Advanced Thermal Hydrolysis: Optimization of a Novel

Mar 31, 2012 - Journal of Environmental Management 2018 213, 271-278 ... Appraisal of changes in sewage sludge management. M. Vochozka , A...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/est

Advanced Thermal Hydrolysis: Optimization of a Novel Thermochemical Process to Aid Sewage Sludge Treatment Jose Abelleira,*,† Sara I. Pérez-Elvira,‡ Juan R. Portela,† Jezabel Sánchez-Oneto,† and Enrique Nebot§ †

Department of Chemical Engineering and Food Technologies, Faculty of Sciences, University of Cádiz, 11510 Puerto Real (Cádiz), Spain ‡ Department of Chemical Engineering and Environmental Technology, University of Valladolid, 47011 Valladolid, Spain § Department of Environmental Technologies, Faculty of Marine and Environmental Sciences, University of Cádiz, 11510 Puerto Real (Cádiz), Spain S Supporting Information *

ABSTRACT: The aim of this work was to study in depth the behavior and optimization of a novel process, called advanced thermal hydrolysis (ATH), to determine its utility as a pretreatment (sludge solubilization) or postreatment (organic matter removal) for anaerobic digestion (AD) in the sludge line of wastewater treatment plants (WWTPs). ATH is based on a thermal hydrolysis (TH) process plus hydrogen peroxide (H2O2) addition and takes advantage of a peroxidation/direct steam injection synergistic effect. On the basis of the response surface methodology (RSM) and a modified Doehlert design, an empirical second-order polynomial model was developed for the total yield of: (a) disintegration degree [DD (%)] (solubilization), (b) filtration constant [Fc (cm2/min)] (dewaterability), and (c) organic matter removal (%). The variables considered were operation time (t), temperature reached after initial heating (T), and oxidant coefficient (n = oxygensupplied/oxygenstoichiometric). As the model predicts, in the case of the ATH process with high levels of oxidant, it is possible to achieve an organic matter removal of up to 92%, but the conditions required are prohibitive on an industrial scale. ATH operated at optimal conditions (oxygen amount 30% of stoichiometric, 115 °C and 24 min) gave promising results as a pretreatment, with similar solubilization and markedly better dewaterability levels in comparison to those obtained with TH at 170 °C. The empirical validation of the model was satisfactory.



INTRODUCTION In a technical-scientific framework, the conventional wastewater treatment plant (WWTP) concept is subjected to continuous review and improvement. The latest trend is the redefinition of the WWTP structure based on the deployment of emerging technologies and the evaluation of novel operational strategies. A key factor is the flexibility in the concept of the WWTP according to the different scenarios in which they may be implemented: location, size, and destination of the final effluent, among many other aspects. Sewage sludge is the main waste generated in the urban wastewater treatment process and its production is increasing significantly every year. In fact, current data show that for large plants [more than 100 000 person equivalent (p.e.)] up to 50% of the total wastewater treatment costs correspond to sludge handling and disposal.1 In any case, sludge treatment and disposal should be highlighted as an integral part of wastewater treatment. There is a wide range of uses for sludge in which its energy or chemical content is exploited and this means that sludge can be considered as a valuable source of recoverable resources rather than a waste material. Nowadays, the sludge line of a large urban WWTP is commonly designed around the anaerobic digestion (AD) © 2012 American Chemical Society

process. However, it is a reality that secondary sewage sludge is becoming more difficult to degrade in anaerobic digesters and this is due to the implementation of stricter nitrogen limits, longer sludge ages, and the removal of primary sedimentation units.2 On the one hand, the hydrolytic stage of AD is regarded as the rate-limiting step in the degradation of complex organic compounds, such as sewage sludge,3 and the consequent generation of biogas (mainly CH4 and CO2). In this sense, processes such as pulping and thermal hydrolysis have been widely studied4 and have become full scale pretreatments for AD to enhance the hydrolytic limiting stage and to improve the performance of the global process.5,6 Moreover, numerous studies have been carried out on different AD pretreatments that are considered unconventional or even novel. Among these approaches, the following can be highlighted: mechanical pretreatments,7 alkaline thermal hydrolysis,8 acid thermal hydrolysis,9 ultrasound,8,10 ozonation,11,12 microwave irradiReceived: Revised: Accepted: Published: 6158

November 23, 2011 March 30, 2012 March 31, 2012 March 31, 2012 dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

ation,13,14 and Fenton’s peroxidation.15 In this field of study, recent investigations concerning the combination of hydrogen peroxide (H2O2) addition and thermal treatment (without steam injection)16−19 or microwave irradiation20 can also be found in the literature. On the other hand, the digestate (stabilized sludge) from the AD needs further treatment to facilitate its management or to eliminate the excess organic matter depending on the final disposal of the sludge and the requirements of the WWTP location. In this sense, postreatment processes such as thickening, dewatering (by, for instance, belt filters, filter press or centrifuge), heat drying, and/or incineration are considered as conventional and commercially applicable examples.21 In almost all cases, dewatering of the AD digestate is required, but this process is very expensive and has limited efficiency.22,23 Therefore, new solutions are required to aid the removal of more water from the digestate while reducing the downstream costs. Moreover, despite significant environmental, capital, and operational expenses, incineration has become a widespread strategy for sludge treatment as the volume of the final product (mainly ash) is around 30% of the initial solids content of sewage sludge.24 In fact, because land application is more and more problematical due to rigorous regulations concerning the tolerated compositions, coincineration is attracting increasing attention where permits can be obtained.25 Overall, anaerobic digestion is commonly considered to be the optimal treatment technology, whereas incineration, particularly if coal-fired, is one of the most environmentally and economically costly.21 Hence, in recent years increasing effort has been focused on the study and implementation of alternative AD postreatment processes that are intended to be much more environmentally friendly than incineration. These approaches include, for instance, (catalytic) wet (air) oxidation (CWAO, WAO, WO),26−28 supercritical water oxidation (SCWO),29,30 and Fenton’s peroxidation,31 all of which are considered as advanced oxidation processes (AOPs). Within this framework of innovation, and for the case of the WWTP sludge line, a modification of the thermal hydrolysis (TH) process was reported by Abelleira et al.32 This novel process, which is called advanced thermal hydrolysis (ATH), exploits the synergistic effect of peroxidation plus direct steam injections, without using catalysts, and is intended to operate under conditions that are less severe than those in other pre- or postreatments. In our previous work, discrete TH and ATH assays were conducted to compare the performance of the system through the examination and corroboration of the following hypotheses: 1 Secondary sludge processed with ATH could exhibit higher dewaterability than that subjected to TH. 2 When low doses of oxidant are added, solubilization of secondary sewage sludge organic matter may be higher in the ATH process than in TH, without removing total organic matter significantly. 3 When moderate doses of oxidant are added, ATH could be an effective process for the partial removal of organic matter present in secondary sewage sludge, even at a pH ∼ 6. The aim of the present work was to study the behavior and performance optimization of the ATH process over a whole experimental domain to determine whether its attributes are suitable for pretreatment (sludge solubilization) and/or postreatment (organic matter removal) of the AD. The way

in which the ATH process affects the sludge dewaterability was also monitored. For this, a set of ATH tests based on the design of experiment (DOE) methodology was carried out to study the influence of the main operating variables (amount of oxidant, temperature, and time) through the development of empirical quadratic polynomial models.



MATERIAL AND METHODS

Materials. The feed in this study was secondary sewage sludge from the municipal WWTP of Valladolid (Spain). The sludge was partially thickened and its average characteristics were as follows: total solids (TS) = 50 g/L, total chemical oxygen demand (TCOD) = 50 315 mgO2/L, soluble chemical oxygen demand (SCOD) = 2 396 mgO2/L, total organic carbon (TOC) = 13 348 mgC/L, and pH 6.12. To prevent putrefaction, the sludge was stored at 4 °C and it was attemperated before every experiment. The oxidizing agent used in the ATH experiments was hydrogen peroxide (PRS Panreac, H2O2, 33% w/v, aqueous solution). System and Procedure. Experiments were performed in a thermal hydrolysis laboratory prototype composed of a steam generator, a 1.5 L stainless steel pressure vessel and a 6 L flash tank. In the case of ATH tests, the respective dose of H2O2 was added at once into the reactor just after introducing the sludge and, immediately afterward, the initial steam injection heating took place until the temperature set point was reached. The increase in temperature only needed a maximum of 1−2 min. The advantage of adding the H2O2 just before injecting steam is that during the heating period both hydrogen peroxide and steam were already acting synergistically. If required, additional slight steam injections were performed to maintain a steady set point temperature during the operation time of the experiments. Direct steam injections from the boiler also acted as the reactor stirrer. At the end of each experiment, the reactor content was depressurized moderately into the flash tank. Further details on the experimental system and procedure can be found in our previous publication.32 Analytical Methods. A set of analyses was carried out in order to characterize both influent and liquid effluent [TCOD, SCOD, TOC, time-to-filter (TTF), and constant-pressure filtration tests]. All of the analyses (excepting constant-pressure filtrations tests) were carried out according to standard methods for the examination of water and wastewater.33 TOC (total amount of organic carbon present in a sample) was determined by a combustion-infrared method using a Shimadzu TOC−SM5000A instrument, which allows the TOC determination of sludge samples without pretreating or diluting them (besides, the homogeneity of all the samples was assured by gentle stirring). Dissolved organic carbon (DOC) analysis was not conducted in this work. TCOD and SCOD were assessed by the closed reflux colorimetric method. SCOD was measured after centrifuging the sample (5000 rpm for 5 min) and filtering the supernatant through a glass fiber filter (AP40). The sewage sludge disintegration degree (DD) was calculated using eq 1; the fraction of particulate substances solubilized by the process with respect to the initial insoluble fraction could be determined directly from this value.34 DD(%) = 6159

SCOD − SCOD0 × 100 TCOD0 − SCOD0

(1)

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

Table 1. Experimental Schedule Based on a Modified Doehlert Matrix and Empirical Results Experimental Factors expt. no.

run order

oxidant coefficient n = U1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

8 6 23 5 17 21 20 2 13 18 3 19 10 12 15 16 24 4 11 1 7 14 22 9

0.9 0.9 0.9 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.1 0.1 0.1 0 0 0 0 0 0 0 0

Coded Variables

temperature (°C) T = U2

time (min) t = U3

X1

X2

138 138 99 164 164 125 125 125 125 125 125 86 86 151 112 112 170 164 151 125 138 112 99 86

34 21 28 34 21 40 28 28 28 28 15 34 21 28 34 21 15 34 28 28 34 34 28 21

0.8 0.8 0.8 0 0 0 0 0 0 0 0 0 0 −0.8 −0.8 −0.8 −1 −1 −1 −1 −1 −1 −1 −1

0.289 0.289 −0.578 0.867 0.867 0 0 0 0 0 0 −0.867 −0.867 0.578 −0.289 −0.289 1 0.867 0.578 0 0.289 −0.289 −0.578 −0.867

where SCOD is the soluble COD of the effluent sludge, SCOD0 is the soluble COD of the raw sludge, and TCOD0 is the total COD of the raw sludge. The removal of organic matter was studied by means of the reduction percentage of total chemical oxygen demand (TCOD) and total organic carbon (TOC). The sludge dewaterability was assessed by means of time-to-filter (TTF) and constant-pressure filtration tests. The basis and procedure for both tests, as well as the mathematical derivation of the filtration constant (Fc), are included in the Supporting Information. Experimental Plan: Design of Experiments (DOE). The basis of this work is a set of secondary sludge ATH assays, for which the experimental domain and results were established and calculated respectively by means of DOE methodology. The relationship between factors and responses was established by empirical mathematical models based on the response surface methodology (RSM) and developed using a modified Doehlert design and the multiple linear regression (MLR) method. (Pages S3-S5 of the Supporting Information for more details on DOE methodology). A priori, and on the basis of a literature survey on TH and AOPs,26,35−38 three factors were considered as significant in the ATH process: oxidant coefficient (n = U1), temperature reached after initial heating (T = U2), and operation time (t = U3). According to eq 2, the oxidant coefficient (n) is the amount of oxidant theoretically necessary to oxidize partially (n < 1) or totally (n = 1) the raw sludge organic matter; (n = 0 for TH):

n=

O2 supplied O2 stoichiometric

Responses

X3

disintegration degree Y1 (%)

filtration constant Y2 (cm2/min)

TCOD removal Y3 (%)

TOC removal Y4 (%)

0.52 −0.52 0.04 0.52 −0.52 1 0.04 0.04 0.04 0.04 −1 0.52 −0.52 0.04 0.52 −0.52 −1 0.52 0.04 0.04 0.52 0.52 0.04 −0.52

29.49 31.45 60.05 42.21 44.23 45.06 44.62 43.56 44.53 41.15 44.15 47.44 44.71 44.57 26.63 24.21 41.12 39.54 31.21 21.62 24.89 19.04 16.74 15.50

1.2801 1.7072 1.2812 1.6354 1.8957 1.4993 1.2985 1.5092 1.1688 1.1460 1.2000 0.0034 0.0092 0.6427 0.0069 0.0083 0.0313 0.0357 0.0224 0.0050 0.0205 0.0010 0.0014 0.0003

58.2 56.7 15.0 40.6 28.3 37.7 35.1 33.7 33.2 28.7 21.5 13.0 10.0 7.20 4.10 8.40 4.30 7.80 7.40 5.30 6.00 5.50 5.70 6.10

47.9 40.3 12.1 48.1 34.8 38.8 32.1 34.0 29.0 29.0 22.4 12.0 9.70 6.34 3.63 6.50 4.40 7.10 7.60 5.00 6.00 5.40 5.30 5.80

In this way, and according to the Doehlert design, the experimental domain was defined as a sphere around the following central point: n = 0.5, T = 125 °C, t = 28 min. A range of variation was assigned for each variable: 0 < n < 1, 80 °C < T < 170 °C, 15 min < t < 40 min. The operation pressure was the equilibrium pressure corresponding to saturated steam inside the reactor. Operation time was fixed at around 30 min because, according to previous experience on TH, this was the optimum value.39−41 The explored range of T includes temperatures equal to or below the optimum for TH (170 °C), as determined in previous studies.39 It is worth noting that an excess of oxygen (n > 1) was not added in the ATH tests. According to the Doehlert design, the total number of experiments to be conducted, N, for a given number of factors, k, is: N = k2 + k + C

(3)

where C is the number of experiments corresponding to the central point. Doehlert’s experimental matrix and the corresponding experimental conditions are shown in Table 1 together with the experimental results for each studied response. In this work k = 3 and C = 4; hence, the original matrix is comprised of 16 ATH experiments (Expt. Nos. [1− 16]). Replications in the central point (Expt. Nos. [7−10]) were used to determine the variance of the experimental error and the reproducibility of the data. In addition, it is important to highlight that for the defined experimental domain, and according to the range of variation assigned to n, the quadratic model will predict results for responses even in the case of TH (i.e., n = 0). Therefore, the inclusion of TH assays in the experimental schedule (Expt. Nos. [17−24]) was beneficial to

(2) 6160

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

Figure 1. Contour plot (a) and response surface plot (b) for disintegration degree (DD) yield. Both plots show DD vs oxidant coefficient (n) and temperature (T), for a constant operation time (t = 30 min).

improve the prediction accuracy. This is the main reason why we refer to a modified Doehlert design. The software Modde 9.0 (Umetrics), which was employed in this study to implement the DOE−RSM methodology, easily allows the inclusion and exclusion of experiments with respect to the original experimental plan. ATH experiments were conducted in random order to avoid the effect of external variability. TH assays (excepting Expt. No. 17) were carried out just before the execution of seven of the ATH tests (Expt. Nos. 1, 3, 4, 8, 13, 14, 15), operating under the same conditions but without the addition of H2O2. The maximum mean experimental pressure (8.3 bar) was registered for Expt. No. 4 at t = 34 min, T = 164 °C, and n = 0.5. The experimental results of four responses (Table 1) were analyzed by RSM: sludge solubilization in terms of DD (Y1), sludge dewaterability in terms of the filtration constant (Y2), TCOD removal (Y3), and TOC removal (Y4).



Y1(DD) = 43.70 + 6.27X1 − 4.71X 2 + 0.02X3 − 13.97X12 + 3.76X 22 − 1.49X32 − 20.83X1X 2 + 0.29X1X3 − 0.43X 2X3

(4)

where X1, X2, and X3 represent the values (in coded units) of n, T, and t, respectively. Analysis of variance (ANOVA) showed that the model is statistically good (probability for the regression is significant at 95% → p = 0.000) and there is no lack of fit (probability for lack of fit is not significant at 95% → p = 0.067). It can be seen from Figure S1 of the Supporting Information that the predicted values are quite close to the observed ones (R2 = 0.9317) indicating that the developed model successfully describes the correlation between factors and the total DD yield. The cross-validated coefficient of determination (Q2) quantifies the ability to predict new data correctly. For eq 4, Q2 = 0.6655 and this denotes good predictive power. Therefore, the quadratic model for the DD yield (eq 4) is a good approximation to the sludge solubilization performance by the TH and ATH processes. The trends followed by DD with n, T, and t are well predicted by the model. As expected according to the literature,35,42,43 DD yield is much more influenced by T and n than by t in the domain under investigation. The low influence of the operation time can be related to the fact that just initial heating and depressurization have a more marked effect than an increase in the operation time has in the studied operation range (from 15 to 40 min). The response surface for predicted DD values (obtained using eq 4) as a function of T and n is shown in Figure 1 for a constant value of t = 30 min, a reasonable operation time in TH at industrial scale.6 Four characteristic regions can be clearly distinguished in this response surface and these correspond to two minima and two maxima. Minimum 1: located in the low T and low n region, where DD values are low for TH (n = 0) and ATH, because both the solubilization and organics removal rates are low and quite similar. Minimum 2: located in the high T and high n region, where DD values decrease with T and n because in this region the operating conditions give rise to a significant TCOD and SCOD removal (according to eq 1 the lower the SCOD, the lower the DD values). In this case, the organics removal rate is higher than the solubilization rate.

RESULTS AND DISCUSSION

Sludge Solubilization: Influence of n and T. As stated in the introduction, it is widely known that the hydrolysis of sludge is the limiting stage during the anaerobic digestion process. This limitation becomes clearer in the case of secondary sewage sludge. The enhancement of the sludge solubilization through industrial pretreatments may facilitate a more efficient hydrolysis by the microorganisms involved in anaerobic digestion and this could improve the production of biogas. In this section, the sludge solubilization performance by means of the ATH process is discussed, in terms of the disintegration degree [DD (%) − eq 1], throughout the whole experimental domain defined in the Experimental Plan section. This procedure was carried out to identify the optimum conditions for the ATH pretreatment that would maximize the production of CH4 in the AD stage. The DD (%) results obtained under the experimental conditions described in Table 1 were used to estimate values for the coefficients of the quadratic polynomial model (pages S3−S5 of the Supporting Information) according to RSM. For the DD yield (eq 4) the model equation is: 6161

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

Figure 2. Contour plot (a) and response surface plot (b) for Filtration Constant (Fc) yield. Both plots show Fc vs oxidant coefficient (n) and temperature (T), for a constant operation time (t = 30 min).

Maximum 1: located in the high T and low n region; in accordance with the model and experimental results, sludge solubilization is moderately improved by the ATH process with respect to TH. On considering Figure 1, it must be highlighted that, for TH, DD changes from 10% at 80 °C to 43% at 170 °C. TH normally operates at 170 °C and 30 min. Under these conditions, the improvement in DD by adding H2O2 is practically negligible (DD = 46% at n = 0.2, 170 °C and 30 min). On working within the range n ∈ [0, 0.3), Takashima and Tanaka19 also reported that there was no difference in the sludge solubilization after thermal (170 °C, 60 min) and thermochemical pretreatment (170 °C, 60 min, H2O2 as oxidant). However, in the present work a DD value of 44.6% was obtained on operating at n = 0.1, 151 °C and 28 min (Expt. 14), while for the same conditions but without adding H2O2 (Expt. 19), DD was only 31.2%. Moreover, a narrow intermediate region can be seen in Figure 1, where ATH upon the addition of low H2O2 doses leads to a considerable increase in the DD (%) in comparison with TH under the same conditions (but without H2O2). This fifth characteristic region is approximately delimited by n ∈ [0, 0.3] and T ∈ [100, 130] °C. Maximum 2: located in the low T and high n region, where the highest DD values are obtained; as the temperature is too low to enhance the organics removal rate, the additional effect of high H2O2 doses is to cause a considerable increment in the sludge solubilization. For instance, a DD value of 60% was obtained on operating at n = 0.9, 99 °C and 28 min (Expt. 3). This is an impressive result but it must be noted that the H2O2 dose would be prohibitive on an industrial scale. According to experimental and predicted results, numerous combinations of factors within the explored experimental domain give rise to DD values for the ATH process that are significantly higher than those for TH (with both processes operated under the same T and t conditions). In some cases the DD values obtained were higher by a factor of 2 (or even more) than those obtained with TH. For the ATH process as an AD pretreatment, desired effluents are those that present high DD values [hence, high soluble chemical oxygen demand (SCOD)] and low organic matter removal percentages.32 In fact, thermochemical pretreatment with H2O2 carried out by Cacho-Rivero et al.17,18 and work reported by Takashima and Tanaka19 led to significant sludge solubilization and little organic matter removal after the pretreatment, which resulted

in a considerable increase in the volatile suspended solids (VSS) removal after AD. However, methane production was not greatly enhanced or, in some cases, it even decreased. This latter finding could be due to the use of operating conditions that were severe enough to degrade the methane precursors19 and/or to produce compounds that were toxic to the methanogenic bacteria. Sludge Dewaterability: Influence of n and T. The improvement of the sludge dewaterability is, in many cases, one of the most prized objectives to make the management of the sludge easier and more feasible, particularly for secondary sewage sludge. In fact, at the outset in this area the main reason to build hydrolysis plants was to enhance sludge dewaterability.6 The difficulty in dewatering and solubilizating the secondary sludge can be explained by the presence of extracellular polymeric substances (EPS).44 EPS are metabolic products that are 60% constituted by proteins and polysaccharides.1 These compounds accumulate on the bacterial cell surface and form a protective gel-like layer against desiccation, which also acts as a carbon and energy reservoir during starvation. According to the literature, EPS constitute up to 80% of the activated sludge mass and are crucial to the flocculation and dewatering of activated sludge.1,45 Jorand et al.46 reported floc structure models to explain the composition of resulting aggregates. It was reported by Abelleira et al.32 that the ATH process can radically improve the sludge dewaterability, in terms of time-tofilter (TTF) values, on increasing the temperature (T) and oxidant coefficient (n). This possibility can be confirmed through the experimental results obtained in the present work (Table 1) as filtration constant (Fc) and TTF values are consistent (Table S1 of the Supporting Information). Low TTF values correspond to high Fc values and vice versa, with these two cases indicating better or worse dewaterability of the studied sludge, respectively; thereby, Fc and TTF can both be used to assess the dewaterability of the sewage sludge. The secondary sewage sludge used as influent in this work had values of TTF = 65.75 min and Fc = 0.0029 cm2/min. From the Fc (cm2/min) results obtained under the experimental conditions described in Table 1, estimated coefficients of the quadratic polynomial model were calculated according to RSM. For the Fc yield (eq 5) the model equation is: 6162

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

be released, so improving sludge dewaterability and organic matter solubilization. In brief, during the ATH process the addition of H2O2 has a synergistic effect with the direct injections of steam from a boiler.32 In other words, steam injections aid the effect of peroxidation. The depressurization of the reactor content into the flash tank at the end of the reaction also has a considerable beneficial effect on the sludge dewaterability and solubilization.35 Organic Matter Removal (%): Influence of n, T, and t. In general, significant differences were not found between TCOD removal (%) and TOC removal (%) under the same experimental conditions. This means that both parameters could be used to predict the reaction conversion. From the TCOD and TOC removal experimental results reported in Table 1, estimated coefficients for the quadratic polynomial model were calculated according to RSM. The model equations obtained for the TCOD removal yield (eq 6) and the TOC removal yield (eq 7) are:

Y2(Fc) = 1.22 + 0.69X1 + 0.74X 2 − 0.04X3 − 0.45X12 − 0.27X 22 − 0.06X32 + 0.49X1X 2 − 0.0006X1X3 − 0.02X 2X3

(5)

where X1, X2, and X3 represent the values (in coded units) of n, T, and t, respectively. ANOVA showed that the model is statistically good (p = 0.000) and there is no lack of fit (p = 0.132). For eq 5, R2 = 0.8861 and Q2 = 0.3709 indicate an acceptable predictive power. The response surface for Fc values (obtained using eq 5) as a function of T and n is shown in Figure 2 for a constant value of t = 30 min. As expected, this response surface shows a clear enhancement of Fc with increasing n and T. However, it was statistically confirmed that the dependence of Fc with t (from 15 to 40 min) is not particularly strong in comparison with its dependence on n or T. This finding is consistent with a previous statement made in this work, namely that the effect of t on the disintegration degree (solubilization) is not particularly marked. In fact, our previous work32 showed certain connection between sludge solubilization and dewaterability enhancement, a situation that is consistent with studies in which it was considered that the mechanism to improve the sludge solubilization and dewaterability by means of thermochemical processes is analogous.1,47 This mechanism will be further discussed in this section. Furthermore, Neyens et al.48 showed that the effect of operation time on the sludge dewaterability was hardly significant in comparison with the effect of pH and temperature, in the acid thermal hydrolysis process. For the ATH process at moderate conditions of n and T (fifth characteristic region in part b of Figure 1), the predicted and experimental results are also promising. In fact, it can be confirmed that the addition of small doses of H2O2 is sufficient to significantly enhance the dewaterability of TH effluents at temperatures below 130 °C even at levels above those achieved under optimal conditions of TH (T = 170 °C, t = 30 min). In addition, TH at temperatures below 130 °C has a negative effect on the sludge dewaterability, as shown in this work and by Bougrier et al.49 This adverse effect may be related to the particle size and the partially modified EPS structure obtained, which are susceptible to clog the filtration test filters more easily. The significant dewaterability enhancement obtained when operating at 151 °C with n = 0.1 (Expt. 14 → Fc = 0.64 cm2/ min) in comparison with n = 0 (Expt. 19 → Fc = 0.02 cm2/ min) is worth highlighting. H2O2 increases the sludge solubilization and dewaterability through a mechanism that involves the reactivity of hydroxyl radicals (OH·) on organic molecules to produce nonselective ruptures and, subsequently, the degradation of the proteins and polysaccharides contained in the EPS of activated sludge. Because of this deterioration, the water retention properties of EPS are diminished thus releasing the EPS-bound water and soluble organic compounds.48 However, because most cells have an intracellular defense system against toxic radicals, the action of peroxides may be confined to the extracellular biopolymer matrix.47 It is at this point that the role of steam injections comes into play as they not only have the capability to denature the EPS, but also to disrupt cell activity and structure,6 thus disabling the intracellular defense system and paving the way to a more effective action of peroxides and hydroxyl radicals. Therefore, even the intracellular content may

Y3(TCOD removal) = 29.03 + 20.43X1 + 17.75X 2 + 4.79X3 − 1.39X12 − 6.90X 22 + 1.99X32 + 17.41X1X 2 + 6.16X1X3 + 3.13X 2X3

(6)

Y4(TOC removal) = 28.37 + 14.35X1 + 18.60X 2 + 6.19X3 − 9.18X12 − 2.80X 22 + 2.26X32 + 17.38X1X 2 + 6.82X1X3 + 3.94X 2X3

(7)

where X1, X2, and X3 represent the values (in coded units) of n, T, and t, respectively (Table 1). ANOVA showed that these models are statistically good (p = 0.000) and there is no lack of fit (p = 0.106 for Y3 and p = 0.219 for Y4). For eq 6, R2 = 0.9324 and Q2 = 0.5986 denote a good predictive power. For eq 7, R2 = 0.9633 and Q2 = 0.8586 indicate a very good predictive power. As can be seen in Figure S2 of the Supporting Information, the factors n, T, and t are significant in the implemented models for TCOD and TOC removal (%). Despite the fact that Figure S2 of the Supporting Information shows a trend of increasing TCOD and TOC removal, T and n have a more marked influence than t on the explored experimental domain. For n = 0.5 and temperatures between 135 and 170 °C, the models predict an organic matter removal of approximately 30−45% (parts b and e of Figure S2 of the Supporting Information). It should be noted that n = 0.5 corresponds to the oxidant quantity that is theoretically required to remove half of the organic matter contained in the raw sludge. The predicted response surfaces for TCOD and TOC removal respectively are shown in parts b and d of Figure 3. These results belong to the whole domain defined for n and T while considering a constant value of t = 40 min. The corresponding contour plots for each response surface are shown in parts a and c of Figure 3. We consider that studying the ATH as a postreatment of AD using secondary sludge is a good estimation to the study with digested sludge. This assumption is based on several works found in the literature.26−28,50,51 It can be seen from Figure 3 that the ATH process can reach a considerable level of effectiveness in sewage sludge treatment: organic matter removal percentages of approximately 85−92% on operating at 170 °C, equilibrium pressure and n = 1; that is, 6163

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

Figure 3. Contour plot (a and c) and response surface plot (b and d) for TCOD removal (%) and TOC removal (%) yields. The contour plot and the response surface show the removal yield vs oxidant coefficient (n) and temperature (T), for a constant operation time (t = 40 min).

view to further studies including anaerobic biodegradability tests. This optimization was carried out by means of a double interpolation routine. The second interpolation was used to ensure that the first was not trapped by a local minimum or maximum. The Optimizer application of the software Modde 9.0 (Umetrics) was used for this purpose. First, the variation limits for each factor must be fixed. A prior optimization considering the whole explored domain (where each factor can vary throughout its original range) was conducted. It is then necessary to define the desirability or target for each response, that is, to establish the desired response profile. According to the studied parameters, a good AD pretreatment should generate an effluent with the maximum solubilization, while the organic matter removal is minimized and its dewaterability maximized. In this case, the prior optimum operation was obtained for n = 0.63, T = 92 °C and t = 18 min. The response values predicted for this optimum point are: (a) an organic matter removal of 11%; (b) DD = 53%; (c) Fc = 0.6 cm2/min. These results are reasonably satisfactory as the operation time is short and the temperature quite low. However, the H2O2 dose is still considered prohibitive as well as severe. A second optimization routine was therefore conducted with the same desirability profile but with limits set on the variation of the factors: n ∈ [0, 0.3], T ∈ [100, 130] °C, and t ∈ [15, 30] min. These operation ranges match the fifth characteristic region indentified in the response surface for DD (section entitled Sludge Solubilization: Influence of n and T). In this

to attain these removal values, the addition of excess oxidant would not be required. By subjecting anaerobically digested sludge to WAO, Wu et al.28 obtained a TCOD removal rate of 85% [at 275 °C, 148 bar, 60 min and n = 1.2 (air as source of oxygen)]. SCWO of anaerobically digested sludge conducted by Xu et al.30 yielded TCOD and TOC removal rates of 99.6% and 98.5%, respectively [at 400 °C, 250 bar, 6 min and n = 2 (oxygen as oxidant agent)]. Furthermore, Abelleira et al.32 reported that in the ATH process conducted at temperatures above 120 °C and n ≥ 0.5, only the initial heating was required to operate close to − or even above − the temperature set point for approximately 20−35 min, a situation due to the exothermic nature of the organic matter oxidation. Despite this finding, the destruction of the organic matter as a postreatment of the AD may not be the main objective of the ATH process, since the H2O2 doses (besides the T and t operating conditions) required to reach these organic matter removal percentages are prohibitive for larger scale operations. As a result, this and further studies are going to be focused on the ATH process as a pretreatment for AD. Optimization − Desirability. The information acquired from the response surfaces discussed in the previous sections provides a rational strategy to optimize the ATH process as a pretreatment for AD. The optimization procedure was focused to obtain the best performance of the ATH process under operating conditions as mild as possible to avoid the degradation of methane precursors and/or the production of compounds that are toxic to the methanogenic bacteria, with a 6164

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

(5) Laurent, J.; Casellas, M.; Carrère, H.; Dagot, C. Effects of thermal hydrolysis on activated sludge solubilization, surface properties and heavy metals biosorption. Chem. Eng. J. 2011, 166 (3), 841−849, DOI: 10.1016/j.cej.2010.11.054. (6) Kepp, U.; Machenbach, I.; Weisz, N.; Solheim, O. E. Enhanced stabilisation of sewage sludge through thermal hydrolysis−three years of experience with a full scale plant. Water Sci. Technol. 2000, 42 (9), 89−96. (7) Nah, I. W.; Kang, Y. W.; Hwang, K.-Y.; Song, W.-K. Mechanical pretreatment of waste activated sludge for anaerobic digestion process. Water Res. 2000, 34 (8), 2362−2368, DOI: 10.1016/S0043-1354(99) 00361-9. (8) Zhang, D.; Chen, Y.; Zhao, Y.; Zhu, X. New sludge pretreatment method to improve methane production in waste activated sludge digestion. Environ. Sci. Technol. 2010, 44 (12), 4802−4808, DOI: 10.1021/es1000209. (9) Liu, X. L.; Liu, H.; Du, G. C.; Chen, J. Improved bioconversion of volatile fatty acids from waste activated sludge by pretreatment. Water Environ. Res. 2009, 81 (1), 13−20, DOI: 10.2175/106143008X304640. (10) Aldin, S.; Elbeshbishy, E.; Nakhla, G.; Ray, M. B. Modeling the effect of sonication on the anaerobic digestion of biosolids. Energy Fuels 2010, 24 (9), 4703−4711, DOI: 10.1021/ef901255k. (11) Weemaes, M.; Grootaerd, H.; Simoens, F.; Verstraete, W. Anaerobic digestion of ozonized biosolids. Water Res. 2000, 34 (8), 2330−2336, DOI: 10.1016/S0043-1354(99)00373-5. (12) Liu, J. C.; Lee, C. H.; Lai, J. Y.; Wang, K. C.; Hsu, Y. C.; Chang, B. V. Extracellular polymers of ozonized waste activated sludge. Water Sci. Technol. 2001, 44 (10), 137−142. (13) Toreci, I.; Kennedy, K. J.; Droste, R. L. Evaluation of continuous mesophilic anaerobic sludge digestion after high temperature microwave pretreatment. Water Res. 2009, 43 (5), 1273−1284, DOI: 10.1016/j.watres.2008.12.022. (14) Coelho, N. M. G.; Droste, R. L.; Kennedy, K. J. Evaluation of continuous mesophilic, thermophilic and temperature phased anaerobic digestion of microwaved activated sludge. Water Res. 2011, 45 (9), 2822−2834, DOI: 10.1016/j.watres.2011.02.032. (15) Erden, G; Filibeli, A. Improving anaerobic biodegradability of biological sludges by Fenton pre-treatment: Effects on single stage and two-stage anaerobic digestion. Desalination 2010, 251 (1−3), 58−63, DOI: 10.1016/j.desal.2009.09.144. (16) Genç, N.; Yonsel, Ş.; Dağaşan, L.; Onar, A. N. Wet oxidation: a pre-treatment procedure for sludge. Waste Manage. 2002, 22 (6), 611−616, DOI: 10.1016/S0956-053X(02)00040-5. (17) Cacho-Rivero, J. A.; Suidan, M. T. Effect of H2O2 dose on the thermo-oxidative co-treatment with anaerobic digestion of excess municipal sludge. Water Sci. Technol. 2006, 54 (2), 253−259, DOI: 10.2166/wst.2006.513. (18) Cacho-Rivero, J. A.; Madhavan, N.; Suidan, M. T.; Ginestet, P.; Audic, J. M. Enhancement of anaerobic digestion of excess municipal sludge with thermal and/or oxidative treatment. J. Environ. Eng. 2006, 132 (6), 638−644, DOI: 10.1061/(ASCE)0733-9372(2006) 132:6(638). (19) Takashima, M.; Tanaka, Y. Comparison of thermo-oxidative treatments for the anaerobic digestion of sewage sludge. J. Chem. Technol. Biotechnol. 2008, 83 (5), 637−642, DOI: 10.1002/jctb.1841. (20) Eskicioglu, C.; Prorot, A.; Marin, J.; Droste, R. L.; Kennedy, K. J. Synergetic pretreatment of sewage sludge by microwave irradiation in presence of H2O2 for enhanced anaerobic digestion. Water Res. 2008, 42 (18), 4674−4682, DOI: 10.1016/j.watres.2008.08.010. (21) Murray, A.; Horvath, A.; Nelson, K. L. Hybrid life-cycle environmental and cost inventory of sewage sludge treatment and enduse scenarios: a case study from China. Environ. Sci. Technol. 2008, 42 (9), 3163−3169, DOI: 10.1021/es702256w. (22) Boráň, J.; Houdková, L.; Elsäßer, T. Processing of sewage sludge: Dependence of sludge dewatering efficiency on amount of flocculant. Resour. Conserv. Recycl. 2010, 54 (5), 278−282, DOI: 10.1016/j.resconrec.2009.08.010. (23) Citeau, M.; Larue, O.; Vorobiev, E. Influence of salt, pH and polyelectrolyte on the pressure electro-dewatering of sewage sludge.

case, the optimum conditions corresponded to n = 0.3, T = 115 °C, and t = 24 min. The response values predicted for this optimum point are: (a) an organic matter removal of 17%, (b) DD = 38%, (c) Fc = 0.75 cm2/min. The empirical validation for this experiment was satisfactory: (a) an organic matter removal of 15%, (b) DD = 41%, (c) Fc = 0.66 cm2/min. It must be taken into account that, for the TH effluent at n = 0, T = 115 °C and t = 30 min, the predicted DD and Fc values are only 20% and 0.0023 cm2/min, respectively. Moreover, results for the optimum case are promising because, under these mild conditions of n and T, DD is close to and Fc is above the values predicted for the TH process under conventional operating conditions (T = 170 °C, t = 30 min → DD = 43%; Fc = 0.054 cm2/min). This DD value for the TH process at optimal conditions is consistent with the results obtained by Laurent et al.5 (DD = 46%). Further studies are required to assess whether the AD of the sludge solubilized by means of the ATH process give rise to a methane production at least similar to that generated by the AD of the TH effluent under conventional operating conditions (T = 170 °C, t = 30 min). These studies should be mainly focused on those ATH experiments that presented satisfactory solubilization and dewaterability results but they should be conducted under conditions that are as mild as possible (n ∈ [0.01, 0.3]; T ∈ [90, 135] °C; t < 30 min) or even under high temperature conditions (T ∈ [150, 170] °C), low H2O2 doses (n ∈ [0.01, 0.2]) and t < 30 min.



ASSOCIATED CONTENT



AUTHOR INFORMATION

* Supporting Information S

Additional information, table and figures on analytical methods, design of experiments and results. This material is available free of charge via the Internet at http://pubs.acs.org. Corresponding Author

*E-mail: [email protected], phone: +34 956 016458, fax: +34 956 016411. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge financial support from the NOVEDAR_Consolider Project (CSD2007-00055) promoted by the Spanish Ministry of Education and Science.



REFERENCES

(1) Neyens, E.; Baeyens, J.; Dewil, R.; De heyder, B. Advanced sludge treatment affects extracellular polymeric substances to improve activated sludge dewatering. J. Hazard. Mater. 2004, 106 (2−3), 83−92, DOI: 10.1016/j.jhazmat.2003.11.014. (2) Dwyer, J.; Starrenburg, D.; Tait, S.; Barr, K.; Batstone, D. J.; Lant, P. Decreasing activated sludge thermal hydrolysis temperature reduces product colour, without decreasing degradability. Water Res. 2008, 42 (18), 4699−4709, DOI: 10.1016/j.watres.2008.08.019. (3) Ponsá, S.; Ferrer, I.; Vázquez, F.; Font, X. Optimization of the hydrolytic−acidogenic anaerobic digestion stage (55 °C) of sewage sludge: Influence of pH and solid content. Water Res. 2008, 42 (14), 3972−3980, DOI: 10.1016/j.watres.2008.07.002. (4) Ma, Y.; Wilson, C. A.; Novak, J. T.; Riffat, R.; Aynur, S.; Murthy, S.; Pruden, A. Effect of various sludge digestion conditions on sulfonamide, macrolide, and tetracycline resistance genes and class I integrons. Environ. Sci. Technol. 2011, 45 (18), 7855−7861, DOI: 10.1021/es200827t. 6165

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166

Environmental Science & Technology

Article

Water Res. 2011, 45 (6), 2167−2180, DOI: 10.1016/ j.watres.2011.01.001. (24) Fytili, D.; Zabaniotou, A. Utilization of sewage sludge in EU application of old and new methods − A review. Renew. Sustain. Energy Rev. 2008, 12 (1), 116−140, DOI: 10.1016/j.rser.2006.05.014. (25) Appels, L.; Baeyens, J.; Degrève, J.; Dewil, R. Principles and potential of the anaerobic digestion of waste-activated sludge. Prog. Energy. Combust. Sci. 2008, 34 (6), 755−781, DOI: 10.1016/ j.pecs.2008.06.002. (26) Khan, Y.; Anderson, G. K.; Elliott, D. J. Wet Oxidation of Activated Sludge. Water Res. 1999, 33 (7), 1681−1687, DOI: 10.1016/ S0043-1354(98)00387-X. (27) Bernardi, M.; Cretenot, D.; Deleris, S.; Descorme, C.; Chauzy, J; Besson, M. Performances of soluble metallic salts in the catalytic wet air oxidation of sewage sludge. Catal. Today 2010, 157 (1−4), 420− 424, DOI: 10.1016/j.cattod.2010.01.030. (28) Wu, Y. C.; Hao, O. J.; Olmstead, D. G.; Hsieh, K. P.; Scholze, R. J. Wet air oxidation of anaerobically digested sludge. J. Water Pollut. Control Fed. 1987, 59 (1), 39−46. (29) Svanström, M.; Fröling, M.; Modell, M.; Peters, W. A.; Tester, J. Environmental assessment of supercritical water oxidation of sewage sludge. Resour. Conserv. Recycl. 2004, 41 (4), 321−338, DOI: 10.1016/ j.resconrec.2003.12.002. (30) Xu, D.; Wang, S.; Tang, X.; Gong, Y.; Guo, Y.; Wang, Y.; Zhang, J. Design of the first pilot scale plant of China for supercritical water oxidation of sewage sludge. Chem. Eng. Res. Des. 2011, DOI: 10.1016/ j.cherd.2011.06.013. (31) Dewil, R.; Baeyens, J.; Neyens, E. Fenton peroxidation improves the drying performance of waste activated sludge. J. Hazard. Mater. 2005, 117 (2−3), 161−170, DOI: 10.1016/j.jhazmat.2004.09.025. (32) Abelleira, J.; Pérez-Elvira, S. I.; Sánchez-Oneto, J.; Portela, J. R.; Nebot, E. Advanced thermal hydrolysis of secondary sewage sludge: A novel process combining thermal hydrolysis and hydrogen peroxide addition. Resour. Conserv. Recycl. 2012, 59, 52−57, DOI: 10.1016/ j.resconrec.2011.03.008. (33) American Public Health Association (APHA). Standard Methods for the Examination of Water and Wastewater; APHA: Washington, D.C., 2005. (34) Kim, D.-H.; Jeong, E.; Oh, S.-E.; Shin, H.-S. Combined (alkaline + ultrasonic) pretreatment effect on sewage sludge disintegration. Water Res. 2010, 44 (10), 3093−3100, DOI: 10.1016/ j.watres.2010.02.032. (35) Donoso-Bravo, A.; Pérez-Elvira, S.; Aymerich, E.; Fdz-Polanco, F. Assessment of the influence of thermal pre-treatment time on the macromolecular composition and anaerobic biodegradability of sewage sludge. Bioresour. Technol. 2011, 102 (2), 660−666, DOI: 10.1016/ j.biortech.2010.08.035. (36) Portela, J. R.; López, J.; Nebot, E.; Martínez de la Ossa, E. Elimination of cutting oil wastes by promoted hydrothermal oxidation. J. Hazard. Mater. 2001, 88 (1), 95−106, DOI: 10.1016/S03043894(01)00295-3. (37) Sánchez-Oneto, J.; Mancini, F.; Portela, J. R.; Nebot, E.; Cansell, F.; Martínez de la Ossa, E. J. Kinetic model for oxygen concentration dependence in the supercritical water oxidation of an industrial wastewater. Chem. Eng. J. 2008, 144 (3), 361−367, DOI: 10.1016/ j.cej.2008.01.041. (38) Zalouk, S.; Barbati, S.; Sergent, M.; Ambrosio, M. Disposal of animal by-products by wet air oxidation: Performance optimization and kinetics. Chemosphere 2009, 74 (2), 193−199, DOI: 10.1016/ j.chemosphere.2008.09.084. (39) Pérez-Elvira, S. I.; Fdz-Polanco, F.; Fdz-Polanco, M.; Rodríguez, P.; Rouge P. Hydrothermal multivariable approach. Full-scale feasibility study. Electron. J. Biotechnol. 2008, 11 (4) DOI:10.2225/ vol11-issue4-fulltext-14. (40) Neyens, E.; Baeyens, J. A review of thermal sludge pre-treatment processes to improve dewaterability. J. Hazard. Mater. 2003, 98 (1−3), 51−67, DOI: 10.1016/S0304-3894(02)00320-5. (41) Penaud, V.; Delgenes, J. P.; Moletta, R. Influence of thermochemical pretreatment conditions on solubilization and

anaerobic biodegradability of a microbial biomass. Environ. Technol. 2000, 21 (1), 87−96, DOI: 10.1080/09593332108618141. (42) Valo, A.; Carrère, H.; Delgenès, J. P. Thermal, chemical and thermo-chemical pre-treatment of waste activated sludge for anaerobic digestion. J. Chem. Technol. Biotechnol. 2004, 79 (11), 1197−1203, DOI: 10.1002/jctb.1106. (43) Chung, J.; Lee, M.; Ahn, J.; Bae, W.; Lee, Y. W.; Shim, H. Effects of operational conditions on sludge degradation and organic acids formation in low-critical wet air oxidation. J. Hazard. Mater. 2009, 162 (1), 10−16, DOI: 10.1016/j.jhazmat.2008.05.038. (44) Liu, Y.; Fang, H. H. P. Influences of extracellular polymeric substances (EPS) on flocculation, settling, and dewatering of activated sludge. Crit. Rev. Environ. Sci. Technol. 2003, 33 (3), 237−273, DOI: 10.1080/10643380390814479. (45) Liu, H.; Fang, H. H. P. Extraction of extracellular polymeric substances (EPS) of sludges. J. Biotechnol. 2002, 95 (3), 249−256, DOI: 10.1016/S0168-1656(02)00025-1. (46) Jorand, F.; Boué-Bigne, F.; Block, J. C.; Urbain, V. Hydrophobic/hydrophilic properties of activated sludge exopolymeric substances. Water Sci. Technol. 1998, 37 (4−5), 307−315, DOI: 10.1016/S0273-1223(98)00123-1. (47) Christensen, B. E.; Trønnes, H. N.; Vollan, K.; Smidsrød, O.; Bakke, R. Biofilm removal by low concentrations of hydrogen peroxide. Biofouling 1990, 2 (2), 165−175, DOI: 10.1080/ 08927019009378142. (48) Neyens, E.; Baeyens, J.; Weemaes, M. De heyder, B. Hot acid hydrolysis as a potential treatment of thickened sewage sludge. J. Hazard. Mater. 2003, 98 (1−3), 275−293, DOI: 10.1016/S03043894(03)00002-5. (49) Bougrier, C.; Delgenès, J. P.; Carrère, H. Effects of thermal treatments on five different waste activated sludge samples solubilisation, physical properties and anaerobic digestion. Chem. Eng. J. 2008, 139 (2), 236−244, DOI: 10.1016/j.cej.2007.07.099. (50) Mikkelsen, L. H.; Keiding, K. Physico-chemical characteristics of full scale sewage sludges with implications to dewatering. Water Res. 2002, 36 (10), 2451−2462, DOI: 10.1016/S0043-1354(01)00477-8. (51) Nielsen, P. H.; Fralund, B.; Keiding, K. Changes in the composition of extracellular polymeric substances in activated sludge during anaerobic storage. Appl. Microbiol. Biotechnol. 1996, 44 (6), 823−830, DOI: 10.1007/s002530050639.

6166

dx.doi.org/10.1021/es204203y | Environ. Sci. Technol. 2012, 46, 6158−6166