielda (mL H2S/gm dry sulfur-containing waste)
H2S recoveryb (%)
Sparged
Non-sparged
Sparged
Non-sparged
2.8 (0.2)c 22.3 (0.6) 30.9 (1.1) 1.0 (0.2) 0.6 (0.1) 16.1 (0.4) 1.6 (0.2) 2.5 (0.4)
2.4 8.6 1.9 0.1 0.2 3.7 0.3 0.2
102.9 (6.0) 39.8 (1.1) 121.2 (4.2)d 13.3 (3.1) 9.2 (0.8) 103.2 (2.6)d 11.3 (1.7) 22.6 (3.5)
88.1 (15.4) 15.5 (1.6) 7.6 (1.1) 1.8 (0.2) 2.7 (0.2) 23.5 (6.5) 2.0 (1.1) 1.7 (0.9)
(0.4) (0.9) (0.3) (0.0) (0.0) (1.0) (0.2) (0.1)
a H2S yield was corrected for background H2S in the control (newsprint only) reactors which ranged from 0.0002 to 0.02 mL H2S/gm newsprint. H2S concentrations in the controls were about 20 ppmv. The reported yield is based on the period before sparging was initiated in some of the non-sparged reactors as described in the text. b H2S recovery is the fraction of the initial sulfate recovered as H2S. c Standard deviations are given parenthetically. d If it is assumed that 100% of the initial solid phase sulfide was released to the gas phase, then the H2S recovery is 48.3 and 40% in Trona Ash A and Lime Ash B, respectively.
controlled toxicity tests were not conducted, the H2S concentration was as high as 7% (70,000 ppm) in the non-sparged reactors and orders of magnitude higher than in the sparged reactors, suggesting that H2S (i.e., end product) inhibition was responsible for the reduced H2S yields and reduced sulfate conversion in the nonsparged reactors. Other possible explanations for the results include additional sulfide precipitation in the non-sparged reactors, and endproduct inhibition due to CH4 and CO2, the concentrations of which would have been lower in the sparged reactors relative to the non-sparged treatments. As solid phase sulfide could not be quantified, the effect of sulfide precipitation cannot be examined. Both CH4 and CO2 routinely accumulate in standard BMP tests which would seem to preclude endproduct inhibition attributable these compounds (Wang et al., 1994). The effect of sparging was minimal in the reactors containing fly ash as calculated sulfate conversion was near 100% in both treatments (Table 1, Fig. 2A). Notably, the H2S concentration only remained elevated for a relatively short time (Fig. 1A) in the fly ash reactors. Methane yields are summarized in Table 3 and cumulative yield curves are presented in Fig. 3. A separate set of control reactors containing newsprint and inoculum was initiated each time that a set of reactors with a sulfur-containing waste was initiated which is why results for three sets of control reactors are presented. As noted in the Methods, all reactor sets were monitored until H2S production was complete so that a recovery and decay rate constant could be calculated. However, reactors in Sets B and C were terminated before methane production was complete. With the exception of Lime Ash A (Fig. 3G), methane yields were significantly higher in the sparged reactors relative to the nonsparged reactors (p < 0.05), which is consistent with the inhibition of H2S production inhibition in the non-sparged reactors. The explanation for the absence of an effect attributable to sparging in the Lime Ash A reactors is unclear though the behavior in these reactors was unique. H2S recovery in these reactors was only 9.2% and 2.7% in the sparged and non-sparged reactors, respectively, and there was little sulfate-reduction activity as evidenced by the low gaseous H2S concentrations and yield (Figs. 1G and 2G). Interestingly, the sparged control reactors produced significantly more methane than the non-sparged controls in each set despite the absence of sulfate in the controls (Table 3). Thus, although the methane yield in the Fly Ash reactors was higher in the sparged treatment, the ratio of the methane yields in the
sparged and non-sparged treatments in the Fly Ash reactors was comparable to that in the controls (Table 3), so no effect can be attributed to sparging. H2S concentrations of up to 30,000 ppm were present in the non-sparged Fly Ash reactors on day 75 (Fig. 1A) at which time methane production was equivalent in the sparged and non-sparged Fly Ash reactors (Fig. 3A). Sulfate conversion was essentially complete in the Fly Ash reactors within 50–100 days (Fig. 2A) while methane production continued for about 300 days. Once the effect of sparging was established, selected nonsparged reactors were sparged with N2 to evaluate whether either the one time or continuous removal of H2S (and other intermediated and endproducts) would stimulate sulfate reduction and/or methane production. For example, the duplicate Trona Ash A non-sparged reactors were sparged one time on days 100 and 175, respectively. While there was no substantial effect on H2S production (Fig. 2E), methane generation was stimulated after the one-time sparging (Fig. 3C) although the cumulative methane yield in the continuously sparged reactor treatment was still higher. Continuous N2 sparging was initiated in Lime Ash A and B and Trona Ash B after 80–110 days of operation. Both H2S and CH4 generation were stimulated in Lime Ash B and Trona Ash B but not in Lime Ash A (Figs. 2 and 3). The absence of an effect for Lime Ash A could be attributed to the fact that H2S accumulations were relatively low (Fig. 1G) so the effect of sparging on H2S concentrations was not as meaningful. This contrasts with Lime Ash B where sparging significantly reduced H2S concentrations (Fig. 1H). The explanation for the behavior of Trona Ash B reactors is not clear as continuous sparging of the previously non-sparged reactors stimulated both H2S (Fig. 2F) and CH4 (Fig. 3F) production despite the low accumulation of H2S in these reactors (Fig. 1F). Despite some inconsistencies, it appears that H2S toxicity on sulfate reduction and methane production is reversible if an appropriate strategy is applied to alleviate the accumulated H2S. 3.3. H2S recovery The fraction of the initial sulfate that was recovered as H2S is presented in Table 2. The recovery ranged from 9.2% to 121.2% in the sparged reactors and was lower in the non-sparged reactors. The calculated recovery is greater than 100% for Fly Ash, Trona Ash A and Lime Ash B. While the average Fly Ash recovery is slightly above 100%, this is likely a result of measurement error and the standard deviation of the averages suggests that the calculated average recovery of 102.9% is not different from 100%. In the case of Trona Ash A and Lime Ash B, these samples contained 7.5% and 6.9% solid phase sulfide, respectively. Thus, the likely explanation for recoveries of greater than 100% is that some or all of the initial sulfide was liberated during the decomposition experiment. If the theoretical H2S recovery is calculated as the measured H2S divided by the sum of the stoichiometric conversion of the initial sulfate to gaseous H2S plus the release of the initial solid phase sulfide to the gas phase, then H2S recoveries for Trona Ash A and Lime Ash B would be 48.3 ± 1.7% and 40.0 ± 1.0%, respectively. Although the recoveries are higher in the sparged treatments relative to the non-sparged treatments, the overall conversions are relatively low. Once the correction for initial solid phase sulfide is considered, the recoveries range from 9.2% to 48.3% for the samples tested, excluding the fly ash as discussed above. Two samples of Trona Ash were tested, and both their initial characterization and their biotransformation behavior were different. Trona Ash A had more sulfate but also contained solid phase sulfide. While Trona Ash A exhibited a high recovery, when the recovery is recalculated assuming that all of the solid phase sulfide
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
5
Fig. 1. H2S concentration in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors.)
was released as gaseous H2S, the recovery is only 48.3%, still well above the 13.3% conversion recorded for Trona Ash B. The effect of one time sparging on H2S production in the non-sparged reac-
tors was minimal for Trona Ash A (Fig. 2E) but significant for Trona Ash B (Fig. 2F), even though its overall recovery was still relatively low.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
6
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Fig. 2. H2S yield in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors; SW stands for sulfur-containing waste.)
Two samples of Lime Ash were tested, and their initial characterization and biotransformation behavior were also different.
Lime Ash B had more sulfate but also contained solid phase sulfide. Sulfate recoveries in Lime Ashes A and B, after correction for the
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
W. Sun et al. / Waste Management xxx (2016) xxx–xxx Table 3 CH4 production in laboratory-scale reactors.a Treatment
CH4 yield (mL CH4/gm dry newsprint) Sparged
Non-sparged
Control reactor (Set A) Fly Ash Trona Ash A C&D fines
165.1 163.4 155.2 151.6
120.2 (4.7) 116.8 (0.3) 1.8 (0.1) 69.7 (15.6)
Control reactor (Set B) Trona Ash B Lime Ash A Lime Ash B
86.3 72.7 77.3 80.5
Control reactor (Set C) MSW Ash A MSW Ash B
84.4 (2.1) 81.4 (3.6) 84.8 (0.7)
(3.4) (1.1) (1.0) (17.1)
(0.6) (4.1) (2.0) (1.4)
67.5 (4.0) 25.9 (12.6) 71.9 (6.7) 6.1 (0.8) 70.1 (1.9) 27.6 (3.1) 22.9 (3.8)
a Three sets of control reactors are presented because a new set of control reactors was set up each time a new set of sulfur-containing wastes was tested. Standard deviations are given parenthetically. In Sets 2 and 3, reactors were terminated before methane production was complete which is why the control reactor methane yields are lower in Control Sets B and C.
initial sulfide in Lime Ash B, were 9.2% and 40%, respectively. Continuous sparging had a significant effect on H2S (Fig. 2H) and CH4 (Fig. 3H) generation in Lime Ash B but no effect on CH4 generation in Lime Ash A (Fig. 2G and 3G). In the C&D fines reactors, H2S accumulated within about 14 days of reactor initiation (Fig. 1B), and the concentrations remained elevated (up to 30,000–35,000 ppm) for almost 200 days which appeared to impact both CH4 (Fig. 3D) and H2S production (Fig. 2B). While both CH4 and H2S production were stimulated by sparging and the resulting shorter duration of H2S accumulation in the sparged reactors (about 100 days), overall H2S recovery was still relatively low even in the sparged reactors compared to the maximum H2S potential calculated from the conversion of all the sulfate in the C&D fine sample. The low H2S yield in the sparged C&D fines reactors is surprising given the solubility of gypsum. In previous work, Yang et al. (2006) reported H2S generation was reduced by the presence of wood and concrete as were present in our samples. Yang et al. explained that the presence of concrete could have increased the system pH, which would inhibit biological sulfate reduction, and that concrete also acted as a sink for the produced sulfide. While concrete was undoubtedly present in the C&D fines tested here, it was not quantified. If pH was important in this research, then the pH effects would have had to have been very localized as the reactor leachate pH was between 7.5 and 7.9 in the C&D fines reactors and comparable to the pH in all other reactors (data not shown). Sulfate recovery in both MSW Ashes was relatively low. While sparging stimulated both CH4 (Fig. 3J and K) and H2S (Fig. 2C and D) production, measured H2S concentrations never exceeded 9000 ppm (Fig. 1C and D) in any of the reactors, a concentration well below those observed in the Fly Ash reactors which had much higher recovery. The effect of the elevated H2S concentrations on sulfate conversion and methane generation appeared to vary between wastes. For example, among the sparged reactors, H2S concentrations were highest in Trona Ash A and Lime Ash B (Fig. 1E and H), yet these two materials had among the highest sulfate recoveries, even after consideration of the impact of the initial sulfide. Similarly, methane production in the sparged reactors containing these wastes is comparable to that in the controls (Fig. 3A, C, E and H). Thus, general statements about the extent of sulfate conversion based on the initial sulfate content are not possible for the wastes tested and there may be factors not identified in this study that also affect sulfate conversion.
7
The measured H2S yields are based on H2S recovery in the reactor gas phase. Aqueous sulfide was minimal in the final leachate sample from the sparged reactors. Although solid phase sulfide was measured in the residual solids at the end of the monitoring period, the results were so variable that they were not useful and hence not reported. Nonetheless, qualitatively, some solid phase sulfide was present. It is possible that the extent of sulfide precipitation varied by waste which would influence gaseous H2S recovery. The presence of metals in the ash samples may also exert toxicity. Metals concentrations in the samples are presented in Table 4. Interestingly, the Trona Ash B and Lime Ash A had the highest Fe concentrations which might suggest that the presence of Fe contributed to the low measured gaseous H2S yield and corresponding low H2S recoveries due to FeS(s) formation. However, when H2S recovery was plotted against the individually measured metals concentrations for all wastes, there were no apparent correlations. 3.4. Estimated H2S decay rate in laboratory and field scales The reactor data were used to estimate laboratory-scale decay rate constants as described in the Methods. The laboratory decay rate constants were then used to estimate a decay rate for H2S that is applicable at field-scale by assuming that the ratio of the decay rate constant of newsprint conversion to methane and that of a sulfur-containing waste to H2S is constant between the lab and the field as expressed by Eq. (5).
CH4 decay rate of newsprint at lab-scale H2 S decay rate at lab-scale CH4 decay rate of newsprint at field-scale ¼ H2 S decay rate at field-scale
ð5Þ
Newsprint was used because it was the source of organic matter in the reactors and its methane production rate constant was measured in parallel with H2S production from the sulfurcontaining wastes (Table 5). The laboratory newsprint decay rate in the reactors of Control Set A, Fly Ash, Trona Ash A and C&D fines was 5.6 yr1 (range 5.4–5.7 yr1) and this average was used in Eq. (5). With reference to Eq. (5), the decay rate of newsprint at field-scale (0.033 yr1) was adopted from a published estimate. This field-scale newsprint decay rate was estimated from laboratory data and an algorithm to scale decay rates measured in laboratory reactors to a landfill as described previously (De la Cruz and Barlaz, 2010). This newsprint decay rate corresponds to a landfill with an MSW decay rate of 0.04 yr1. Using Eq. (5), estimates for decay rates (ksfield) for the tested sulfurcontaining wastes are presented in Table 5 and range from 0.11 to 0.38 yr1. The field-scale H2S decay rate constants presented in Table 5 are based on the field-scale CH4 decay rate constant for newsprint, which, as described above, was based on a landfill with a bulk MSW decay rate constant of 0.04 yr1. In previous research on U. S. landfills, it has been reported that a field-scale MSW methane decay rate constant of 0.09–0.12 yr1 may be more appropriate (Wang et al., 2013). If a field-scale bulk MSW decay rate of 0.1 yr1 were assumed, then the field-scale CH4 decay rate constant for newsprint and the calculated ksfield values given in Table 5 would increase by a factor of 2.5. The only published estimate of field-scale rate constants of H2S is that presented in Anderson et al. (2010) who estimated a decay rate constant of 0.7 yr1 for landfills that used C&D fines as daily cover. There are, however, so many differences between the values in Anderson et al. (2010) and the value estimated here that a direct comparison is flawed. To estimate a decay rate from field data, Anderson et al. (2010) allowed ksfield and the
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
8
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Fig. 3. CH4 yield in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors; NP stands for newsprint.)
ultimate sulfide generation potential (Ls0) to vary simultaneously to get the best fit between the model and measured data. For the estimates developed here, Ls0 was treated as an intrinsic property of the waste and was not allowed to vary when calculating kslab. In addition, Anderson et al. (2010) had minimal data
on the actual H2S generation rate and did not consider a timevarying collection efficiency, which is a confounding factor in working with data from field-scale landfills. Finally, there was uncertainty in both the mass of sulfur actually buried over time and the mass of H2S recovered through the gas collection and
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
9
W. Sun et al. / Waste Management xxx (2016) xxx–xxx Table 5 Estimation of field-scale H2S decay rate constants for sulfur-containing wastes.a Treatment
H2S decay rate constant (Lab scale) (yr1)
CH4 decay rate constant (Lab scale) (yr1)
H2S decay rate constant (Field scale) (yr1)
Control reactor Fly Ash C&D fines Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
N/Ab 30.8 (1.3) 18.4 (6.7) 25.6 (0.2) 35.5 (1.9) 39.0 (0.1) 37.3 (2.4) 61.4 (0.6) 64.0 (0.2)
5.6 5.7 5.4 5.5 5.6 5.6 5.6 5.6 5.6
N/A 0.18 0.11 0.15 0.21 0.23 0.22 0.36 0.38
a b
(0.3) (0.1) (0.2) (0.2) (0.1) (0.1) (0.1) (0.1) (0.1)
(0.01) (0.04) (0.01) (0.01) (0.00) (0.01) (0.00) (0.00)
Standard deviations are given parenthetically. N/A means not applicable.
control system in Anderson et al. The estimated decay rate constant for C&D fines derived in this study was 0.11 yr1 and this would increase to 0.27 yr1 at a bulk MSW decay rate constant of 0.1 yr1, which is likely more representative of a wet landfill in the northeastern U.S. While there is considerable uncertainty in the decay rate estimates, it is noteworthy that both the values in Table 5 and the values proposed by Anderson et al. (2010) are consistent in that the laboratory-scale decay rate constant for H2S production from C&D fines (18.4 yr1) is higher than the methane production decay rate constant (5.6 yr1). The implication of this is that H2S production from C&D fines will decrease faster than methane generation from MSW. 4. Summary and conclusions The H2S production potential of a range of sulfur-containing wastes was measured in 8-L reactors that were operated to maximize the rate of decomposition by the use of leachate neutralization and recirculation, the presence of excess organic carbon, inoculation, and incubation at 37 °C. Duplicate reactors were operated with and without sparging to evaluate the effect of endproduct (H2S(g), CH4 and CO2) removal. Sparging was shown to consistently increase both H2S and CH4 yields, suggesting that H2S endproduct accumulation was inhibitory to the biological processes occurring in the reactors. H2S recovery varied and was not correlated with either the initial sulfate or total S in a sample which suggests that chemical analyses are not suitable for prediction of H2S yields. The laboratory-scale H2S production rate constants were used to estimate field-scale rate constants which ranged from 0.11 to 0.38 yr1. These values are generally higher than the parallel values estimated for methane production, suggesting that sulfate will be depleted more rapidly that the cellulosic component of MSW.
Fig. 3 (continued)
Table 4 Metals in the sulfur-containing wastes.a Sulfur-containing waste
Al Wt.%
As mg/kg
Cd mg/kg
Co mg/kg
Cr mg/kg
Cu mg/kg
Fe Wt.%
Ni mg/kg
Pb mg/kg
Sb mg/kg
Se mg/kg
Zn mg/kg
Hg mg/kg
Fly Ash Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
2.23 0.78 1.35 1.29 0.51 2.63 2.51
68.0 17.1 80.6 100 69.8 13.4 8.99
<5.00 <5.00 <5.00 <5.00 <5.00 54.7 43.7
20.9 9.70 10.8 13.6 5.19 9.37 9.03
53.6 12.6 72.6 48.6 15.5 50.3 62.0
72.9 24.0 31.9 40.1 40.4 1089 1195
1.85 0.35 3.61 4.00 0.97 1.24 2.26
37.3 17.1 34.0 45.3 11.4 35.3 36.7
28.3 15.2 27.6 17.4 8.94 926 718
<5.00 <5.00 7.59 <5.00 5.47 136 40.2
5.93 52.4 <5.00 <5.00 52.4 <5.00 <5.00
47.0 24.4 126 110 24.8 5287 3952
<5.00 <5.00 <5.00 <5.00 <5.00 <5.00 <5.00
a
Most analytes are reported in mg/kg – Al and Fe are reported in weight% (1% by weight is equivalent to 10,000 mg/kg). The minimum detection limit (MDL) is 5 mg/kg on
ICP.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
10
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Acknowledgements This work was supported by Waste Management Inc. and the Environmental Research and Education Foundation. We appreciate the analytical support of David Black at North Carolina State University.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.wasman.2016.01. 028.
References Anderson, R., Jambeck, J.R., McCarron, 2010. Modeling of hydrogen sulfide generation from landfills beneficially utilizing processed construction and demolition materials, final report to Environmental Research and Education Foundation. Beauchamp, R.O., Bus, J.S., Popp, J.A., Boreiko, C.J., Andjelkovich, D.A., 1984. A critical review of the literature on hydrogen-sulfide toxicity. Crit. Rev. Toxicol. 13, 25– 97. Cochran, K., Townsend, T., Reinhart, D., Heck, H., 2007. Estimation of regional building-related C&D debris generation and composition: case study for Florida, US. Waste Manage. 27, 921–931. De la Cruz, F.B., Barlaz, M.A., 2010. Estimation of waste component-specific landfill decay rates using laboratory-scale decomposition data. Environ. Sci. Technol. 44, 4722–4728. Eun, S., Reinhart, D., Cooper, D., Townsend, T., Faour, A., 2007. Hydrogen sulfide flux measurements from construction and demolition debris (C&D) landfills. Waste Manage. 27, 220–227. IPCC Intergovernmental Panel on Climate Change, 2006. IPCC Guidelines for National Greenhouse Gas Inventories, Chapter 3; IPCC National Greenhouse Gas Inventory Program; IPCC, WMO: Geneva, Switzerland, vol. 5. Ko, J.H., Xu, Q.Y., Jang, Y.C., 2015. Emissions and control of hydrogen sulfide at landfills: a review. Crit. Rev. Environ. Sci. Tec. 45 (19), 2043–2083. Lee, S., Xu, Q.Y., Booth, M., Townsend, T.G., Chadik, P., Bitton, G., 2006. Reduced sulfur compounds in gas from construction and demolition debris landfills. Waste Manage. 26, 526–533.
McDonald, H.B., Parkin, G.F., 2009. Effect of sulfide inhibition and organic shock loading on anaerobic biofilm reactors treating a low-temperature, high-sulfate wastewater. Water Environ. Res. 81 (3), 265–288. O’Flaherty, V., Mahony, T., O’Kennedy, R., Colleran, E., 1998. Effect of pH on growth kinetics and sulphide toxicity thresholds of a range of methanogenic, syntrophic and sulphate-reducing bacteria. Process Biochem. 33 (5), 555–569. OSHA, 2005. OSHA fact sheet: Hydrogen sulfide (H2S); U.S. Department of Labor. Reis, M.A.M., Almeida, J.S., Lemos, P.C., Carrondo, M.J.T., 1992. Effect of hydrogen sulfide on growth of sulfate reducing bacteria. Biotechnol. Bioeng. 40, 593–600. Selene, C.H., Chou, J., 2003. Hydrogen Sulfide: Human Health Aspects. World Health Organization, Geneva. Sun, W.J., Barlaz, M.A., 2015. Measurement of chemical leaching potential of sulfate from landfill disposed sulfate containing wastes. Waste Manage. 36, 191–196. Sun, M., Sun, W.J., Barlaz, M.A., 2015. A batch assay to measure microbial hydrogen sulfide production from sulfur-containing solid wastes. Sci. Total Environ. http://dx.doi.org/10.1016/j.scitotenv.2016.01.161. Tolaymat, T.M., El Badawy, A.M., Carson, D.A., 2013. Estimate of the decay rate constant of hydrogen sulfide from drywall in a simulated bench-scale study. J. Envrion. Eng. 139, 538–544. U.S. EPA, 1996. Method 9030B:Acid-soluble and acid-insoluble sulfides: distillation (Revision 2). U.S. EPA, 1998. Characterization of building-related construction and demolition debris in the United States. United States Environmental Protection Agency, Municipal and Industrial Solid Waste Division Office of Solid Waste. U. S. EPA, 2005. Landfill Gas Emissions Model (LandGEM) Version 3.02 User’s Guide; U.S. EPA: Washington, D.C. U.S. EPA, 2014. Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures for 2012. Wang, Y.S., Byrd, C.S., Barlaz, M.A., 1994. Anaerobic biodegradability of cellulose and hemicellulose in excavated refuse samples using a biochemical methane potential assay. J. Ind. Microbiol. 13, 147–153. Wang, Y.S., Odle, W.S., Eleazer, W.E., Barlaz, M.A., 1997. Methane potential of food waste and anaerobic toxicity of leachate produced during food waste decomposition. Waste Manage. Res. 15, 149–167. Wang, X., Nagpure, A.S., DeCarolis, J.F., Barlaz, M.A., 2013. Using observed data to improve estimated methane collection from select U.S. landfills. Environ. Sci. Technol. 47 (7), 3251–3257. Wang, X., De la Cruz, F., Ximenes, F., Barlaz, M.A., 2015. Decomposition and carbon storage of selected paper products in laboratory-scale landfills. Sci. Total Environ. 532 (1), 70–79. WHO, 2000. Hydrogen sulfide. Air quality guidelines for europe. Second edition. World Health Organization, Regional Office for Europe, Copenhagen, Denmark. Yang, K., Xu, Q., Townsend, T.G., Chadik, P., Bitton, G., Booth, M., 2006. Hydrogen sulfide generation in simulated construction and demolition debris landfills: impact of waste composition. J. Air Waste Manage. Assoc. 56, 1130–1138.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors Wenjie Sun a,⇑, Mei Sun b, Morton A. Barlaz b a b
Department of Civil and Environmental Engineering, Southern Methodist University, PO Box 750340, Dallas, TX, United States Department of Civil, Construction, and Environmental Engineering, North Carolina State University, Campus Box 7908, Raleigh, NC, United States
a r t i c l e
i n f o
Article history: Received 17 October 2015 Revised 22 January 2016 Accepted 25 January 2016 Available online xxxx Keywords: Sulfur-containing waste Coal combustion residue Ash Hydrogen sulfide Landfill
a b s t r a c t Landfills that accept municipal solid waste (MSW) in the U.S. may also accept a number of sulfurcontaining wastes including residues from coal or MSW combustion, and construction and demolition (C&D) waste. Under anaerobic conditions that dominate landfills, microbially mediated processes can convert sulfate to hydrogen sulfide (H2S). The presence of H2S in landfill gas is problematic for several reasons including its low odor threshold, human toxicity, and corrosive nature. The objective of this study was to develop and demonstrate a laboratory-scale reactor method to measure the H2S production potential of a range of sulfur-containing wastes. The H2S production potential was measured in 8-L reactors that were filled with a mixture of the target waste, newsprint as a source of organic carbon required for microbial sulfate reduction, and leachate from decomposed residential MSW as an inoculum. Reactors were operated with and without N2 sparging through the reactors, which was designed to reduce H2S accumulation and toxicity. Both H2S and CH4 yields were consistently higher in reactors that were sparged with N2 although the magnitude of the effect varied. The laboratory-measured first order decay rate constants for H2S and CH4 production were used to estimate constants that were applicable in landfills. The estimated constants ranged from 0.11 yr1 for C&D fines to 0.38 yr1 for a mixed fly ash and bottom ash from MSW combustion. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction There are approximately 2000 landfills in the U.S that are permitted to receive municipal solid waste (MSW) (U.S. EPA, 2014). In addition to MSW, many of these landfills receive a variety of non-hazardous industrial wastes. Examples of such wastes include (1) construction and demolition (C&D) waste that contains gypsum wallboard (i.e., CaSO4), (2) the fines fraction from C&D recycling facilities that contains small pieces of wallboard, (3) flue gas desulfurization (FGD) residue that is generated from processes to remove SOx from combustion off-gas at both coal-fired power plants and MSW combustion facilities, and (4) fly ash that may or may not be mixed with FGD. A common feature of these wastes is that they contain sulfate, which, when co-disposed with MSW in landfills, can be biologically reduced to hydrogen sulfide (H2S). For example, while the composition of C&D waste is highly variable, the drywall (gypsum or calcium sulfate) content has been estimated at 17–27% by weight (Cochran et al., 2007; U.S. EPA, 1998).
⇑ Corresponding author. E-mail address: [email protected] (W. Sun).
The presence of H2S in landfill gas (LFG) has been reported to occur at landfills throughout the U.S. and globally at concentrations as high as 12,000 ppm (Lee et al., 2006; Eun et al., 2007; Ko et al., 2015). The presence of H2S in landfill gas is problematic for several reasons: (1) its low odor threshold, 0.02–0.13 ppm, may result in odors due to fugitive LFG emissions (Beauchamp et al., 1984; OSHA, 2005), (2) it is toxic to humans and presents challenges for occupational safety in enclosed areas at landfills such as subgrade elements of the leachate collection system (Selene and Chou, 2003; WHO, 2000), and (3) it is corrosive to LFG collection and control systems. In addition to these well-recognized problems, the toxicity of H2S to anaerobic microbial activity is often overlooked. The biological formation of H2S via sulfate reduction has been reported to inhibit the activity of both sulfatereducing (SRB) (Reis et al., 1992; O’Flaherty et al., 1998) and methanogenic microorganisms (McDonald and Parkin, 2009). A series of biochemical reactions is initiated when MSW is disposed in landfills with CH4 and CO2 as the major end products (Eq. (1)). When sulfate is present, it can be biologically reduced to H2S (Eq. (2)). In both reactions, MSW or MSW leachate is the source of organic matter. In ecosystems in which organic matter is limiting, there is competition for organic matter between
http://dx.doi.org/10.1016/j.wasman.2016.01.028 0956-053X/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
2
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
sulfate-reducing and methane-producing microorganisms. However, there is an excess of organic matter in MSW landfills such that sulfate reduction and methane production proceed concurrently.
Organic matter þ H2 O $ CO2 þ CH4
ð1Þ
Organic matter þ SO2 4 þ H2 O $ H2 S þ CO2
ð2Þ
While the microbial processes associated with H2S production are well understood, the ability to predict the rate and extent of H2S production from sulfur-containing wastes in landfills is poor. Predictability of H2S production would allow landfill operators to estimate H2S concentrations in LFG as a function of the quantity of sulfur-containing waste received and may inform operating guidelines on the upper limit of a sulfur-containing waste to be accepted. Unfortunately, field-scale data on sulfate-reduction are limited. Anderson et al., (2010) analyzed landfill gas data from four landfills in the northeastern U.S. that used C&D fines for daily cover and estimated an average decay rate of 0.7 yr1. Tolaymat et al. (2013) estimated a decay rate of 0.14 yr1 for sulfate in drywall in a 1250 cm3 lysimeter. However, this decay rate was measured under carbon-limited conditions representative of a landfill that only accepts C&D waste. In previous work, Sun and Barlaz (2015) presented a procedure to measure the chemically leachable sulfate content for a range of sulfur-containing wastes including residuals from the combustion of coal, wood and MSW, as well as fines from the processing for C&D waste. However, the relationship between sulfate content, sulfate mineral form, and the rate and extent of sulfate conversion in simulated landfill systems has not as yet been established. The overall objectives of this research were to (1) demonstrate a method to measure H2S production from a range of sulfurcontaining wastes in laboratory-scale reactors that are not inhibited by the accumulation of H2S as an endproduct; (2) characterize the effect of gaseous H2S on both H2S and CH4 production; and (3) estimate field-scale rate constants for H2S production by using laboratory reactor data on H2S and CH4 production. 2. Materials and methods 2.1. Experimental design Tests were conducted in 8-L reactors that were filled with a mixture of newsprint to serve as the source of organic matter required for microbial sulfate reduction, leachate as an inoculum, and a sulfur-containing waste. Newsprint was selected as a lignocellulosic waste representative of MSW. The sulfur-containing wastes that were tested are listed in Table 1. It was hypothesized that the accumulation of H2S in the reactor system could result in inhibition of methane production and sulfate reduction. To eval-
uate the potential for H2S inhibition, reactors were operated with and without continuous N2 sparging to reduce gaseous H2S concentrations in the reactor headspace. Duplicate reactors were operated for each treatment. Reactors were operated under conditions designed to maximize the rate and extent of decomposition. This included inoculation of each reactor with leachate from a 300-L reactor of decomposing residential MSW, leachate neutralization and recirculation, the use of shredded newsprint as a moderately degradable carbon source, and incubation in a room maintained at 37 °C. All reactors were monitored until the H2S concentration was near zero, though some reactors were terminated before methane production was complete (120–450 days) as described in the Results. Control reactors with newsprint and inoculum only (no sulfur-containing waste added) were operated to verify the activity of the inoculum, and to measure background H2S and CH4 production in the absence of sulfate. 2.2. Materials The sulfur-containing wastes tested in this study included fly ash, FGD residues in which either trona (a sodium carbonate mineral) or lime was used for SOx removal, MSW combustion ash, and C&D fines (Table 1). All sulfur-containing wastes were dried in an oven at 75 °C, ground in a Wiley Mill to pass a 1 mm screen as needed, and redried prior to use. The newsprint was collected from the North Carolina State University daily newspaper and shredded in a copy paper shredder to long narrow ( 0.5 cm) strips before use. 2.3. Experimental equipment and reactor operation Reactors were made from 8-L polypropylene mason jars sealed with a screw cap (U.S. Plastics Corp., Lima, OH, USA). Jars were modified for installation of a leachate inlet port, a leachate collection port, a gas sparging port, and a gas collection port. Each port accepted 6.35 mm I.D. tubing. Gas was collected in flex-foil gas bags (SKC Corp., Houston, TX, USA) fitted with a hose valve. Leachate was collected in 2-L intravenous bags (Baxter Healthcare, Deerfield, IL, USA). Assembled reactors together with gas bags, leachate bags and gas connections were tested for leaks using a vacuum pump. The reactors were operated as stationary solid phase beds with leachate recirculated in down flow mode. In control reactors, 500 dry gm of shredded newsprint was added to each reactor in the absence of a sulfur-containing waste. In the treatment reactors, 500 dry gm of shredded newsprint and 125 dry gm of ground sulfur-containing waste were added to each reactor. One liter of leachate was added to each reactor, including the controls, as an inoculum. Leachate was obtained from a 300-L reactor that contained actively decomposing residential MSW. The
Table 1 Sulfur-containing wastes tested in laboratory-scale reactors.
a b
Sulfurcontaining waste
NCSU ID
Waste generation process description
Sulfate content a (% SO2 4 –S)
Sulfide content (% S)
Theoretical H2S yield from stoichiometric sulfate reduction (mL H2S/gm dry waste)
Fly Ash C&D fines
12–307 11–200
Fly Ash only without flue gas desulfurization (FGD) residue C&D fines that contain gypsum (CaSO4)
0.38 (0.02) 7.99 (0.15)
Not detected Not detected
2.68 55.93
Trona b Ash A Trona b Ash B Lime Ash A Lime Ash B
12–280 12–393 12–392 13–03
Fly Ash residue from wet scrubbers, dry sorbent injection, or spray dryer treatment in which trona or lime was used to neutralize SOx
3.64 1.07 0.96 2.28
(0.74) (0.01) (0.01) (0.31)
7.48 (0.96) Not detected Not detected 6.94 (0.68)
77.89 7.49 6.70 64.53
MSW Ash A MSW Ash B
12–365 13–199
Mixed fly ash and bottom ash from MSW combustion
2.00 (0.02) 1.55 (0.01)
Not detected Not detected
14.00 10.85
The results were adopted from Sun and Barlaz, 2015. Standard deviations are given parenthetically. Trona is a natural sodium carbonate mineral.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
3
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
presence of sulfate-reducing activity in the leachate inoculum was verified in preliminary work. Sufficient deionized (DI) water was added initially to ensure the generation of leachate for recirculation and then added periodically to maintain 500–600 mL leachate. Leachate was neutralized to 7.4–7.6 as necessary and recirculated about three times a week throughout the incubation period. Duplicate reactors were continuously sparged with humidified N2 (i.e., bubbled through water); while there was no sparging in a parallel set of two reactors. The N2 flow rate was set at 5 mL min1 and controlled by using a borosilicate glass tube flow meter (VWR, Radnor, PA, USA). This rate was designed to flush each reactor 3 times per day based on an estimate of the reactor pore space. Gas was added to a diffuser stone placed at the bottom of each reactor and a layer of Ottawa sand was placed above the stone to distribute the gas. Ottawa sand was also added above the newsprint/ sulfurcontaining waste mixture in each reactor to reduce the overall reactor pore space and subsequently the required N2 flow rate. All gas produced or sparged through the reactors was collected in 40-L flex-foil gas bags, and both total volume and gas composition were measured once or twice a week as needed. 2.4. Analytical methods The sulfate content was measured by a 5-day acid extraction at room temperature as described previously (Sun and Barlaz, 2015). Briefly, 10 g of sample were extracted in 1.95-L of deionized water and 50-mL of 10 N HCl in a 2-L plastic bottle that was shaken at 2 100 rpm for 5 days. Sulfate (SO2 4 ) and sulfite (SO3 ) were analyzed by suppressed conductivity ion chromatography (IC) using a Dionex IC-2500 system (Sunnyvale, CA, USA) fitted with a Dionex IonPac AS19 analytical column (4 250 mm) and an AG18 guard column (4 50 mm). The eluent was 20 mM KOH. For reporting, sulfate and sulfite were combined and expressed as sulfate-S. Sulfide in the liquid phase was analyzed by iodometric titration (U.S. EPA, 1996). An acid extraction method modified from U. S. EPA method 9030B (U.S. EPA, 1996) was used to measure the solid phase sulfide content (Sun et al., 2015). The newsprint was characterized by measurement of its cellulose, hemicellulose, and lignin content as described previously (Wang et al., 2015). Briefly, the cellulose and hemicellulose were converted to their monomeric sugars using a two-stage sulfuric acid hydrolysis. The monomeric sugars were then quantified using a high performance liquid chromatograph (Shimadzu, Columbia, Maryland, USA) equipped with an electrochemical detector. Klason lignin was determined by loss on ignition (2 h, 550 °C) of the solids remaining after 72% sulfuric acid hydrolysis. The biochemical methane potential (BMP) of newsprint was measured by incubation of a known mass of ground sample at 37 °C in triplicate 160 mL serum bottles that included anaerobic growth medium and an anaerobic consortium maintained and acclimated to the conversion of MSW to methane (Wang et al., 1997). BMPs were corrected for methane attributable to the inoculum. The methods employed for analyses of CH4 gas concentration and volume have been presented previously (Wang et al., 2015). Gaseous H2S was measured by using a Jerome H2S analyzer (Model Jerome J645) manufactured by Arizona Instrument LLC. (Chandler, AZ, USA). The Jerome H2S analyzer is an industrially calibrated ambient air analyzer with a range of 3 ppb (ppb) to 5 ppm (ppm). Gaseous H2S samples were prepared by using serial dilution; 1–10 mL of gas was removed from the reactor headspace using a gas tight glass syringe, and diluted in a 1-L flexfoil gas bag (SKC Corp., Houston, TX, USA) filled with N2 to a total volume was 1 L. Further dilution was repeated as needed to achieve a final H2S concentration in the range of 3 ppb to 5 ppm. In preliminary work, the accuracy of the instrument was confirmed by preparing dilutions from an H2S standard gas. The ratio of the measured to
theoretical H2S concentration ranged from 91.2% to 115% for samples between 0.025 and 5.35 ppm (Table S1 of the Supporting Information). 2.5. Data analysis All gas data were corrected to dry gas at 0 °C and 1 atm (STP). The H2S yield was corrected for H2S measured in the control reactors and attributable to the inoculum. Data from the sparged reactors were used to calculate a decay rate constant for each sulfur-containing waste by adopting the method described in De la Cruz and Barlaz (2010). To begin, it was assumed that H2S production could be estimated by using a first order reaction. A first order equation, such as the U.S. EPA’s Landfill Gas Emissions Model (LandGEM), is commonly used to represent the biological production of methane. For this study, the same equation was used to describe H2S production in a batch reactor in which all waste is disposed in one increment (Eq. (3)) (U.S. EPA, 2005).
Q ¼ ks L0 M ekt
ð3Þ 3
where Q is the rate of gas production (H2S or CH4) (m /yr); k is the first-order decay rate constant (H2S or CH4) (yr1); L0 is ultimate gas production (H2S or CH4) (m3/Mg); M is the mass of sulfate or MSW disposed (Mg/yr); and t is time (yr). Note that k in Eq. (3) is equivalent to k in the IPCC landfill gas model (IPCC, 2006). Eq. (3) was converted to a linear form (Eq. (4)) to calculate laboratory-scale first order decay rate constants for both H2S (kslab) and CH4 (kCH4-lab).
lnðY 0 Y t Þ ¼ klab t þ lnðY 0 Þ
ð4Þ
where Y0 is the ultimate yield (L0 for CH4 and Ls0 for H2S), Yt is the yield (H2S or CH4) at any time t during the reactor monitoring period, and klab (kslab or kCH4-lab) is the lab-scale first order decay rate constant. The cumulative H2S yield (Ls0) was calculated from the measured H2S yield in the sparged reactors and includes sulfide from both sulfate reduction and that released from the initial solid sample (in samples where solid phase sulfide was present). Time course data between 1% and 99% of the measured ultimate yield were used for the regression in Eq. (4). The square of the correlation coefficient (r2) for all linear regressions ranged from 0.85 to 1.0, which suggests that representing the data with a first order decay model was appropriate. 3. Results and discussion 3.1. Materials characterization The sulfur content of the wastes tested is presented in Table 1. Two of eight tested samples were shown to have non-sulfate sulfur in previous work and as a result, these samples were analyzed for solid phase sulfide by acid extraction (Sun et al., 2015) (Table 1). The volatile solids, cellulose, hemicellulose, and lignin contents of the ground newsprint were 95.3, 46.5, 16.3, and 17.1% of dry weight, respectively. The BMP was 125.7 mL CH4/dry gm newsprint. 3.2. Hydrogen sulfide and methane production in landfill simulation reactors The ultimate H2S yield of each waste is summarized in Table 2, and the H2S concentration and yield with time are presented in Figs. 1 and 2, respectively. H2S yields in the sparged reactors were significantly higher (p < 0.05) than those in the non-sparged reactors for all eight materials tested, which indicates that sparging enhanced the conversion of sulfate to sulfide for all wastes. While
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
4
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Table 2 H2S yield and recovery in laboratory-scale reactors. Treatment
Fly Ash C&D Fines Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
H2S yielda (mL H2S/gm dry sulfur-containing waste)
H2S recoveryb (%)
Sparged
Non-sparged
Sparged
Non-sparged
2.8 (0.2)c 22.3 (0.6) 30.9 (1.1) 1.0 (0.2) 0.6 (0.1) 16.1 (0.4) 1.6 (0.2) 2.5 (0.4)
2.4 8.6 1.9 0.1 0.2 3.7 0.3 0.2
102.9 (6.0) 39.8 (1.1) 121.2 (4.2)d 13.3 (3.1) 9.2 (0.8) 103.2 (2.6)d 11.3 (1.7) 22.6 (3.5)
88.1 (15.4) 15.5 (1.6) 7.6 (1.1) 1.8 (0.2) 2.7 (0.2) 23.5 (6.5) 2.0 (1.1) 1.7 (0.9)
(0.4) (0.9) (0.3) (0.0) (0.0) (1.0) (0.2) (0.1)
a H2S yield was corrected for background H2S in the control (newsprint only) reactors which ranged from 0.0002 to 0.02 mL H2S/gm newsprint. H2S concentrations in the controls were about 20 ppmv. The reported yield is based on the period before sparging was initiated in some of the non-sparged reactors as described in the text. b H2S recovery is the fraction of the initial sulfate recovered as H2S. c Standard deviations are given parenthetically. d If it is assumed that 100% of the initial solid phase sulfide was released to the gas phase, then the H2S recovery is 48.3 and 40% in Trona Ash A and Lime Ash B, respectively.
controlled toxicity tests were not conducted, the H2S concentration was as high as 7% (70,000 ppm) in the non-sparged reactors and orders of magnitude higher than in the sparged reactors, suggesting that H2S (i.e., end product) inhibition was responsible for the reduced H2S yields and reduced sulfate conversion in the nonsparged reactors. Other possible explanations for the results include additional sulfide precipitation in the non-sparged reactors, and endproduct inhibition due to CH4 and CO2, the concentrations of which would have been lower in the sparged reactors relative to the non-sparged treatments. As solid phase sulfide could not be quantified, the effect of sulfide precipitation cannot be examined. Both CH4 and CO2 routinely accumulate in standard BMP tests which would seem to preclude endproduct inhibition attributable these compounds (Wang et al., 1994). The effect of sparging was minimal in the reactors containing fly ash as calculated sulfate conversion was near 100% in both treatments (Table 1, Fig. 2A). Notably, the H2S concentration only remained elevated for a relatively short time (Fig. 1A) in the fly ash reactors. Methane yields are summarized in Table 3 and cumulative yield curves are presented in Fig. 3. A separate set of control reactors containing newsprint and inoculum was initiated each time that a set of reactors with a sulfur-containing waste was initiated which is why results for three sets of control reactors are presented. As noted in the Methods, all reactor sets were monitored until H2S production was complete so that a recovery and decay rate constant could be calculated. However, reactors in Sets B and C were terminated before methane production was complete. With the exception of Lime Ash A (Fig. 3G), methane yields were significantly higher in the sparged reactors relative to the nonsparged reactors (p < 0.05), which is consistent with the inhibition of H2S production inhibition in the non-sparged reactors. The explanation for the absence of an effect attributable to sparging in the Lime Ash A reactors is unclear though the behavior in these reactors was unique. H2S recovery in these reactors was only 9.2% and 2.7% in the sparged and non-sparged reactors, respectively, and there was little sulfate-reduction activity as evidenced by the low gaseous H2S concentrations and yield (Figs. 1G and 2G). Interestingly, the sparged control reactors produced significantly more methane than the non-sparged controls in each set despite the absence of sulfate in the controls (Table 3). Thus, although the methane yield in the Fly Ash reactors was higher in the sparged treatment, the ratio of the methane yields in the
sparged and non-sparged treatments in the Fly Ash reactors was comparable to that in the controls (Table 3), so no effect can be attributed to sparging. H2S concentrations of up to 30,000 ppm were present in the non-sparged Fly Ash reactors on day 75 (Fig. 1A) at which time methane production was equivalent in the sparged and non-sparged Fly Ash reactors (Fig. 3A). Sulfate conversion was essentially complete in the Fly Ash reactors within 50–100 days (Fig. 2A) while methane production continued for about 300 days. Once the effect of sparging was established, selected nonsparged reactors were sparged with N2 to evaluate whether either the one time or continuous removal of H2S (and other intermediated and endproducts) would stimulate sulfate reduction and/or methane production. For example, the duplicate Trona Ash A non-sparged reactors were sparged one time on days 100 and 175, respectively. While there was no substantial effect on H2S production (Fig. 2E), methane generation was stimulated after the one-time sparging (Fig. 3C) although the cumulative methane yield in the continuously sparged reactor treatment was still higher. Continuous N2 sparging was initiated in Lime Ash A and B and Trona Ash B after 80–110 days of operation. Both H2S and CH4 generation were stimulated in Lime Ash B and Trona Ash B but not in Lime Ash A (Figs. 2 and 3). The absence of an effect for Lime Ash A could be attributed to the fact that H2S accumulations were relatively low (Fig. 1G) so the effect of sparging on H2S concentrations was not as meaningful. This contrasts with Lime Ash B where sparging significantly reduced H2S concentrations (Fig. 1H). The explanation for the behavior of Trona Ash B reactors is not clear as continuous sparging of the previously non-sparged reactors stimulated both H2S (Fig. 2F) and CH4 (Fig. 3F) production despite the low accumulation of H2S in these reactors (Fig. 1F). Despite some inconsistencies, it appears that H2S toxicity on sulfate reduction and methane production is reversible if an appropriate strategy is applied to alleviate the accumulated H2S. 3.3. H2S recovery The fraction of the initial sulfate that was recovered as H2S is presented in Table 2. The recovery ranged from 9.2% to 121.2% in the sparged reactors and was lower in the non-sparged reactors. The calculated recovery is greater than 100% for Fly Ash, Trona Ash A and Lime Ash B. While the average Fly Ash recovery is slightly above 100%, this is likely a result of measurement error and the standard deviation of the averages suggests that the calculated average recovery of 102.9% is not different from 100%. In the case of Trona Ash A and Lime Ash B, these samples contained 7.5% and 6.9% solid phase sulfide, respectively. Thus, the likely explanation for recoveries of greater than 100% is that some or all of the initial sulfide was liberated during the decomposition experiment. If the theoretical H2S recovery is calculated as the measured H2S divided by the sum of the stoichiometric conversion of the initial sulfate to gaseous H2S plus the release of the initial solid phase sulfide to the gas phase, then H2S recoveries for Trona Ash A and Lime Ash B would be 48.3 ± 1.7% and 40.0 ± 1.0%, respectively. Although the recoveries are higher in the sparged treatments relative to the non-sparged treatments, the overall conversions are relatively low. Once the correction for initial solid phase sulfide is considered, the recoveries range from 9.2% to 48.3% for the samples tested, excluding the fly ash as discussed above. Two samples of Trona Ash were tested, and both their initial characterization and their biotransformation behavior were different. Trona Ash A had more sulfate but also contained solid phase sulfide. While Trona Ash A exhibited a high recovery, when the recovery is recalculated assuming that all of the solid phase sulfide
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
5
Fig. 1. H2S concentration in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors.)
was released as gaseous H2S, the recovery is only 48.3%, still well above the 13.3% conversion recorded for Trona Ash B. The effect of one time sparging on H2S production in the non-sparged reac-
tors was minimal for Trona Ash A (Fig. 2E) but significant for Trona Ash B (Fig. 2F), even though its overall recovery was still relatively low.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
6
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Fig. 2. H2S yield in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors; SW stands for sulfur-containing waste.)
Two samples of Lime Ash were tested, and their initial characterization and biotransformation behavior were also different.
Lime Ash B had more sulfate but also contained solid phase sulfide. Sulfate recoveries in Lime Ashes A and B, after correction for the
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
W. Sun et al. / Waste Management xxx (2016) xxx–xxx Table 3 CH4 production in laboratory-scale reactors.a Treatment
CH4 yield (mL CH4/gm dry newsprint) Sparged
Non-sparged
Control reactor (Set A) Fly Ash Trona Ash A C&D fines
165.1 163.4 155.2 151.6
120.2 (4.7) 116.8 (0.3) 1.8 (0.1) 69.7 (15.6)
Control reactor (Set B) Trona Ash B Lime Ash A Lime Ash B
86.3 72.7 77.3 80.5
Control reactor (Set C) MSW Ash A MSW Ash B
84.4 (2.1) 81.4 (3.6) 84.8 (0.7)
(3.4) (1.1) (1.0) (17.1)
(0.6) (4.1) (2.0) (1.4)
67.5 (4.0) 25.9 (12.6) 71.9 (6.7) 6.1 (0.8) 70.1 (1.9) 27.6 (3.1) 22.9 (3.8)
a Three sets of control reactors are presented because a new set of control reactors was set up each time a new set of sulfur-containing wastes was tested. Standard deviations are given parenthetically. In Sets 2 and 3, reactors were terminated before methane production was complete which is why the control reactor methane yields are lower in Control Sets B and C.
initial sulfide in Lime Ash B, were 9.2% and 40%, respectively. Continuous sparging had a significant effect on H2S (Fig. 2H) and CH4 (Fig. 3H) generation in Lime Ash B but no effect on CH4 generation in Lime Ash A (Fig. 2G and 3G). In the C&D fines reactors, H2S accumulated within about 14 days of reactor initiation (Fig. 1B), and the concentrations remained elevated (up to 30,000–35,000 ppm) for almost 200 days which appeared to impact both CH4 (Fig. 3D) and H2S production (Fig. 2B). While both CH4 and H2S production were stimulated by sparging and the resulting shorter duration of H2S accumulation in the sparged reactors (about 100 days), overall H2S recovery was still relatively low even in the sparged reactors compared to the maximum H2S potential calculated from the conversion of all the sulfate in the C&D fine sample. The low H2S yield in the sparged C&D fines reactors is surprising given the solubility of gypsum. In previous work, Yang et al. (2006) reported H2S generation was reduced by the presence of wood and concrete as were present in our samples. Yang et al. explained that the presence of concrete could have increased the system pH, which would inhibit biological sulfate reduction, and that concrete also acted as a sink for the produced sulfide. While concrete was undoubtedly present in the C&D fines tested here, it was not quantified. If pH was important in this research, then the pH effects would have had to have been very localized as the reactor leachate pH was between 7.5 and 7.9 in the C&D fines reactors and comparable to the pH in all other reactors (data not shown). Sulfate recovery in both MSW Ashes was relatively low. While sparging stimulated both CH4 (Fig. 3J and K) and H2S (Fig. 2C and D) production, measured H2S concentrations never exceeded 9000 ppm (Fig. 1C and D) in any of the reactors, a concentration well below those observed in the Fly Ash reactors which had much higher recovery. The effect of the elevated H2S concentrations on sulfate conversion and methane generation appeared to vary between wastes. For example, among the sparged reactors, H2S concentrations were highest in Trona Ash A and Lime Ash B (Fig. 1E and H), yet these two materials had among the highest sulfate recoveries, even after consideration of the impact of the initial sulfide. Similarly, methane production in the sparged reactors containing these wastes is comparable to that in the controls (Fig. 3A, C, E and H). Thus, general statements about the extent of sulfate conversion based on the initial sulfate content are not possible for the wastes tested and there may be factors not identified in this study that also affect sulfate conversion.
7
The measured H2S yields are based on H2S recovery in the reactor gas phase. Aqueous sulfide was minimal in the final leachate sample from the sparged reactors. Although solid phase sulfide was measured in the residual solids at the end of the monitoring period, the results were so variable that they were not useful and hence not reported. Nonetheless, qualitatively, some solid phase sulfide was present. It is possible that the extent of sulfide precipitation varied by waste which would influence gaseous H2S recovery. The presence of metals in the ash samples may also exert toxicity. Metals concentrations in the samples are presented in Table 4. Interestingly, the Trona Ash B and Lime Ash A had the highest Fe concentrations which might suggest that the presence of Fe contributed to the low measured gaseous H2S yield and corresponding low H2S recoveries due to FeS(s) formation. However, when H2S recovery was plotted against the individually measured metals concentrations for all wastes, there were no apparent correlations. 3.4. Estimated H2S decay rate in laboratory and field scales The reactor data were used to estimate laboratory-scale decay rate constants as described in the Methods. The laboratory decay rate constants were then used to estimate a decay rate for H2S that is applicable at field-scale by assuming that the ratio of the decay rate constant of newsprint conversion to methane and that of a sulfur-containing waste to H2S is constant between the lab and the field as expressed by Eq. (5).
CH4 decay rate of newsprint at lab-scale H2 S decay rate at lab-scale CH4 decay rate of newsprint at field-scale ¼ H2 S decay rate at field-scale
ð5Þ
Newsprint was used because it was the source of organic matter in the reactors and its methane production rate constant was measured in parallel with H2S production from the sulfurcontaining wastes (Table 5). The laboratory newsprint decay rate in the reactors of Control Set A, Fly Ash, Trona Ash A and C&D fines was 5.6 yr1 (range 5.4–5.7 yr1) and this average was used in Eq. (5). With reference to Eq. (5), the decay rate of newsprint at field-scale (0.033 yr1) was adopted from a published estimate. This field-scale newsprint decay rate was estimated from laboratory data and an algorithm to scale decay rates measured in laboratory reactors to a landfill as described previously (De la Cruz and Barlaz, 2010). This newsprint decay rate corresponds to a landfill with an MSW decay rate of 0.04 yr1. Using Eq. (5), estimates for decay rates (ksfield) for the tested sulfurcontaining wastes are presented in Table 5 and range from 0.11 to 0.38 yr1. The field-scale H2S decay rate constants presented in Table 5 are based on the field-scale CH4 decay rate constant for newsprint, which, as described above, was based on a landfill with a bulk MSW decay rate constant of 0.04 yr1. In previous research on U. S. landfills, it has been reported that a field-scale MSW methane decay rate constant of 0.09–0.12 yr1 may be more appropriate (Wang et al., 2013). If a field-scale bulk MSW decay rate of 0.1 yr1 were assumed, then the field-scale CH4 decay rate constant for newsprint and the calculated ksfield values given in Table 5 would increase by a factor of 2.5. The only published estimate of field-scale rate constants of H2S is that presented in Anderson et al. (2010) who estimated a decay rate constant of 0.7 yr1 for landfills that used C&D fines as daily cover. There are, however, so many differences between the values in Anderson et al. (2010) and the value estimated here that a direct comparison is flawed. To estimate a decay rate from field data, Anderson et al. (2010) allowed ksfield and the
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
8
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Fig. 3. CH4 yield in the sparged and non-sparged reactors for sulfur-containing wastes included in this study. (Legend: the filled symbols represent sparged reactors; the empty symbols represent non-sparged reactors; NP stands for newsprint.)
ultimate sulfide generation potential (Ls0) to vary simultaneously to get the best fit between the model and measured data. For the estimates developed here, Ls0 was treated as an intrinsic property of the waste and was not allowed to vary when calculating kslab. In addition, Anderson et al. (2010) had minimal data
on the actual H2S generation rate and did not consider a timevarying collection efficiency, which is a confounding factor in working with data from field-scale landfills. Finally, there was uncertainty in both the mass of sulfur actually buried over time and the mass of H2S recovered through the gas collection and
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
9
W. Sun et al. / Waste Management xxx (2016) xxx–xxx Table 5 Estimation of field-scale H2S decay rate constants for sulfur-containing wastes.a Treatment
H2S decay rate constant (Lab scale) (yr1)
CH4 decay rate constant (Lab scale) (yr1)
H2S decay rate constant (Field scale) (yr1)
Control reactor Fly Ash C&D fines Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
N/Ab 30.8 (1.3) 18.4 (6.7) 25.6 (0.2) 35.5 (1.9) 39.0 (0.1) 37.3 (2.4) 61.4 (0.6) 64.0 (0.2)
5.6 5.7 5.4 5.5 5.6 5.6 5.6 5.6 5.6
N/A 0.18 0.11 0.15 0.21 0.23 0.22 0.36 0.38
a b
(0.3) (0.1) (0.2) (0.2) (0.1) (0.1) (0.1) (0.1) (0.1)
(0.01) (0.04) (0.01) (0.01) (0.00) (0.01) (0.00) (0.00)
Standard deviations are given parenthetically. N/A means not applicable.
control system in Anderson et al. The estimated decay rate constant for C&D fines derived in this study was 0.11 yr1 and this would increase to 0.27 yr1 at a bulk MSW decay rate constant of 0.1 yr1, which is likely more representative of a wet landfill in the northeastern U.S. While there is considerable uncertainty in the decay rate estimates, it is noteworthy that both the values in Table 5 and the values proposed by Anderson et al. (2010) are consistent in that the laboratory-scale decay rate constant for H2S production from C&D fines (18.4 yr1) is higher than the methane production decay rate constant (5.6 yr1). The implication of this is that H2S production from C&D fines will decrease faster than methane generation from MSW. 4. Summary and conclusions The H2S production potential of a range of sulfur-containing wastes was measured in 8-L reactors that were operated to maximize the rate of decomposition by the use of leachate neutralization and recirculation, the presence of excess organic carbon, inoculation, and incubation at 37 °C. Duplicate reactors were operated with and without sparging to evaluate the effect of endproduct (H2S(g), CH4 and CO2) removal. Sparging was shown to consistently increase both H2S and CH4 yields, suggesting that H2S endproduct accumulation was inhibitory to the biological processes occurring in the reactors. H2S recovery varied and was not correlated with either the initial sulfate or total S in a sample which suggests that chemical analyses are not suitable for prediction of H2S yields. The laboratory-scale H2S production rate constants were used to estimate field-scale rate constants which ranged from 0.11 to 0.38 yr1. These values are generally higher than the parallel values estimated for methane production, suggesting that sulfate will be depleted more rapidly that the cellulosic component of MSW.
Fig. 3 (continued)
Table 4 Metals in the sulfur-containing wastes.a Sulfur-containing waste
Al Wt.%
As mg/kg
Cd mg/kg
Co mg/kg
Cr mg/kg
Cu mg/kg
Fe Wt.%
Ni mg/kg
Pb mg/kg
Sb mg/kg
Se mg/kg
Zn mg/kg
Hg mg/kg
Fly Ash Trona Ash A Trona Ash B Lime Ash A Lime Ash B MSW Ash A MSW Ash B
2.23 0.78 1.35 1.29 0.51 2.63 2.51
68.0 17.1 80.6 100 69.8 13.4 8.99
<5.00 <5.00 <5.00 <5.00 <5.00 54.7 43.7
20.9 9.70 10.8 13.6 5.19 9.37 9.03
53.6 12.6 72.6 48.6 15.5 50.3 62.0
72.9 24.0 31.9 40.1 40.4 1089 1195
1.85 0.35 3.61 4.00 0.97 1.24 2.26
37.3 17.1 34.0 45.3 11.4 35.3 36.7
28.3 15.2 27.6 17.4 8.94 926 718
<5.00 <5.00 7.59 <5.00 5.47 136 40.2
5.93 52.4 <5.00 <5.00 52.4 <5.00 <5.00
47.0 24.4 126 110 24.8 5287 3952
<5.00 <5.00 <5.00 <5.00 <5.00 <5.00 <5.00
a
Most analytes are reported in mg/kg – Al and Fe are reported in weight% (1% by weight is equivalent to 10,000 mg/kg). The minimum detection limit (MDL) is 5 mg/kg on
ICP.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028
10
W. Sun et al. / Waste Management xxx (2016) xxx–xxx
Acknowledgements This work was supported by Waste Management Inc. and the Environmental Research and Education Foundation. We appreciate the analytical support of David Black at North Carolina State University.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.wasman.2016.01. 028.
References Anderson, R., Jambeck, J.R., McCarron, 2010. Modeling of hydrogen sulfide generation from landfills beneficially utilizing processed construction and demolition materials, final report to Environmental Research and Education Foundation. Beauchamp, R.O., Bus, J.S., Popp, J.A., Boreiko, C.J., Andjelkovich, D.A., 1984. A critical review of the literature on hydrogen-sulfide toxicity. Crit. Rev. Toxicol. 13, 25– 97. Cochran, K., Townsend, T., Reinhart, D., Heck, H., 2007. Estimation of regional building-related C&D debris generation and composition: case study for Florida, US. Waste Manage. 27, 921–931. De la Cruz, F.B., Barlaz, M.A., 2010. Estimation of waste component-specific landfill decay rates using laboratory-scale decomposition data. Environ. Sci. Technol. 44, 4722–4728. Eun, S., Reinhart, D., Cooper, D., Townsend, T., Faour, A., 2007. Hydrogen sulfide flux measurements from construction and demolition debris (C&D) landfills. Waste Manage. 27, 220–227. IPCC Intergovernmental Panel on Climate Change, 2006. IPCC Guidelines for National Greenhouse Gas Inventories, Chapter 3; IPCC National Greenhouse Gas Inventory Program; IPCC, WMO: Geneva, Switzerland, vol. 5. Ko, J.H., Xu, Q.Y., Jang, Y.C., 2015. Emissions and control of hydrogen sulfide at landfills: a review. Crit. Rev. Environ. Sci. Tec. 45 (19), 2043–2083. Lee, S., Xu, Q.Y., Booth, M., Townsend, T.G., Chadik, P., Bitton, G., 2006. Reduced sulfur compounds in gas from construction and demolition debris landfills. Waste Manage. 26, 526–533.
McDonald, H.B., Parkin, G.F., 2009. Effect of sulfide inhibition and organic shock loading on anaerobic biofilm reactors treating a low-temperature, high-sulfate wastewater. Water Environ. Res. 81 (3), 265–288. O’Flaherty, V., Mahony, T., O’Kennedy, R., Colleran, E., 1998. Effect of pH on growth kinetics and sulphide toxicity thresholds of a range of methanogenic, syntrophic and sulphate-reducing bacteria. Process Biochem. 33 (5), 555–569. OSHA, 2005. OSHA fact sheet: Hydrogen sulfide (H2S); U.S. Department of Labor. Reis, M.A.M., Almeida, J.S., Lemos, P.C., Carrondo, M.J.T., 1992. Effect of hydrogen sulfide on growth of sulfate reducing bacteria. Biotechnol. Bioeng. 40, 593–600. Selene, C.H., Chou, J., 2003. Hydrogen Sulfide: Human Health Aspects. World Health Organization, Geneva. Sun, W.J., Barlaz, M.A., 2015. Measurement of chemical leaching potential of sulfate from landfill disposed sulfate containing wastes. Waste Manage. 36, 191–196. Sun, M., Sun, W.J., Barlaz, M.A., 2015. A batch assay to measure microbial hydrogen sulfide production from sulfur-containing solid wastes. Sci. Total Environ. http://dx.doi.org/10.1016/j.scitotenv.2016.01.161. Tolaymat, T.M., El Badawy, A.M., Carson, D.A., 2013. Estimate of the decay rate constant of hydrogen sulfide from drywall in a simulated bench-scale study. J. Envrion. Eng. 139, 538–544. U.S. EPA, 1996. Method 9030B:Acid-soluble and acid-insoluble sulfides: distillation (Revision 2). U.S. EPA, 1998. Characterization of building-related construction and demolition debris in the United States. United States Environmental Protection Agency, Municipal and Industrial Solid Waste Division Office of Solid Waste. U. S. EPA, 2005. Landfill Gas Emissions Model (LandGEM) Version 3.02 User’s Guide; U.S. EPA: Washington, D.C. U.S. EPA, 2014. Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures for 2012. Wang, Y.S., Byrd, C.S., Barlaz, M.A., 1994. Anaerobic biodegradability of cellulose and hemicellulose in excavated refuse samples using a biochemical methane potential assay. J. Ind. Microbiol. 13, 147–153. Wang, Y.S., Odle, W.S., Eleazer, W.E., Barlaz, M.A., 1997. Methane potential of food waste and anaerobic toxicity of leachate produced during food waste decomposition. Waste Manage. Res. 15, 149–167. Wang, X., Nagpure, A.S., DeCarolis, J.F., Barlaz, M.A., 2013. Using observed data to improve estimated methane collection from select U.S. landfills. Environ. Sci. Technol. 47 (7), 3251–3257. Wang, X., De la Cruz, F., Ximenes, F., Barlaz, M.A., 2015. Decomposition and carbon storage of selected paper products in laboratory-scale landfills. Sci. Total Environ. 532 (1), 70–79. WHO, 2000. Hydrogen sulfide. Air quality guidelines for europe. Second edition. World Health Organization, Regional Office for Europe, Copenhagen, Denmark. Yang, K., Xu, Q., Townsend, T.G., Chadik, P., Bitton, G., Booth, M., 2006. Hydrogen sulfide generation in simulated construction and demolition debris landfills: impact of waste composition. J. Air Waste Manage. Assoc. 56, 1130–1138.
Please cite this article in press as: Sun, W., et al. Characterizing the biotransformation of sulfur-containing wastes in simulated landfill reactors. Waste Management (2016), http://dx.doi.org/10.1016/j.wasman.2016.01.028