• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Activated Sludge Inhibition By Chemical Stressors-A Comprehensive Study

Posted on: Friday, 28 September 2007, 06:00 CDT

By Henriques, Ines D S Kelly, Richard T II; Dauphinais, Jennifer L; Love, Nancy G

ABSTRACT: The effects of shock loads of 1-chloro-2,4- dinitrobenzene (CDNB); cadmium; 1-octanol; 2,4-dinitrophenol (DNP); weakly complexed cyanide; pH 5, 9, and 11; and high ammonia levels on activated sludge biomass growth, respiration rate, flocculation, chemical oxygen demand removal, dewaterability, and settleability were studied. For all chemical shocks, except ammonia and pH, concentrations that caused 15, 25, and 50% respiration inhibition were used to provide a single pulse shock to sequencing batch reactor systems containing a nitrifying or non-nitrifying biomass. Cadmium and pH 11 shocks were most detrimental to all processes, followed by CDNB. The DNP and cyanide primarily affected respiration, while pH 5, pH 9, octanol, and ammonia did not affect the treatment process to a significant extent. A chemical source- process effect matrix is provided, which we believe will aid in the development of methods that prevent and/or attenuate the effects of toxic shock loads on activated sludge systems.

Water Environ. Res., 79, 940 (2007).

KEYWORDS: toxins, shock load, treatment upset, SBRs.

doi: 10.2175/106143007X156709

Introduction

Biological wastewater treatment systems are susceptible to toxic shock loads of industrial chemicals, which can adversely affect the treatment process efficiency (Love and Bott, 2000). Many transient upset events are known to be caused by shock loads of toxic chemicals. Furthermore, studies have shown that chemical toxins can detrimentally affect all the essential processes within an activated sludge treatment system.

For example, deflocculation events, which involve the breakup of biomass floes resulting in increased effluent total suspended solids (TSS), have been found to result from chemical insults to treatment systems. Phenol (Schwartz-Mittelmann and Galil, 2000), heavy metals (Bott and Love, 2001; Neufeld, 1976), and organic electrophilic chemicals (Bott and Love, 2002) are some of the sources that have been found to induce deflocculation of activated sludge mixed liquor. Similarly, chemical oxygen demand (COD) removal has been found to be affected by heavy metals (Bott and Love, 2001; Weber and Sherrard, 1980), high ammonia loadings (Li and Zhao, 1999), and organic compounds, such as 3-chloroaniline (Boon et al., 2003) or high concentrations of 2,4-dinitrophenol (DNP) (Rich and Yates, 1955).

Many studies have found that respiration inhibition occurs in response to the presence of toxic chemicals, such as heavy metals (Bott et al., 2001; Lajoie et al, 2003; Madoni et al., 1999) and different classes of organic compounds (Bott et al., 2001). Consequently, respiration inhibition is frequently used to detect incoming toxicity to wastewater treatment plants, through the use of online respirometers (Buitron et al, 2005; Kong, Vaerewijck, and Verstraete, 1996; Kong, Vanrolleghem, Willems, and Verstraete, 1996; Vanrolleghem et al., 1994).

Industrial chemical toxins have also been found to cause detrimental effects on settleability and dewaterability of activated sludges. In a study performed by Kjellerup et al. (2001), both settleability and dewaterability were found to decrease at a full- scale industrial wastewater treatment facility in response to an unknown chemical shock. A study done by Boon et al. (2003) also found that decreased settleability occurred in activated sludge exposed to chloroaniline, and Novak (2001) has correlated poor settleability of mixed liquor with the presence of high ammonium concentrations.

Although some studies found in the literature have been conducted to assess the effects of toxins on the activated sludge process, these reports generally focus on only one aspect of the process, such as COD removal, nitrification, or settleability of the biomass, and do not make a comprehensive analysis on all the potential effects of that toxin. Additionally, the laboratory-scale reactors used in these studies are typically fed synthetic wastewater, which may change the characteristics of the biomass to a great extent and provide limited information about real-world systems. Moreover, continuous feed of a certain concentration of a toxin is typically preferred over providing shock loads of the same source, and the criteria to select those concentrations is not consistent across the literature, which does not allow a systematic comparison between the effects of different toxins. Furthermore, some chemical classes have been studied more than others. For example, studies on the effect of organic electrophiles, hydrophobic chemicals, or extreme pH levels on activated sludge are scarce or nonexistent in the literature. Therefore, there is a need for comprehensive studies that can adequately mimic the conditions at a treatment plant, both in terms of biomass, influent, and toxic-shock characteristics.

The objective of this research was to establish source-effect relationships for activated sludge exposed to shock loads of chemical toxins. Six different classes of industrially relevant chemicals were selected as sources, and the effects of varying shock concentrations of those toxins on activated sludge COD removal ability, flocculation ability, biomass growth, respiration rates, settleability, and dewaterability were assessed. These studies were conducted on both nitrifying and non-nitrifying activated sludge mixed liquors. Results are reported for both activated sludges in terms of the magnitude of the effect and time of recovery of shocked laboratory-scale reactors relative to a control reactor, to which no toxin was added. The chemical classes (and model compound within each class) chosen as toxic shock sources included heavy metals (cadmium), uncouplers of oxidative phosphorylation (DNP), organic electrophilic chemicals (1-chloro-2,4-dinitrobenzene [CDNB]), hydrophobic chemicals (1octanol), respiration inhibitors (weakly complexed cyanide), high ammonia loadings, and alkaline and acidic pH conditions. This work is part of a broader project focused on developing comprehensive source-effect relationships for different chemical sources and identifying the causal mechanisms linking these sources and their effects. The effects of the chemical toxins used in this study on nitrification were also assessed, and those results were reported elsewhere (Kelly et al., 2004).

Methodology

Pilot-Plant Reactors. A pilot-plant unit served as the source of mixed liquor for sequencing batch reactor (SBR) laboratory experiments. It consisted of two SBRs maintained at two different solids retention times (SRTs)-10 days (nitrifying system) and 2 days (non-nitrifying biomass). Both reactors had a working volume of 180 L, a hydraulic retention time (HRT) of 1 day, and were maintained at approximately 20[degrees]C. The SBRs were operated using four cycles per day, with a cycle consisting of a 5-hour reaction/aeration time and a 1-hour settling/decant period. The 10-day SRT reactor was fed over a period of less than 10 minutes at the beginning of each cycle. The 2-day SRT reactor was step-fed one-third of the influent at the beginning of each cycle, and at 1 hour and 45 minutes, and 3 hours and 30 minutes into the cycle. This was done to prevent nitrification from occurring in the 2-day SRT system. Feed consisted of raw domestic wastewater supplemented with a carbon source containing equal COD quantities of acetate and glucose and KGRO All- Purpose Plant Food fertilizer (KMart, Troy, Michigan) with a total nitrogen (N):available phosphate (P^ sub 2^O^ sub 5^):soluble potash (K^ sub 2^O) content of 15:30:15 on a mass basis. For the nitrifying reactor, approximately 400 mg/L as COD supplement was added, while approximately 800 mg/L as COD was added to the non-nitrifying system to achieve realistic biomass concentrations. The feed was settled for 30 minutes before the fill period to mimic primary clarification. Details regarding the operation and treatment performance of the pilot-plant unit can be found in Love et al. (2005).

Laboratory Reactor Configuration and Operation. The laboratory- scale reactors were designed to mimic the pilot-plant SBRs regarding the number of cycles, reaction/settling periods, the SRTs, and feed characteristics (influent was refrigerated in the laboratory). Both nitrifying and non-nitrifying laboratory reactors consisted of 3.5- L working volume beakers and were fed at the beginning of each cycle for a period of less than 10 minutes. No supplement was added to the 10-day SRT cadmium experiment, and no fertilizer was added with the COD supplement for the 10-day SRT octanol experiment. The 10-day SRT laboratory-scale reactor system was maintained at room temperature (approximately 23[degrees]C), while the 2-day SRT system was kept at 18[degrees]C to prevent nitrification from occurring throughout the experiment. For the 2-day SRT laboratory reactors, temperature control was used as a nitrification prevention strategy instead of the step-feeding approach used in the pilot plant, to simplify reactor operation. One reactor was used as a control reactor, to which no contaminant was added, and three other reactors were shocked at the beginning of cycle one (single pulse event), with set amounts of each contaminant, and monitored over time. Manual biomass wastage was conducted daily during the reaction/aeration period of a cycle and did not account for the solids lost in the effluent. For all the toxins tested, except ammonia and pH shock, the concentrations of each contaminant added to the reactors were defined as the concentrations that inhibited short-term (less than 30 minutes) oxygen uptake by the biomass by 15, 25, and 50% (termed IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^, respectively, as discussed below). Experiments with the nitrifying biomass (10-day SRT) were monitored for 30 days (3 x SRT) or less if recovery to control levels was observed to occur earlier, while the 2-day SRT system was monitored for 6 days (3 x SRT). Parameters monitored included effluent soluble COD, effluent TSS, mixed liquor total and volatile suspended solids (MLSS and MLVSS), mixed liquor specific oxygen uptake rate (SOUR), mixed liquor capillary suction time (CST), and mixed liquor sludge volume index (SVI). The characteristics of the influent wastewater were also monitored throughout each of the shock experiments, and detailed results are reported in Love et al. (2005). The concentrations of contaminants in the effluent and mixed liquor of the shocked reactors were also determined whenever possible and are reported in the same publication (Love et al., 2005).

Toxins. Contaminants tested for inhibition included cadmium (added as CdCl^ sub 2^), CDNB, DNP, 1-octanol, cyanide (added as a zinc-cyanide complex [Zn^ sub x^(CN)^ sub y^^ sup +z^] solution to mimic electroplating waste), ammonia (added as NH^ sub 4^HCO^ sub 3^), and pH shock. Zinc sulfate (ZnSO^ sub 4^) was added to the control reactor of the complexed cyanide shock experiments to offset any potential toxicity from the metal in the shocked reactors. Cadmium, cyanide, and DNP were added as concentrated stock solutions. Because of the low water solubility of octanol and CDNB, extra steps were required for addition of these toxins to the activated sludge. To shock the SBR reactors, octanol was either dissolved overnight or through sonication in autoclaved or raw influent, respectively. The CDNB was melted (at approximately 100[degrees]C) and dissolved using sonication into preheated raw influent before introducing the shock load to the SBRs reactors. In both cases, the mixture was cooled to room temperature before addition to the mixed liquor. For pH shock, sulfuric acid or sodium/ calcium hydroxides were added to adjust the mixed liquor pH to 5, 9, and 11. For pH 11 shock, equal normality sodium and calcium hydroxides were used to maintain a similar monovalent-to-divalent- cation ratio in the mixed liquor. For ammonia shock, the use of the bicarbonate form of ammonium prevented potential low pH effects resulting from increased nitrification from interfering with the interpretation of effects by ammonium alone.

Inhibitory Concentration (IC^ sub x^) Determination. The SOUR- based IC^ sub xx^ assays were conducted 1 day before initiating each shock experiment, to ensure that mixed liquor composition was representative of the biomass that would be used to inoculate the laboratory-scale SBRs. The SOUR procedure is described in the Analytical Procedures section. The IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ concentrations of cadmium, DNP, CDNB, octanol, and complexed cyanide used to shock the reactors were determined with short-term respiration inhibition experiments, by placing a specific quantity of the toxin and the standard amount of soluble COD in the oxygen uptake rate (OUR) assay biochemical oxygen demand (BOD) bottle and subsequently adding the previously aerated mixed liquor sample to the bottle (mixed liquor from the pilot plant was used). Dissolved oxygen readings were started within 30 seconds of blending the toxin with the mixed liquor. A control SOUR was also run, to which no toxin was added. The inhibitory concentrations (IC^ sub xx^) for each toxin were determined by plotting SOUR versus contaminant concentration and fitting a second-order polynomial or a double or single exponential decay function through the data (best fit determined by least squares analysis. Excel 2000 [Microsoft Corporation, Bellevue, Washington] or SigmaPlot 8.0 [Systat Software Inc., San Jose, California]). The R^ sup 2^ values for the fits were all above 0.90. The equations for the best-fit curves were used to determine the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ concentrations, corresponding to 15, 25, and 50% respiration inhibition. In all cases,

Respiration Inhibition (%) [(SOUR^sub control^ - SOUR)/SOUR^sub control^] x 100.

The reactors exposed to ammonia upset were shocked with 40, 130, and 280 mg/L nitrogen (approximately 3, 9, and 18 times the average ammonia in the influent, respectively) for the 2-day SRT system and 70, 190, and 390 mg/L nitrogen (approximately 2, 5, and 10 times the average ammonia in the influent, respectively) for the 10-day SRT system.

Analytical Procedures. The SOUR assays were used to monitor respiration inhibition in the stressed reactors. A mixed liquor sample was aerated for at least 3 to 5 minutes and placed in a 300- mL BOD bottle, to which 100 to 120 mg/L of soluble COD was previously added, to ensure that respiration was not substrate- limited during the test. Soluble COD was composed of 34% protein (beef extract, bacto-casitone, and yeast extract); 18% carbohydrate (fructose, galactose, and glucose); and 48% organic acids/alcohols (glacial acetic acid and glycerol) on a COD basis. The SOUR tests were performed in duplicate using a dual-channel Accumet Research AR25 pH/mV/[degrees]C/ISE meter (Fisher Scientific International, Hampton, New Hampshire) and two Orion 97-08 oxygen electrodes coupled with localized mixers (Thermo Electron Corporation, Waltham, Massachusetts). Dissolved oxygen readings were recorded every 6 seconds using an automated data acquisition system (Labview 6i, National Instruments, Austin, Texas). The OUR was determined through the slope of the linear portion of the dissolved oxygen versus time curve. The SOUR was calculated by dividing the OUR by the MLVSS concentration.

Effluent soluble COD, effluent TSS, MLSS, MLVSS, CST, and SVI were conducted according to Standard Methods (APHA et al., 1998). The CST test was conducted with mixed liquor taken directly from the SBR reactors during the reaction period. The SVI test was performed using 250-mL graduated cylinders. Samples for effluent soluble COD were filtered through a 0.45-[mu]m nitrocellulose filter immediately after effluent collection.

Data Presentation and Statistical Analysis. To assess the significance of an effect in a shocked reactor in comparison with the control reactor and the recovery of the shocked reactors to control reactor levels, statistics were performed using Dunnett's method for multiple comparisons with a control (Berthouex and Brown, 1994). The significance level (alpha) used was 0.05, while the criterion to determine recovery to control levels was defined as three consecutive data points not significantly different from the control. Data included in the figures represents the average of triplicate (COD, MLVSS, and effluent TSS) or duplicate (SOUR) measurements. For easier visualization of the data, error bars were not included. Because of the large amount of data generated during these studies, only the most significant results are shown in the figures.

Results

Severe Inhibition of Chemical Oxygen Demand Removal by Cadmium and pH 11 Shocks. Effluent soluble COD data show that cadmium had a strong effect on COD removal ability for both the 2- and 10-day SRT reactors (Figures 1a and 1b). In the 10-day SRT system, the effluent soluble COD remained significantly elevated in the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ cadmium-stressed reactors relative to the control until cycle 15 (0.37 x SRT), 23 (0.57 x SRT), and 51 (1.3 x SRT), respectively, after which the performance of all reactors was similar. Although partial statistical recovery was seen, complete recovery to control levels was not achieved in any of the 2-day SRT reactors. In this system, the maximum effluent COD level observed in the IC^ sub 50^ reactor was 13 times higher than the control, while, in the 10-day SRT IC^ sub 50^ reactor, the maximum effluent COD level observed was 3 times higher than the control. In both cases, the degree to which COD removal decreased corresponded with increasing cadmium concentrations.

The COD removal efficiency was also affected by CDNB in both the 10- and 2-day SRT systems; however, despite being statistically significant, the effects from the lower concentrations tested (IC^ sub 15^ and IC^ sub 25^ reactors) were modest (Figures 1c and 1d). The effluent soluble COD in the 10-day SRT system remained elevated in the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^-stressed reactors relative to the control until cycle 20 (0.50 x SRT), 24 (0.60 x SRT), and 36 (0.90 x SRT), respectively. Inhibition of COD removal efficiency was highest for the IC^ sub 50^-shocked reactor, which had its maximum COD level (2.5 times higher than the control) during cycle 2, but the inhibition in the IC^ sub 15^ and IC^ sub 25^ reactors did not significantly differ from each other and was much less severe (data not shown). For the 2-day SRT, a similar response was observed. The IC^ sub 50^ reactor showed elevated effluent soluble COD with respect to the control reactor for the first 8 cycles (1.0 x SRT) and had the highest inhibition of all reactors, with COD levels reaching as much as 4 times those of the control reactor.

The pH 11 shock had a significant detrimental effect on COD removal. The shock condition caused a maximum increase in effluent soluble COD relative to the control of 8 and 15 times for the 10- and 2-day SRT systems, respectively, and was the most severe COD removal response of all toxins tested (Figure 1e). Similar to cadmium, the 10-day SRT reactor recovered to control levels by cycle 52 (1.3 x SRT), while the 2-day SRT reactor never recovered to control levels over a period of 3 x SRT. The other pH shock conditions did not cause significant effluent COD increases. The DNP, octanol, cyanide, and ammonia shocks did not affect activated sludge COD removal ability, or the effect was rather modest and/or short-lived (data not shown). Among these conditions, only DNP at the highest concentration tested showed a slight effect on the COD removal capacity of the 10-day SRT biomass, as shown in Figure If.

Deflocculation Effects Similar to Chemical Oxygen Demand Removal Effects. Deflocculation effects were observed for several of the toxic conditions tested on both the 10- and 2-day SRT biomasses and were detected based on an increase in the effluent TSS of the stressed reactors.

Cadmium shock was found to have the strongest negative effect on the flocculation ability of both the 10- and 2-day SRT biomasses, given that effluent TSS increased significantly in all shocked reactors on exposure to the chemical and took a long time to recover or did not recover to control levels during the experiment (Figures 2a and 2b). However, contrary to what was observed with COD removal, the severity of the increase did not correspond with the cadmium dosage, as all tested concentrations yielded similar effluent TSS values. For the 10-day SRT system, maximum increases of 4 to 5 times the control were observed, while, for the 2-day SRT system, a maximum increase of approximately 2.5 times was noted for all reactors. Statistical recovery of the 10-day SRT reactors occurred after 83 cycles (2.1 x SRT), while, for the 2-day SRT system, the IC^ sub 25^ and IC^ sub 50^-shocked biomasses did not recover their flocculation ability within the experimental period of 3 x SRT. The 2-day SRT IC^ sub 15^ reactor temporarily recovered to control levels after 8 cycles (1 x SRT), but elevated effluent TSS levels were again noticed towards the end of the experimental period.

The CDNB was also found to cause elevated TSS levels, with deflocculation patterns similar to cadmium, but not as severe (Figures 2c and 2d). For the IC^ sub 50^ CDNB-shocked reactors, effluent TSS levels exceeded those in the control reactor by more than 1.5 and 3 to 4 times, respectively, for the 2- and 10-day SRT systems, but the CDNB shock load did not affect the 10-day SRT IC^ sub 15^ reactor or the 2-day SRT IC^ sub 15^ and IC^ sub 25^ reactors. Statistical recovery in the reactors where significant effects were observed was also much faster than in the cadmium- shocked reactors, with the 10-day SRT IC^ sub 25^ and IC^ sub 50^ reactors recovering after 10 (0.25 x SRT) and 28 (0.70 x SRT) cycles, respectively, and the 2-day SRT IC^ sub 50^ reactor recovering after 4 cycles (0.50 x SRT).

The effluent TSS levels for both the 10- and 2-day SRT pH 11 reactors showed immediate and substantial increases in effluent TSS, corresponding to 6 and 2.5 times the control levels, respectively (Figure 2e). Despite the severity of the increase, recovery was more rapid than found for both cadmium and CDNB, with the 10-day SRT system recovering after 20 cycles (0.50 x SRT) and the 2-day SRT reactor recovering after 4 cycles (0.50 x SRT). The reactors receiving pH 5 and pH 9 shocks did not show significant increases in effluent TSS relative to the control reactors (data not shown), and, in fact, a slight but statistically significant improvement was observed in the effluent TSS of the 2-day SRT pH 5 reactor relative to the control.

The remaining toxic conditions (octanol, DNP, cyanide, and ammonia shock loads) did not produce strong deflocculation events or long recovery times from initial increases in effluent TSS. Of these, only octanol and DNP showed some deflocculation effects. For the first 20 cycles (0.50 x SRT) of the 10-day SRT IC^ sub 50^ reactor shocked with DNP, moderate but statistically significant elevated effluent TSS levels were observed (Figure 2f). Effluent TSS increased in the 2- and 10-day SRT IC^ sub 50^ reactors shocked with octanol, to approximately 1.5 to 2 times that of the respective control reactor; however, in both cases, the values returned to control levels within 8 cycles (data not shown). In addition to the brief deflocculation effects observed for octanol, a foaming event was observed immediately after octanol addition, and increasing amounts of foam were observed with increasing concentrations of the contaminant.

Biomass Growth Mostly Affected by Cadmium, pH 11, and DNP Shocks. Mixed liquor concentrations in the laboratory-scale reactors varied during the experiments, as a result, in part, of variations in wastewater strength (see Love et al., 2005); nevertheless, a comparison of shocked reactors with the controls was still possible.

In the 10-day SRT system (Figure 3b), cadmium shock resulted in a maximum decrease in MLVSS levels of 18 and 26%, respectively, in the IC^ sub 25^ and IC^ sub 50^ reactors, after which the reactors returned to control levels at cycle 20 (IC^ sub 25^ reactor, 0.50 x SRT) and cycle 35 (IC^ sub 50^ reactor, 0.88 x SRT). No significant decrease in MLVSS was noted in the IC^ sub 15^ reactor. Although increased effluent TSS levels in the stressed reactors may partially account for the decrease in MLVSS, a mass balance on the reactor solids showed that the increase in effluent TSS alone could not explain the decrease in biomass after cadmium shock (data not shown). Similar results were obtained for the 2-day SRT system (Figure 3a), which experienced a 25% (IC^ sub 15^ reactor) and 40% (IC^ sub 25^ and IC^ sub 50^ reactors) decrease in MLVSS relative to the control. Recovery of MLVSS levels was not achieved in any of the shocked reactors.

The pH 11 strongly affected biomass growth in both the 10- and 2- day SRT systems, starting immediately after the shock event (Figure 3c). In the 2-day SRT reactor, decreased MLVSS levels relative to the control reactor were observed throughout the experiment. In the 10-day SRT system, a significant decrease in MLVSS was observed until cycle 29 (0.73 x SRT), after which, the pH 11-shocked reactor performed as the control. In both systems, strong deflocculation events were observed; however, as with cadmium, the increased effluent TSS levels alone did not justify the decrease in MLVSS, indicating that biomass growth was impaired by pH 11 shock.

The results concerning biomass growth in the 10-day SRT DNP- shocked reactors (Figure 3d) indicate that strong growth inhibition occurred since the very first cycle of the experiment. In comparison with the control, the MLVSS concentrations in the DNP treated reactors were statistically significantly lower since the beginning of the experiment and until cycle 33 (0.83 x SRT). In addition, the difference between the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ MLVSS profiles was not significant, although the effect in the IC^ sub 50^ reactor seemed to have been slightly stronger. Combined with the COD removal data, the MLVSS results show the uncoupling effects associated with DNP, whereby biomass growth or anabolic processes were inhibited to a significant extent, but substrate removal or catabolic processes continued to occur near control levels during the same period. The 2-day SRT shocked reactors did not show a significant decrease in MLVSS relative to control levels (data not shown), which is also consistent with the results obtained for COD removal and OUR (see below). It is not clear why DNP shock did not yield typical results for an uncoupler of oxidative phosphorylation in the non-nitrifying, low SRT system.

The pH 5, pH 9, cyanide, ammonia, octanol, and CDNB shocks did not affect biomass growth to a great extent. The MLVSS concentration in the 10-day SRT IC50 reactor shocked with CDNB dropped a maximum of 25% relative to the control and recovered to control values after 29 cycles (0.72 x SRT, data not shown). This reactor was the only one affected by CDNB shock regarding biomass growth.

Inhibition of Respiratory Functions Observed for All Toxic Conditions Tested. Oxygen uptake by activated sludge biomass was severely affected by most of the shock conditions tested, which is not surprising, because toxin doses were determined based on SOUR.

Cadmium detrimentally affected biomass oxygen uptake (Figures 4a and 4b). The SOUR inhibition relative to the control reactor reached 74, 84, and 92% in the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ 10- day SRT reactors and 93, 96, and 98% in the 2-day SRT reactors 7 hours after the shock. In subsequent cycles, the degree of inhibition continued to correlate with increases in contaminant dose. The SOUR levels started to increase immediately after cycle 1 and statistically recovered to control levels by cycle 15 (0.37 x SRT) in the 10-day SRT system. In the 2-day SRT system, the IC^ sub 15^ reactor recovered to control levels by cycle 17 (2.1 x SRT), but both the IC^ sub 25^ and IC^ sub 50^ reactors did not recover during the experiment.

The CDNB shock affected biomass SOUR to a lesser extent than cadmium (Figures 4c and 4d). The maximum respiration rate decrease relative to the control was 37, 64, and 89%, respectively, for the IC^ sub 15^, IC^ sub 25^, and IC^ sub 50^ 10-day SRT reactors and 14, 20, and 83% in the 2-day SRT system. The SOUR values in both systems returned to control levels in approximately 9 cycles (1.1 x SRT for the 2-day SRT system and 0.23 x SRT for the 10-day SRT system). As with cadmium, recovery of respiration rate was faster than recovery of COD removal.

As an uncoupler of oxidative phosphorylation, DNP was expected to cause an increase in OUR, because electron transfer across the electron transport chain tends to increase to compensate for the disruption of the proton motive force (PMF). The results obtained during this study indicate that, at high concentrations, DNP did not stimulate oxygen uptake, but instead had an inhibitory effect on respiration rates. In the 10-day SRT reactors (Figure 4e), the SOUR levels were 6, 21, and 52% lower than the control 7 hours after the shock, which is close to the short-term SOUR inhibition levels. Starting with cycle 11, all the reactors showed a significantly higher SOUR than the control, until the end of the experimental period (cycle 37), which indicated that, at lower concentrations- 1.1, 2.2, and 7.2 mg/L DNP in the IC^sub 15^, IC^sub 25^, and IC^sub 50^ reactors in cycle 11, respectively (data not shown)-the uncoupling effect was noticeable. This effect continued to be observed throughout the remaining cycles, even though DNP was not detected in the reactors after cycle 20 (data not shown). It is unclear why uncoupling continued to occur, even after DNP was no longer detected in the reactors. The 2-day SRT reactors (Figure 4f) responded differently to DNP, with SOUR measurements throughout the experiment showing that moderate inhibition of oxygen uptake was caused by DNP shock. The initial concentrations clearly inhibited cellular respiration, and the respiration rates remained lower than the control throughout the entire experimental period. Although no stimulation of respiration rates was observed for the 2-day SRT biomass, the results are consistent with what was observed at the beginning of the 10-day SRT DNP stress experiment and with the absence of biomass growth inhibition in this system. Contrary to cadmium, CDNB, and pH 1 1, DNP affected respiration rates for longer periods and to a greater extent than it affected COD removal or biomass flocculation. This is interesting to note, as DNP directly affects the respiratory chain of bacterial cells by disrupting the PMF.

Interestingly, like DNP, the greatest effect of cyanide shock was on the respiration rate of both biomasses tested (Figures 5a and 5b). Again, this is presumably related to the fact that the mode of action of this toxin directly affects the bacterial respiratory chain, by inactivating key enzymes in that system. At the IC^sub 50^ shock levels, complexed cyanide resulted in long respiration inhibition effects in both systems, in comparison with most of the other toxic conditions. The increase in the SOUR of the control reactors with time during the complexed cyanide experiments was most likely the result of an initial inhibitory effect by zinc, which has been observed to reduce the SOUR of activated sludge (Lajoie et al., 2003).

The pH 11 shock had a severe, but short-lived, effect on biomass respiration (2-day SRT results in Figure 5c, 10-day SRT data not shown). In the 10-day SRT reactor, pH 11 shock caused an 86% inhibition of respiration 7 hours after the shock, while, in the 2- day SRT system, this value was 98%. Recovery to control levels was achieved in the 2-day SRT system in 8 cycles (1.0 x SRT), while a very fast recovery of respiration rate (less than 5 cycles, 0.1 3 x SRT) was observed in the 10-day SRT reactor. Relative to COD removal, SOUR recovery times were substantially faster. The results suggest that biomass respiration was severely inhibited by the initial pH values, but that cell death was not a major consequence of high pH shock, because the biomass was able to recover its original respiration rate when the pH levels in the reactor dropped to 8.0 and 8.3 in cycle 5 for the 10- and 2-day SRT systems, respectively (data not shown).

The other tested shock conditions, namely pH 5, pH 9, octanol, and ammonia shocks, had some inhibitory effect on the biomass SOUR, but those effects were modest and/or short-lived. At the highest concentration tested, ammonia shock inhibited respiration in the 2- day SRT system to some extent, and this was one of the only process effects observed for ammonia stress (Figure 5d).

Mild Effect of Toxins on Biomass Settleability. We observed a deterioration in biomass settleability resulting from shock events by DNP, pH 5, and cyanide (data not shown). However, the increase in SVI never resulted in values greater than 1 35 mL/g MLVSS and 150 mL/ g MLVSS, for the 10- and 2-day SRT reactors, respectively. The only exception to this occurred in the 2-day SRT pH 5-shocked reactor, in which the SVI increased from 140 mL/g MLVSS in the control reactor to 190 mL/g MLVSS in the stressed reactor. Therefore, in most of the conditions tested and even in the reactors in which negative effects were observed, the biomass settling characteristics were not affected to an extent that would severely compromise its settleability in a final clarification step.

Biomass Dewaterability Significantly Affected by pH 11 Shock. For most of the toxic conditions tested, the effects on dewaterability were modest and short-lived. The pH 1 1 shock was the only condition that substantially decreased the dewaterability of the mixed liquor immediately after the shock, as CST increased approximately 7 times (to 307 seconds) and 9 times (to 135 seconds) in comparison with the control reactor (45 and 14 seconds), for the 2-and 10-day SRT reactors, respectively (data not shown). In addition, the CST of mixed liquor from the 10-day SRT IC^sub 50^ reactor shocked with cadmium remained elevated relative to the control levels for a period of approximately 3 x SRT, although the magnitude of the effect was fairly small (data not shown).

Overall, as with settleability, effects on dewaterability tended to be mild, and, for all of the toxins tested except pH 1 1 , it is not likely that the biosolids dewatering operations at a wastewater treatment facility would be severely affected by any of these toxic conditions. As CST is typically conducted with thickened biosolids, the results from our studies most likely would have yielded greater differences between the CSTs of the control versus the shocked reactors if the CSTs had been determined with concentrated mixed liquor. Therefore, caution should be used when looking at these results.

Discussion

To our knowledge, this study is the first to systematically investigate the effects of toxins on the activated sludge process to establish comprehensive source-effect relationships. This work investigated a wide range of process effects caused by industrially relevant classes of chemical compounds (sources) on activated sludge systems. The effects were related to biomass metabolism (COD removal, OUR, and biomass growth) and to floe structure properties (flocculation ability, dewaterability, and settleability). Table 1 summarizes the severity of each shock, both in terms of the intensity of the effect and recovery time to control levels, for both the 2- and 10-day SRT activated sludges. A qualitative classification was used to allow an easier visualization of the main trends observed in the data.

Among the process effects studied, inhibition of respiratory functions and COD removal were the most affected processes. All the toxins tested detrimentally affected COD removal and respiration rates to some extent in both or just one of the biomasses used in this study. The next most prevalent process effect was loss of flocculation ability (increase in effluent TSS), followed by reduced biomass growth (decrease in MLVSS). In general, both mixed liquor settleability and dewaterability were not affected to a great extent by the toxins evaluated in this study. The only exception was the significant increase in CST caused by pH 11 stress.

The fact that SOUR was inhibited by most of the toxic conditions tested was not unexpected, as the initial criterion for selection of the toxin concentrations for shock experiments was based on respiration inhibition. However, it is important to note that respiration inhibition was not always accompanied by significant deterioration of other process effects. For example, in the case of ammonia, octanol, and cyanide shocks, although reduction in respiration rates was observed immediately after the shock for at least one of the mixed liquors used in this study, the effects on other process parameters, including those related to effluent quality, were very modest and recovered rapidly to control levels. This information is relevant for the optimization of upset early warning devices that rely on respiration rate measurements to trigger alarms in wastewater treatment plants. Cyanide, octanol, or ammonia shock loads in the incoming influent could potentially result in false positive alarms in plants using online respirometers to detect influent toxicity.

Interestingly, when significant effects were observed to occur at the level of COD removal, there was a clear tendency for that effect to take place with biomass deflocculation in the shocked reactors. An increase in soluble COD in the effluent of the SBR reactors can originate from different mechanisms, such as the following:

(1) Inhibition of catabolic functions, which results in a decrease in biomass substrate uptake ability;

(2) Passive release of intracellular soluble materials resulting from cell lysis;

(3) Active excretion of intracellular substances from specific stress responses; and

(4) Release of materials from the extracellular polymeric substances (EPS) matrix that embeds the bacterial cells in activated sludge floes.

For different toxins, it is likely that these different mechanisms contribute, to varying extents, to the observed effluent soluble COD increases. Given the relatively fast recovery of SOUR values to control levels, which occurred in 1 x SRT or less, in most cases, cell lysis probably did not occur, to a great extent, in the shocked reactors. The link between deflocculation and increases in soluble COD may come from the contribution of EPS materials released into the bulk liquid resulting from the deflocculation process. It is plausible that floe breakup leads to rearrangement of the polymer matrix (or vice-versa), and that at least part of that matrix is lost into the bulk liquid, leading to increases in soluble COD.

The COD removal efficiency and biomass flocculation ability were most severely affected by cadmium, pH 11, and CDNB shocks, in both the 2- and 10-day SRT system. This was observed both in terms of the intensity of the effect and the time it took for the stressed biomass to recover to control levels. Similarly, biomass growth was most severely affected by cadmium and pH 1 1 shocks, both in terms of the intensity of the effect and recovery time, but CDNB did not seem to inhibit growth to a significant extent, as only the 10-day SRT IC^sub 50^ biomass was affected by this chemical. The DNP also considerably inhibited biomass growth in the 10-day SRT system. Therefore, in the case of COD removal, effluent TSS levels and inhibition of growth (MLVSS), the intensity of an effect correlated well with the recovery time, as the strongest effects were also the ones that tended to take longer to recover. However, this was not observed for the effects on SOUR. As mentioned above, SOUR values decreased after the shock for most of the conditions tested. Overall, decreases in SOUR tended to be short-lived, and SOUR levels recovered to control levels much faster than COD removal efficiency or flocculation ability in the case of the compounds that caused the strongest inhibitory effects in those process parameters (cadmium, CDNB, and pH 11). However, there was a clear difference between the toxins that produced the most dramatic reductions in respiration rates (cadmium, pH 11 , and CDNB) and the toxins that induced the longest effects on SOUR inhibition (DNP and cyanide). This is interesting to note, as the modes of action of both DNP and cyanide directly target the respiratory functions and will most probably affect biomass respiration for as long as they persist in the system. For the case of cadmium, pH 11, and CDNB, the toxicity mechanisms at the cellular level are likely nonspecific, and respiration is affected as a consequence of generalized damage to cellular structures. Therefore, respiration rates should recover as soon as the concentration of the contaminant decreases and basic cellular functions are operational. In fact, for these three chemicals, recovery of respiration rates was probably a priority for the bacterial community, as SOURs recovered much faster than COD removal or biomass growth. This priority may be linked to the need to generate energy to restore other cellular processes.

Ultimately, the data obtained from the source-effect studies suggest that the effects of a toxic shock load on specific processes, such as COD removal, flocculation ability, respiration, or biomass growth, are intrinsically related to the toxicity mechanisms that each toxin elicits at the cellular level. A discussion of such toxicity mechanisms for each of the toxins used in this study is presented next.

Cadmium resulted in severe process upset, both in the 2- and 10- day SRT systems, and affected all the processes analyzed in this paper, to some extent. This is consistent with what is found in the literature regarding its effects on activated sludge systems and toxicity mechanisms induced at the molecular level. Both decreases in COD removal efficiency (Weber and Sherrard, 1980) and SOUR (Madoni et al., 1999) have been reported as effects of cadmium on activated sludge. In addition, cadmium-induced deflocculation was also found in previous studies (Bott and Love, 2001; Neufeld, 1976). Activated sludge deflocculation by electrophilic chemicals, such as cadmium and CDNB, has been shown to be linked to a microbial stress response called the glutathione gated potassium efflux (GGKE) (Bott and Love, 2002, 2004). For this reason, we have also measured soluble potassium levels in the effluent of the shocked reactors. An analysis and discussion of the effluent potassium levels in the stressed reactors and their correlation with effluent TSS concentrations is given in Henriques et al. (2004). In addition, cadmium was found to induce synthesis of GroEL, a heat shock protein, in activated sludge cultures stressed with concen- trations as low as 5 mg/L (Bott and Love, 2001). Cadmium has also been reported to interfere with numerous cellular functions, such as DNA mutagenesis (Hughes and Poole, 1989), enzyme inactivation, and complex formation with phosphate groups in membrane phospholipids (Collins and Stotzky, 1989). Given its numerous effects at the cellular level, it is not possible to determine what specific mechanisms led to each of the observed effects, with the only exception being deflocculation. Nevertheless, the inhibition trends tended to correlate with the presence of measurable soluble cadmium in the stressed reactors (data not shown), corroborating previous reports that indicate that the soluble form is the predominant stressor controlling cadmium toxicity (Collins and Stotzky, 1989; Hughes and Poole, 1989). Complete recovery was not achieved for the 2-day SRT SBRs, but was reached in the 10-day SRT stressed bioreactors within 3 x SRT. Recovery in the 10-day SRT systems was linked to washout of soluble cadmium from the reactors. Soluble cadmium in the 2-day SRT reactors was much higher on a moles per cell basis than in the 10-day SRT reactors and might explain why the 2-day SRT system did not recover as promptly.

From the three pH values selected to shock the SBR reactors, only pH 11 resulted in significant process upset. Extreme pH values, such as pH 11, can directly damage external cell structures outside the cytoplasmic membrane, such as flagella, chemoreceptors, and cell walls (Chong et al., 1997; Dilworth et al., 1999). The results presented here showed that, although severe inhibitory effects were noted for COD removal, flocculation ability, and biomass growth, which could be related to external cell structure damage, respiration inhibition recovered very rapidly, suggesting that cell lysis/death was not a major consequence of pH 11 shock. High pH is used as a technique to extract EPS from activated sludge floes (Frolund et al., 1996; Higgins and Novak, 1997). Therefore, pH 11 shock most likely released polymer molecules from the floes into the bulk solution, which is consistent with its effects on effluent soluble COD, effluent TSS, and CST. The case of pH 1 1 shock is probably the most obvious example of the connection between deflocculation and increases in soluble COD, as the solubilization of EPS molecules is the most likely mechanism behind such effects. The dramatic increase in CST values can be also linked to an increase in the concentration of biopolymers, which leads to an increase in the viscosity of the liquid.

The CDNB affected mostly COD removal and biomass flocculation, but its influence on biomass respiration rate was shortlived. The CDNB is an electrophilic chemical, and, as such, it has the potential to inflict oxidative damage to cells. Although reports on the effects of CDNB in activated sludge systems are scarce, studies on the effects of organic electrophilic chemicals on pure cultures of gram-negative bacteria indicate that electrophiles, such as CDNB, can damage DNA and proteins (Apontoweil and Berends, 1975a, 1975b; Ferguson et al., 1995), and, in the latter case, the mechanism typically involves reaction with thiol bonds (Ferguson et al., 1995, 1997; Mclaggan et al., 2000). As an electrophile, CDNB has been reported to elicit the GGKE stress response (Ferguson and Booth, 1998; Ferguson et al., 1995, 1998) and to form a conjugate with glutathione (Vuilleumier, 1997). The results for the CDNB source- effect experiments are consistent with these previous findings. Activated sludge deflocculation by CDNB has been previously connected to the GGKE mechanism (Bott and Love, 2002, 2004). In addition, although statistical recovery of COD removal took longer, the magnitude of COD removal inhibition in the 10-day SRT CDNB- stressed reactors was not severe. In association with the soluble CDNB (data not shown) and MLVSS data in these reactors, these results suggest that the GGKE protective mechanism might have a limited capacity in terms of the CDNB concentration/mass that it can take up, for the mechanism to be efficiently used by bacteria; for example, the IC^sub 15^ and IC^sub 25^ did not seem to be greatly affected, while the effects on the IC50 reactor, where part of the dosed CDNB was found in its free soluble form during the cycles after the shock, were more dramatic.

In both the 10- and 2-day SRT systems, the process that was most severely affected by DNP shock was respiration. Biomass growth was also strongly inhibited in the 10-day SRT system. Given its known mode of action as an uncoupler of oxidative phosphorylation, the results in the 10-day SRT are consistent with uncoupling between biomass growth and substrate uptake. Most of the literature that analyzes the effects of DNP on activated sludge systems is related to the reduction of biomass yield to reduce biosolids production in activated sludge plants (Liu, 2000; Mayhew and Stephenson, 1998). However, there are few reports on the inhibitory effects of DNP on different process parameters, such as the ones presented here. It was interesting to note that, at the initial high DNP concentrations, there was an inhibitory effect on oxygen uptake rather than a stimulation, both in the nitrifying (10-day SRT)and non-nitrifying (2-day SRT) systems, showing that, at high concentrations, other toxicity mechanisms rather than disruption of the PMF alone were probably occurring or that the energy content in the cells decreased to levels that did not allow the basic metabolic processes to proceed. Stimulation of oxygen uptake started to occur in the 10-day SRT system when the DNP levels in the reactors dropped to under 10 mg/L (data not shown), and elevated SOUR values remained even after the DNP concentrations decreased to below the detection limit. No stimulation of respiration was observed in the 2-day SRT system. The reason for such results is not clear. Similar to DNP, cyanide shock affected respiration functions more severely than COD removal or biomass growth. Weak cyanometal complexes, such as the zinc-cyanide complex used in this study, can dissociate extensively to the metal ion and cyanide ion/hydrogen cyanide, which creates a greater availability of the toxic form of cyanide and, therefore, a greater potential for biomass toxicity than strong cyanide complexes (Torrens, 2000). The results of the source-effect cyanide experiment are consistent with the primary toxicity mechanism identified for the toxin, which consists of chelation reactions with divalent and trivalent metals in metallic enzymes. Specifically, chelation of the iron center in the heme group of cytochrome oxidases, the terminal enzymes in the electron transport chain, prevents the reoxidation of the enzymes and leads to partial or complete inactivation of respiration (Arden et al., 1998; Knowles, 1988; Solomonson, 1981; Yoshikawa and Caughey, 1990).

Octanol shock had a very modest effect on both activated sludge systems. Given the weak effects that octanol induced on biomass respiration, COD removal, bioflocculation and growth, and its fast disappearance from the bulk liquid (data not shown), octanol was either consumed as a carbon source or became associated with the mixed liquor particles, which would reduce its bioavailability and toxicity. Data regarding effluent TSS and CST analysis in the shocked reactors showed that these parameters were moderately affected by octanol, which may suggest that sorption to activated sludge floes was the predominant effect during the early stages of the experiment. Hydrophobic compounds, such as octanol, have been found to insert to biological membranes and to change the membrane structure (Heipieper et al., 1994; Ingram, 1977; Sikkema et al., 1995). Therefore, such a mechanism could potentially result in a destruction and/or rearrangement of the interactions between different cells or between cells and EPS molecules within activated sludge floes, thereby causing alterations at the level of floe structure. This could result in deflocculation and increased water retention, which would explain the mild effects observed in both activated sludge systems.

Ammonia shock was the condition that had the least effect on both activated sludge systems. Because ammonium bicarbonate was used as the source of ammonia shock, the added alkalinity contributed to maintain a stable pH in the reactors (data not shown). Therefore, the results reported in this study pertain to the effects of ammonia (NH^sub 3^ and NH^sub 4^^sup +^) alone. No significant effects were noted on any of the monitored process parameters in the 10-day SRT system, while only minor effects were noted on respiration, COD removal, CST, and SVI on the 2-day SRT biomass. The brief effects on COD removal and respiration rates seem to suggest that inhibition of metabolic activities occurred, to some extent, and are consistent with previous studies, specifically in terms of a decrease in COD removal efficiency (Li and Zhao, 1999). We hypothesize that this inhibition may be related to free ammonia in the system, as free ammonia has been reported to inhibit some classes of bacteria, such as nitrifiers (Neufeld et al., 1980). The effects on SVI and CST could be related to an increase in the monovalent-to-divalent cation ratio in the mixed liquor (resulting from NH^sub 4^^sup +^ addition), which has been linked to a deterioration of the settleability properties of activated sludge (Novak, 2001).

A comparison between the source-effect matrices generated for the non-nitrifying (2-day SRT) and the nitrifying (10-day SRT) mixed liquor reveals that it is not possible to identify a general trend of increased sensitivity for either biomass, under the conditions used in the present work. In terms of the intensity of an effect, the 2-day SRT biomass seemed to be more sensitive to increases in soluble COD. The effects on SOUR and SVI were comparable for both biomasses, but the effects on MLVSS reduction and CST increase were not consistent between the two systems. However, in terms of recovery time of the shocked reactors to control levels, in most cases, the 2-day SRT biomass took longer to recover from the stress event than the 10-day SRT biomass, which indicates an increased susceptibility of the low SRT biomass to toxic shock loads. For some of the most intense effects produced by cadmium and pH 11 shocks, the 2-day SRT biomass was not able to recover to control levels during the experimental period of 3 x SRT, while the 10-day SRT biomass always recovered within the same period. While physiological and community structure differences between the two systems may contribute to this observed trend, the differences in floe structure between the two activated sludges are also noteworthy. Floe structure differences have been previously suggested to affect the toxicity response of activated sludge, with smaller floes leading to increased susceptibility to toxic conditions (Henriques et al., 2005).

It is important to note that, when a process effect did recover to control levels, for most cases, the recovery occurred in less than 1 to 1.5 x SRT (2-day SRT system) or 1 x SRT (10-day SRT system), which suggests that, in general, the biomass was able to react to the shock load and recover to a physiologically active state in a fairly short period. In other words, although this study shows that significant deterioration of treatment process efficiency can potentially occur as a result of upset events from many different sources, it also shows that activated sludge systems are likely to overcome the shock event within 1 to 2 x SRT. However, this trend does not hold if the effects of nitrification are considered, as nitrification was the process most detrimentally affected by chemical shock (Kelly et al., 2004).

Comprehensive assessments of the potential effects of a specific class of toxins on biological treatment processes are extremely important, because they may contribute to the development of both mitigation and prevention measures and/or technology. This work showed that COD removal, bioflocculation, biomass growth, and biomass respiration were inhibited, to different extents, by distinct classes of industrial chemical toxins. Interestingly, the process that was most severely affected by the different toxins varied with chemical class and seemed to be intrinsically related to the nature of the chemical and its predominant mode of action on bacterial cells. Therefore, it becomes important to understand the causal mechanisms behind the source-effect relationships for each chemical class, as this information may help develop smart biosensors that can differentiate between different chemical shock loads and prevent major process upset.

Conclusions

Based on the results presented above, the following conclusions were made:

* Cadmium, pH 11, and CDNB shocks adversely affected effluent soluble COD, effluent TSS, SOUR, and MLVSS levels relative to the control reactor. Cyanide and DNP shocks resulted in significant decreases in biomass SOUR, and respiration inhibition was the most affected process. Octanol, pH 5, pH 9, and ammonia shocks did not cause significant process upset.

* The processes that were most severely affected by the toxic shock loads were respiration, COD removal, and bioflocculation. Loss of COD removal and deflocculation tended to occur concurrently. Respiration inhibition did not necessarily translate into other process effects.

* The observed process effects were intrinsically related to the toxicity mechanisms elicited by the toxic source.

* In general, when recovery to control levels was observed after a shock event for respiration, COD removal, flocculation, biomass growth, settleability, and dewaterability, the biomass was able to overcome the shock in less than 2 x SRT.

Credits

This work was funded by the Water Environment Research Foundation (Alexandria, Virginia), project no. 01-CTS-2. I. D. S. Henriques received a fellowship (SFRH/BD/4689/2001) from the Fundacao para a Ciencia e a Tecnologia, Portugal, and R. T. Kelly received a Charles E. Via Fellowship. The authors would like to acknowledge the assistance of several undergraduate assistants and Julie Petruska and Jody Smiley during this work.

Submitted for publication March 15, 2006; accepted for publication September 25, 2006.

The deadline to submit Discussions of this paper is December 15, 2007.

References

American Public Health Association; American Water Works Association; Water Environment Federation (1998) Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association: Washington, D.C.

Apontoweil, P.; Berends, W. (1975a) Glutathione Biosynthesis in Escherichia coli K- 12 - Properties of Enzymes and Regulation. Biochim. Biophys. Acta, 399 (1), 1-9.

Apontoweil, P.; Berends, W. (1975b) Isolation and Initial Characterization of Glutathione-Deficient Mutants of Escherichia coli K- 12. Biochim. Biophys. Acta, 399 (1), 10-22.

Arden, S. R.; Sinor, J. D.; Potthoff, W. K.; Aizenman, E. (1998) SubunitSpecific Interactions of Cyanide with the N-Methyl-D- Aspartate Receptor. J. Biol. Chem., 273 (34), 21505-21511.

Berthouex, P. M.; Brown, L. C. (1994) Statistics for Environmental Engineers; Lewis Publishers: Boca Raton, Florida.

Boon, N.; Top, E. M.; Verstraete, W.; Siciliano, S. D. (2003) Bioaugmentation as a Tool to Protect the Structure and Function of an Activated Sludge Microbial Community Against a 3-Chloroaniline Shock Load. Appl. Environ. Microbiol., 69 (3), 1511-1520.

Bott, C. B.; Duncan, A. J.; Love, N. G. (2001) Stress Protein Expression in Domestic Activated Sludge in Response to Xenobiotic Shock Loading. Water Sci. Technol., 43 (1), 123-130.

Bott, C. B.; Love, N. G. (2004) Implicating the Glutathione- Gated Potassium Efflux System as a Cause of Electrophile-Induced Activated Sludge Deflocculation. Appl. Environ. Microbiol., 70 (9), 5569-5578. Bott, C. B.; Love, N. G. (2002) Investigating a Mechanistic Cause for Activated Sludge Deflocculation in Response to Shock Loads of Toxic Electrophilic Chemicals. Water Environ. Res., 74 (3), 306-315.

Bott, C. B.; Love, N. G. (2001) The Immunochemical Detection of Stress Proteins in Activated Sludge Exposed to Toxic Chemicals. Water Res., 35 (1), 91-100.

Buitron, G.; Schoeb, M. E.; Moreno-Andrade, L.; Moreno, J. A. (2005) Evaluation of Two Control Strategies for a Sequencing Batch Reactor Degrading High Concentration Peaks of 4-Chlorophenol. Water Res., 39 (6), 1015-1024.

Chong, N. M.; Pai, S. L.; Chen, C. H. (1997) Bioaugmentation of an Activated Sludge Receiving pH Shock Loadings. Bioresour. Technol., 59 (2-3), 235-240.

Collins, Y. E.; Stotzky, G. (1989) Factors Affecting the Toxicity of Heavy Metals to Microbes. In Metal Ions and Bacteria, Beveridge, T. J., Doyle, R. J. (Eds.); John Wiley & Sons: New York.

Dilworth, M. J.; Glenn, A. R.; Konings, W. N.; Booth, I. R.; Poole, R. K.; Krulwich, T. A.; Rowbury, R. J.; Stock, J. B.; Slonczewski, J. L.; Cook, G. M.; et al. (1999) Problems of Adverse pH and Bacterial Strategies to Combat It. Novartis Found. Symp., 221, 4-14; John Wiley & Sons: West Sussex, England.

Ferguson, G. P.; Booth, I. R. (1998) Importance of Glutathione for Growth and Survival of Escherichia coli Cells: Detoxification of Methylglyoxal and Maintenance of Intracellular K+. J. Bacteriol., 180 (16), 4314-4318.

Ferguson, G. P.; Creighton, R. I.; Nikolaev, Y.; Booth, I. R. (1998) Importance of RpoS and Dps in Survival of Exposure of Both Exponential- and Stationary-Phase Escherichia coli Cells to the Electrophile N-Ethylmaleimide. J. Bacteriol, 180 (5), 1030-1036.

Ferguson, G. P.; Mclaggan, D.; Booth, I. R. (1995) Potassium Channel Activation by Glutathione-S-Conjugates in Escherichia coli - Protection Against Methylglyoxal is Mediated by Cytoplasmic Acidification. MoI. Microbiol., 17 (6), 1025-1033.

Ferguson, G. P.; Nikolaev, Y.; Mclaggan, D.; Maclean, M.; Booth, I. R. (1997) Survival During Exposure to the Electrophilic Reagent N- Ethylmaleimide in Escherichia coli: Role of KefB and KefC Potassium Channels. J. Bacteriol., 179 (4), 1007-1012.

Frolund, B.; Palmgren, R.; Keiding, K.; Nielsen, P. H. (1996) Extraction of Extracellular Polymers from Activated Sludge Using a Cation Exchange Resin. Water Res., 30 (8), 1749-1758.

Heipieper, H. J.; Weber, F. J.; Sikkema, J.; Keweloh, H.; Debont, J. A. M. (1994) Mechanisms of Resistance of Whole Cells to Toxic Organic Solvents. Trends Biotechnol., 12 (10), 409-415.

Henriques, I. D. S.; Holbrook, R. D.; Kelly. R. T.; Love, N. G. (2005) The Impact of Floe Size on Respiration Inhibition by Soluble Toxicants - A Comparative Investigation. Water Res., 39 (12), 2559- 2568.

Henriques, I. D. S.; Kelly, R. T.; Love, N. G. (2004) Deflocculation Effects Due to Chemical Perturbations in Sequencing Batch Reactors. Water Sci. Technol., 50 (10), 287-294.

Higgins, M. J.; Novak, J. T. (1997) The Effect of Cations on the Settling and Dewatering of Activated Sludges: Laboratory Results. Water Environ. Res., 69 (2), 215-223.

Hughes, M. N.; Poole, R. K. (1989) Metals and Microorganisms; Chapman and Hall: New York.

Ingram, L. O. (1977) Changes in Lipid Composition of Escherichia coli Resulting from Growth with Organic Solvents and with Food Additives. Appl. Environ. Microbiol., 33 (5), 1233-1236.

Kelly, R. T.; Henriques, I. D. S.; Love, N. G. (2004) Chemical Inhibition of Nitrification in Activated Sludge. Biotechnol. Bioeng., 85 (6), 683-694.

Kjellerup, B. V.; Keiding, K.; Nielsen, P. H. (2001) Monitoring and Troubleshooting of Non-Filamentous Settling and Dewatering Problems in an Industrial Activated Sludge Treatment Plant. Water Sci. Technol., 44 (2-3), 155-162.

Knowles, C. (1988) Cyanide Utilization and Degradation by Microorganisms. Ciba Found. Symp., 40, 3-15.

Kong, Z.; Vaerewijck, M.; Verstraete, W. (1996a) On-Line StBOD Measurement and Toxicity Control of Wastewaters with a Respirographic Biosensor. Environ. Technol, 17 (4), 399-406.

Kong, Z.; Vanrolleghem, P.; Willems, P.; Verstraete, W. (1996b) Simultaneous Determination of Inhibition Kinetics of Carbon Oxidation and Nitrification with a Respirometer. Water Res., 30 (4), 825-836.

Lajoie, C. A.; Lin, S. C; Kelly, C. J. (2003) Comparison of Bacterial Bioluminescence with Activated Sludge Oxygen Uptake Rates During Zinc Toxic Shock Loads in a Wastewater Treatment System. ASCE J. Environ. Eng., 129 (9), 879-883.

Li, X. Z.; Zhao, Q. L. (1999) Inhibition of Microbial Activity of Activated Sludge by Ammonia in Leachate. Environ. Int., 25 (8), 961- 968.

Liu, Y. (2000) Reduced Growth Yield of Activated Sludge in Organic Protonophore-Containing Batch Culture. Microb. Ecol., 39 (2), 168-173.

Love, N. G.; Bott, C. B. (2000) A Review and Needs Survey of Upset Early Warning Devices, Project 99-WWF-2; Water Environment Research Foundation: Alexandria, Virginia.

Love, N. G; Henriques, I. D. S.; Kelly, R. T. (2005) Upset Early Warning Systems for Biological Treatment Processes - Source and Effect Relationships, Report No. 01-CTS-2; Water Environment Research Foundation: Alexandria, Virginia.

Madoni, P.; Davoli, D.; Guglielmi, L. (1999) Response of SOUR and AUR to Heavy Metal Contamination in Activated Sludge. Water Res., 33 (10), 2459-2464.

Mayhew, M.; Stephenson, T. (1998) Biomass Yield Reduction: Is Biochemical Manipulation Possible without Affecting Activated Sludge Process Efficiency? Water Sci. Technol, 38 (8-9), 137-144.

Mclaggan, D.; Rufino, H.; Jaspars, M.; Booth, I. R. (2000) Glutathione-Dependent Conversion of N-Ethylmaleimide to the Maleamic Acid by Escherichia coli: An Intracellular Detoxification Process. Appl. Environ. Microbiol., 66 (4), 1393-1399.

Neufeld, R. D. (1976) Heavy Metals-Induced Deflocculation of Activated Sludge. J. Water Pollut. Control F

Source: Water Environment Research

More News in this Category