Bioresource Technology

Transcription

1 Bioresource Technology 129 (213) Contents lists available at SciVerse ScienceDirect Bioresource Technology journal homepage: Minimization of sludge production by a side-stream reactor under anoxic conditions in a pilot plant M. Coma a,, S. Rovira b, J. Canals b, J. Colprim a a LEQUIA, Institute of the Environment, University of Girona, Girona, Catalonia, Spain b GS Inima, Barcelona, Spain highlights " An anoxic side-stream reactor (SSR) at 15 mv minimizes sludge production. " Treating the whole sludge line in the SSR reduces the observed yield by 18.31%. " The applied sludge loading rate is the key parameter for sludge minimization. " Application of an SSR improves settleability and maintains removal efficiencies. graphical abstract article info abstract Article history: Received 23 August 212 Received in revised form 8 November 212 Accepted 11 November 212 Available online 27 November 212 Keywords: Activated sludge Energy uncoupling Side-stream reactor (SSR) Sludge reduction Yield This study evaluates the application of an anoxic side-stream reactor in the sludge return line of a conventional activated sludge system for the reduction of biomass production. The oxidation reduction potential was maintained at 15 mv while the applied sludge loading rate was modified by changing the percentage of return sludge treated in this reactor. The observed yield from the conventional system (.513 kg VSS kg 1 COD) was continuously reduced when the portion of return sludge treated was increased. A maximum reduction of 18.3% of the observed yield was obtained treating the whole sludge return line. The sludge age maintained through the experiment. The organic matter removal was not deteriorated, even improved, by the proposed plant modification. Thus, simply applying an anoxic side-stream reactor would decrease the final volume of waste sludge while maintaining the sludge retention time and would, in fact, decrease the economic costs in terms of sludge handling. Ó 212 Elsevier Ltd. All rights reserved. 1. Introduction Wastewater treatment by activated sludge has become the most used treatment for both urban and industrial wastewater influents. The major limitation of this technology is the excess of sludge generated, which increases the post-treatment and management costs. These costs can go up by 5 6% of the operating expenses in a wastewater treatment plant (WWTP) (Yang et al., Corresponding author. Address: Laboratory of Chemical and Environmental Engineering (LEQUIA), Institute of the Environment, University of Girona, Campus Montilivi s/n, Facultat de Ciències, E-1771 Girona, Catalonia, Spain. Tel.: ; fax: address: marta@lequia.udg.cat (M. Coma). 211). Furthermore, legislation on sludge disposal is becoming stricter. The biomass growth yield, which is around.4.6 kg VSS kg 1 COD (Tchobanoglous et al., 23), leads to high energy consumption in the processes used to reduce the excess sludge. Therefore, new sustainable strategies for sludge reduction are required (Khursheed and Kazmi, 211). Two differentiated approaches may be applied to reduce excess sludge disposed from WWTPs: (i) post-treatment methods, applying extra technologies into the waste flow, or (ii) process reduction methods, applying other strategies in the water line to reduce sludge production (Yang et al., 211). On the one hand, the most common methods of sludge stabilization for final disposal purposes are the biological processes of anaerobic mesophilic digestion and aerobic digestion at ambient /$ - see front matter Ó 212 Elsevier Ltd. All rights reserved.

2 23 M. Coma et al. / Bioresource Technology 129 (213) conditions (Borowski and Szopa, 27). Due to the high operating costs of the heating, post-treatment and chemical dosing requirements, anaerobic digestion is usually applied in large WWTPs where biogas co-production can recover the energy used. Aerobic digestion is typically used in smaller treatment facilities of less than 2, m 3 d 1, as it is more flexible in operation, less prone to process failure and has a low odour potential (Bernard and Gray, 2). Both anaerobic and aerobic digestion are implemented between the activated sludge and dewatering processes. However, digestion processes account for more than 57% of the power consumption because the existing sludge reduction technologies are capital intensive and process-wise complex (Khursheed and Kazmi, 211). On the other hand, an ideal approach to the sludge problem would be the reduction of excess sludge in the wastewater treatment rather than the post-treatment of the sludge produced. Microbial metabolism liberates a portion of the carbon from organic substrate in respiration and assimilates a portion into biomass. To reduce the production of biomass, wastewater processes must be engineered so that substrate is diverted from assimilation for biosynthesis to fuel exothermic, non-growth activities (Wei et al., 23). There are three possible solutions to create a feasible engineering approach to this problem: physico-chemical, mechanical and biological. The physico-chemical and mechanical methods are fairly well understood in how they might function, i.e. through the oxidation of organic material or the lysis of microbial material, thus making the overall mass more biodegradable; but the biological systems of sludge reduction, and the mechanisms behind them, are much less well understood (Khursheed and Kazmi, 211). Different strategies are currently under development for an engineering solution to sludge reduction based on several mechanisms: lysis-cryptic growth, uncoupling metabolism, maintenance metabolism and predation on bacteria (Wei et al., 23). Regarding biomass metabolism, adenosine triphosphate (ATP) plays an important role between the substrate oxidation process (ATP is produced during catabolism) and biomass synthesis reactions (anabolism). Uncoupling of both processes may be carried out when microorganisms are subjected to a physiological shock created by a lack of oxygen and substrate (i.e. oxic, anoxic or anaerobic cycling) and they use ATP as a source of energy. When they are returned to aerobiosis they rebuild their energy reserves at the expense of growth (Chudoba et al., 1992). The purposeful promotion of uncoupled metabolism can not only achieve the effective reduction of excess sludge production, but also does not require an important change in the configuration of the conventional activated sludge process (Ye and Li, 21). Chudoba et al. (1991) proposed a modification of the conventional activated sludge (CAS) process by inserting an anaerobic side-stream reactor (SSR) within the returned activated sludge (RAS) circuit after starvation conditions (settling tank). The oxic-settling-anaerobic (OSA) process may reduce the excess sludge production by 4 5%, according to Chudoba et al. (1992). The OSA process marks a significant difference compared to a conventional biological nitrogen removal system where anoxic or anaerobic tanks contain an adequate amount of substrate to achieve denitrification or phosphorus release, respectively (Saby et al., 23). Sludge reduction in the OSA system can be explained by sludge decay, which is accelerated effectively under low oxidation reduction potentials (ORP) in the anaerobic tank (Saby et al., 23). In addition, a short retention time stimulates uncoupled metabolism and decreases the cost of building the sludge tank (Ye et al., 28). Furthermore, it has been demonstrated that around half of the sludge production takes place in the aerobic tank when the sludge retention time (SRT) is increased by adding an anaerobic reactor (Chon et al., 211). Therefore, the sludge production might be reduced by applying smother conditions than anaerobic, allowing the maintenance of optimal SRT values for nutrient removal purposes. In this sense, Troiani et al. (211) studied different ORP values by modifying the hydraulic retention time (HRT) of the holding tank, so a higher HRT resulted in a lower ORP, and concluding that neither completely anaerobic nor completely aerobic conditions reduced sludge production. Finally, it must be stated that OSA-like processes have usually been applied to aerobic systems, but sludge reduction has not been tested with combined anaerobic, anoxic and aerobic processes for nutrient removal. Therefore, the BIMINEX process has been developed and evaluated in this study. This process integrates all conditions for nutrient removal by a CAS, such as a University of Cape Town (UCT) system, and biological minimization of sludge production. The main objective of this study is to evaluate the effect of an anoxic SSR at ambient temperature for the reduction of sludge production in a UCT system. The SSR was situated in the sludge return between the settling and the anaerobic tanks, contrarily to the OSA process in which it is returned to the aerobic tank. The control of the anoxic conditions (not as restrictive as anaerobic conditions) was carried out with an ORP set-point, regulated by punctual aeration. Different recirculation ratios were tested to evaluate the sludge reduction potential of the SSR at different applied sludge loading rates (ASLR) at a fixed ORP. 2. Methods 2.1. Pilot plant configuration The BIMINEX pilot plant was situated in La Garriga WWTP (La Garriga, Catalonia, Spain). The water line was based on a UCT configuration with a total volume of 6 m 3 (Fig. 1). Anaerobic, anoxic and aerobic reactor volumes were 1, 2 and 3 m 3, respectively. The settler presented a volume of.7 m 3. The return activated sludge (RAS) circuit was divided into two lines that allowed either direct recirculation of the sludge into the anoxic tank or treatment of the sludge in the anoxic SSR before being returned to the anaerobic reactor. The volume of the SSR was 1.4 m 3. The pilot plant was equipped with ph, ORP and dissolved oxygen (DO) probes and a programmable logic controller (PLC) which allowed monitoring and control of the system. Temperature was monitored but not controlled, so seasonal variations affected the pilot plant during the entire study Experimental procedure During the whole experimental period the HRT and the SRT of the water line were maintained around 24 h and 17 days, respectively. Waste activated sludge (WAS) was modified on a weekly basis to achieve the desired operational SRT. The internal recycle sludge lines were defined as four times the influent flow. The flow of the RAS was defined to be the same as the influent flow; however, different percentages of the sludge line were deviated into the SSR to be treated while the rest of the sludge was directly returned to the anoxic tank. Four configurations were tested: (i) % SSR, where RAS was completely returned to the anoxic tank; (ii) Fig. 1. Schematic diagram of the BIMINEX process.

3 M. Coma et al. / Bioresource Technology 129 (213) % SSR, where 1% of the RAS flow was treated in the SSR; (iii) 5% SSR, where half of the RAS was treated in the SSR; and (iv) 1% SSR, where RAS was completely treated in the SSR before being returned to the anaerobic tank. Table 1 summarizes the operating conditions of each configuration. Influent for the pilot plant was taken up from La Garriga WWTP after the grids and the sand trap chamber and before the primary settler. Table 2 summarizes the organic matter, nitrogen and phosphorus concentrations of this wastewater Analysis and calculations Chemical oxygen demand (COD), total suspended solids (TSS), volatile suspended solids (VSS), sludge volumetric index (SVI), ammonium, Kjeldahl nitrogen and alkalinity were measured according to Standard Methods (APHA, 25). Nitrites, nitrates and phosphates were analysed by ionic chromatography (Metrhom Ò 761-Compact; Metrosep A Supp 5; 25/4. mm). Total nitrogen was calculated as the sum of Kjeldahl nitrogen, nitrites and nitrates. The SRT was calculated with mean values of volatile suspended solids as stated in Eq. (1) in order to maintain a mean operating value of about 17 days. V VSS reactor SRT ¼ F W VSS W þ F ef VSS ef where V stands for the volume of the water line (6 m 3 ), F w and F ef correspond to the waste and effluent flows (m 3 d 1 ), respectively; and VSS reactor, VSS w and VSS ef correspond to the volatile suspended solids in the water line, the wastage and the effluent (mg L 1 ), respectively. Total SRT was calculated taking into account the anoxic side-stream reactor as stated in Eq. (2). Total SRT ¼ V VSS reactor þ V SSR VSS SSR F W VSS W þ F ef VSS ef where V SSR and VSS SSR stand for the volume (1.7 m 3 ) and the volatile suspended solids of the side-stream reactor, respectively. In this study the observed sludge yield (Y obs ) was calculated to evaluate the sludge production, taking into account the growth and death of the biomass. Since solids concentrations in both the UCT and SSR systems changed, cumulative terms during the period of study were used to quantify changes in both solids and substrates Eq. (3). For this reason, COD and biomass concentrations (VSS) were quantified daily. Y obs ¼ RVSSproduced RCOD removed P ðfw VSS W þ F ef VSS ef þ DVSS system Þ kg VSS ¼ P ¼ ðfin ðcod in COD ef ÞÞ kg COD where F w, F in and F ef correspond to the waste, influent and effluent flows (m 3 d 1 ), respectively; COD in and COD ef correspond to the total organic matter in the influent (including the solids in the influent) and the soluble organic matter in the effluent ð1þ ð2þ ð3þ (kg COD m 3 ), respectively; VSS w and VSS ef correspond to the volatile suspended solids in the waste and effluent (kg VSS m 3 ), respectively; and DVSS system corresponds to the biomass accumulated in the system (kg VSS d 1 ) and calculated as the daily variations of VSS in the three tanks of the reactor, in the SSR and in the settler, as described in Eq. (4). DSSV system ¼ DSSV reactor þ DSSV SSR þ DSSV settler Daily values were evaluated during a long period of time to take into account the stability of the system. The cumulative organic matter removal and the sludge production were calculated taking into account the variation of the previous day. Biomass production was obtained by regression of the experimental data, using a method similar to that of Chon et al. (211). Y obs accounted for the slope of the regression line of the cumulative biomass production versus the cumulative organic removal and it is considered the mean experimental value for the observed biomass yield. The ASLR was calculated taking into account the volatile suspended solids (kg m 3 ), the flow (m 3 d 1 ) and the volume (m 3 ) from the SSR as stated in Eq. (5). ASLR ¼ VSS SSR F SSR V SSR 3. Results 3.1. Stability of organic matter removal with the SSR Influent wastewater presented a total COD concentration from 15 to 9 mg O 2 L 1 with a mean value of 464 mg O 2 L 1. Despite the wide variability of the wastewater during all the experimental periods, the maximum amount of soluble organic matter presented in the effluent was 87 mg O 2 L 1, lower than the limits of discharge allowed by legislation (125 mg O 2 L 1 ; 91/271/EC). The organic matter removals obtained for each period of study were 87%, 88%, 92% and 93% for % SSR, 1% SSR, 5% SSR and 1% SSR, respectively. The slight improvement of organic matter removal was attributed to an increase of temperature from % SSR to 1% SSR (Table 2). Therefore, the application of a SSR itself did not affect the treatment performance of the system Maintenance of biomass concentration when applying the SSR The SRT from the water line was maintained for about 17 days, taking into account the effluent solids and regulating the sludge waste. Because of that, solids within the water line and the SSR presented high variability. TSS from the water line and SSR systems and from the effluent are presented in Fig. 2 for all operational conditions. The biomass concentration within the water line was maintained at around 3 mg TSS L 1, except for the 1% SSR configuration which reached values over 4 mg TSS L 1. Mean values obtained were 279, 475, 2949 and 2439 mg TSS L 1 for % SSR, ð4þ ð5þ Table 1 Operating conditions of the pilot plant for all configurations tested. System Influent (m 3 h 1 ) HRT (h) SRT (days) Total SRT (days) Temperature ( C) % SSR Water line SSR. 1% SSR Water line SSR % SSR Water line SSR % SSR Water line SSR

4 232 M. Coma et al. / Bioresource Technology 129 (213) Table 2 Average and standard deviation of wastewater influent parameters. COD (mg O 2 L 1 ) Total Nitrogen (mg N L 1 ) Ammonium (mg N NH þ 4 L 1 ) Phosphorus (mg P PO 3 4 L 1 ) Solids (mg TSS L 1 ) Alkalinity (mg CaCO 3 L 1 ) 464 ± 3 53 ± 2 31 ± 9 3 ± 1 23 ± ± 115 mg TSS L -1 mg TSS L -1 mg TSS L % SSR 1% SSR 5% SSR 1% SSR WATER LINE SSR EFFLUENT Days Fig. 2. Total suspended solids within the water line, the SSR and the effluent from %, 1%, 5% and 1% SSR operation. 1% SSR, 5% SSR and 1% SSR configurations, respectively. Regarding the SSR system, biomass increased inside the reactor over time in the 1% SSR and 1% SSR configurations, reaching a mean value of 4941 and 5366 mg TSS L 1, respectively. Meanwhile, the solids concentration in the 5% SSR configuration remained stable during the entire performance with a mean value of 4267 mg TSS L 1. Finally, a bad settler design of the pilot plant did not allow proper sedimentation of the sludge, causing biomass washout through the effluent on some punctual occasions. The mean values of biomass in the effluent for all configurations were 44, 15, 118 and 128 mg TSS L 1 for % SSR, 1% SSR, 5% SSR and 1% SSR, respectively. These values increased with the increase of sludge treated within the SSR. Nevertheless, the application of the anoxic SSR did not worsen the settling properties of the whole system, as reported in the next section. calculated SVI curves for the % SSR, 5% SSR and 1% SSR configurations. The % SSR, which corresponded to a CAS operation with UCT, presented poor settling properties, as can be seen in Fig. 3 by the smooth decrease of the index during the curve analysis. The value of SVI at 3 min was 195 ml g 1 TSS. In contrast, both the 5% SSR and 1% SSR profiles presented a depletion of the index during the first 1 min and were stabilized at 3 min to 162 and 122 ml g 1 TSS, respectively. Therefore, the more sludge treated within the anoxic SSR, the lower the SVI obtained Minimization of sludge production by increasing the sludge treated in the SSR The cumulative biomass production was plotted against the cumulative organic matter removal for each operating condition. The slope of the regression line obtained showed, as described in Eq. (3), the observed yields from each system taking into account the growth and death of heterotrophic and autotrophic biomass. Fig. 4 depicts regression lines and observed yields obtained for each configuration. The % SSR configuration presented the highest observed yield (.513 kg VSS kg 1 COD). This was expected as this configuration represented a CAS system. When treating some of the RAS within an anoxic SSR, the observed yield decreased. In fact, the recirculation of 1% of the RAS to the SSR (1% SSR) reduced the Y obs from.513 to.434 kg VSS kg 1 COD. The increase to half of the volume of sludge treated to the SSR (5% SSR) decreased the Y obs even more to.332 kg VSS kg 1 COD. However, the experimental values obtained from the treatment of the entire RAS into the SSR (1% SSR) did not differ significantly from the 5% SSR configuration as the Y obs were similar (.327 versus.332 kg VSS kg 1 COD). According to the literature, the biomass yield is highly affected by temperature and sludge age (Tchobanoglous et al., 23). Therefore, the fact that the whole experimental period was carried out in an outdoor pilot plant must be taken into account for the final calculation of the observed yield. Seasonal variations affected the temperature of the water line, ranging from 7 to 27 C. Therefore, 3.3. Improvement of the settling properties by the SSR introduction Sludge volumetric index and settling curves were analysed during the experimental study for all the configurations applied, providing an indication of the settling properties of the sludge and how the SSR was affecting the system. Fig. 3 presents the Cummulative biomass production (Kg VSS) 8 8 Y % obs =.513 Kg VSS Kg -1 COD Y 1% obs =.434 Kg VSS Kg -1 COD 6 R 2 = R 2 = % SSR 1% SSR Y 5% obs =.332 Kg VSS Kg -1 COD Y 1% obs =.327 Kg VSS Kg -1 COD 6 R 2 =.99 6 R 2 = % SSR 1% SSR Cummulative organic matter removed (Kg COD) Fig. 3. SVI curves from %, 5% and 1% SSR operation. Fig. 4. Regression lines obtained from biomass production and organic matter removal from % SSR, 1% SSR, 5% SSR and 1% SSR configurations.

5 M. Coma et al. / Bioresource Technology 129 (213) this parameter should be included within the calculation of Y obs.to that end, the experimental value of Y obs was temperature normalised to 2 C (Y 2 C obs ) by an Arrhenius term as the ones used in activated sludge models (U 1 = U h 2 T where h = 1.29) (Gujer et al., 1999). Biomass reduction percentages were calculated by comparing the normalised yields to % SSR according to Eq. (6). Table 3 summarizes the results obtained. %Sludge reduction ¼ Y2 C obs; i%ssr Y 2 C obs; %SSR Y 2 C obs; %SSRE 1 ð6þ On the one hand, a higher temperature in the system would enhance the auto-digestion of the sludge, decreasing the observed yield. In this light, when normalizing the experimental values obtained, yields were reduced for the experiments carried out at less than 2 C and were increased when working at higher temperatures. Finally, as can be observed in Table 3, the treatment of 1% of the RAS within the anoxic SSR did not improve as much as observed in the preliminary results, but the recirculation of the entire sludge line into the SSR (1% SSR) decreased the sludge production by 18.31%. On the other hand, long sludge age results in increased energy consumption for maintenance, which leaves less energy for cell synthesis and reduces sludge production (van Loosdrecth and Henze, 1999). Even though a high SRT would enhance the yield reduction through auto-digestion, it would deteriorate nutrient removal processes. Therefore, the SRT in the water line was fixed and controlled to affect neither the yield reduction nor the removal efficiency. The introduction of the SSR increased the total SRT compared with the CAS scenario (Table 2). However, the minimum reduction of the observed yield (.22%) was observed from % SSR to 1% SSR with an increase of the total SRT from 16.5 to 23.3 days. A reduction of 8.88% of the observed yield was observed from 1% SSR to 5% SSR with exactly the same total SRT of 23.3 days. Therefore, the more sludge treated in the SSR, the lower biomass production obtained Effect of applied sludge loading rate in the SSR The application of an anoxic SSR in the sludge line reduced the overall sludge production of a UCT system. A major yield reduction was observed when increasing the percentage of RAS treated within the SSR, therefore decreasing the HRT. In fact, the decrease of the time in which the sludge was kept under anoxic and endogenous conditions was inversely proportional to the quantity of sludge under such conditions. In this light, the applied sludge loading rate (ASLR) in the SSR was evaluated against the normalised observed yield. Results are presented in Fig. 5. The ASLR increased from 3.26 kg VSS m 3 d 1 when applying 1% SSR to 7.31 kg VSS m 3 d 1 at 5% SSR and to kg VSS m 3 d 1 at 1% SSR conditions. The ASLR presented a linear behaviour with the reduction of Y 2 C obs, which confirmed the influence of the quantity of sludge treated in the SSR to the biomass Y obs 2ºC (Kg VSS Kg -1 COD) % SSR 1% SSR 5% SSR 2º C Yobs.46.4 ASLR = Applied Sludge Loading Rate (Kg VSS m -3 d -1 ) R 2 =.956 1% SSR Fig. 5. Normalized observed yields and applied sludge loading rate from % SSR, 1% SSR, 5% SSR and 1% SSR configurations. growth reduction. Theoretically, the yield at % SSR with kg VSS m 3 d 1 of ASLR was identified as.46 kg VSS kg 1 COD from the regression line. This value was similar to.43 kg VSS kg 1 COD obtained as the normalised experimental value, which confirmed the linear behaviour. Therefore, the ASLR, and not the hydraulic retention time, was the key parameter for sludge production reduction. Even though a lower HRT will increase the quantity of sludge treated within the SSR, ASLR will depend on the biomass concentration, and is therefore the operational condition to be regulated for sludge reduction Evaluation of operational conditions for minimization of sludge production The most accepted hypothesis for minimization of sludge production in the sludge line is the decoupling of anabolism and catabolism (Chudoba et al., 1992), but how that will be achieved is still uncertain. Table 4 summarizes different operating conditions tested in the literature for sludge production minimization. Some researchers proposed short retention times in the holding tank (Ye et al., 28), but the global operation, including the returned activated sludge (RAS), was not defined. Low ORP was also considered to be a key parameter for sludge reduction (Saby et al., 23), but not under completely anaerobic conditions (Troiani et al., 211). In most cases, variations in ORP values were obtained by modifying the HRT of the reactor, and the higher retention times enhanced anaerobic conditions. In this study, the ORP was fixed independently of the HRT, which ensures low values but without reaching anaerobic conditions. Novak et al. (27) even introduced the Cannibal process, which includes an inert solid removal device to increase sludge reduction, but at the cost of increasing operating and investment costs. Most of the studies from the literature presented higher sludge reduction results than the ones presented in this study, but some operating systems must be taken into account. First of all, all the experiments presenting sludge production reductions higher than Table 3 Experimental and temperature normalised observed yields and biomass reduction. HRT a (h) Y b obs (kg VSS kg 1 COD) Temperature c ( C) Y 2 C obs (kg VSS kg 1 COD) Sludge reduction (%) % SSR % SSR % SSR % SSR a b c d Hydraulic retention time in the SSR. Observed yield calculated from regression line of experimental data. Mean water line temperature during the period of study. Observed yield corrected at 2 C.

6 234 M. Coma et al. / Bioresource Technology 129 (213) Table 4 Literature review of sludge production reduction by OSA-like processes. Process a SRT (days) HRT (h) ORP (mv) Temperature ( C) Sludge reduction (%) Reference OSA Chudoba et al. (1992) OSA Chen et al. (23) OSA Chen et al. (23) OSA Chen et al. (23) OSA Saby et al. (23) OSA Saby et al. (23) OSA Saby et al. (23) CANNIBAL Novak et al. (27) CANNIBAL Novak et al. (27) OSA Ye et al. (28) OSA Ye et al. (28) OSA Ye et al. (28) OSA + TCS Ye and Li (21) ACSL >5 Room Troiani et al. (211) ACSL 1/ 5 Room 22 Troiani et al. (211) ACSL 2/+5 Room Troiani et al. (211) ACSL 3/ 1 Room Troiani et al. (211) ACSL < 4 Room Troiani et al. (211) ACSL / 5 Real 15 Troiani et al. (211) Low DO Yang et al. (211) BIMINEX Real.2 This study BIMINEX Real 8.9 This study BIMINEX Real 18.3 This study a OSA: Oxic Settling Anaerobic; CANNIBAL: OSA + Inert Solids Removal; TCS: Tetrachlorosalicylanilide; ACSL: Alternate Cycle process in the Sludge Line; BIMINEX: UCT + anoxic SSR. 2% were carried out in lab-scale reactors using synthetic wastewater and ideal operating conditions (Table 4). Secondly, the ones presenting reductions higher than 4% were operating at long SRT (Novak et al., 27; Saby et al., 23), which may deteriorate the nutrient performance, or adding external uncouplers such as TCS (Ye and Li, 21). Ye et al. (28) obtained results slightly better than in this study because the system was kept at a higher temperature, which decrease the biomass yield and enhances sludge minimization (Khursheed and Kazmi, 211). Troiani et al. (211) obtained values similar to this study when testing different conditions in batch tests. However, the sludge reduction percentage of their study was lower when they applied their proposed system in a WWTP, even working at long SRT. Finally, most of the experiments were performed in OSA configuration, which is based on the introduction of an anaerobic tank for sludge reduction in a complete aerated system. Contrarily, the proposed configuration in this manuscript is a variation of a UCT configuration, which contains anaerobic, anoxic and aerobic tanks for integrated biological nutrient removal. The modification of a conventional configuration by adding an anoxic SSR in the sludge line may considerably decrease the costs associated with sludge handling. According to this study, a 1% SSR configuration obtained the best results in biomass yield reduction. Therefore, all of the returned activated sludge should be treated within the SSR and no extra pumping would be required to return the sludge from the settler to the main reactor. Thus, the application of an anoxic SSR would decrease the final volume of waste sludge to maintain the same SRT in a biological nutrient removal WWTP and this would have a direct influence on reducing economic costs in terms of sludge handling. 4. Conclusions Sludge production can be minimized by inserting an anoxic side-stream reactor controlled at 15 mv in the sludge line of an UCT system. Different configurations treating part or all of the returned activated sludge were tested. It has been found that the higher the applied sludge loading rate in the reactor, the lower the observed yield obtained, making the treatment of the whole return line the best option to reduce sludge production. With this configuration, SRT was maintained, and so the treatment efficiencies, while sludge settleability was improved. Thus, an anoxic SSR would improve sludge treatment management. Acknowledgements This research was financially supported by the Spanish Government (CSD27-55, CTQ , CTQ /PPQ, 33/RN8/3.2 and IPT ) and GS Inima. References Bernard, S., Gray, N.F., 2. Aerobic digestion of pharmaceutical and domestic wastewater sludges at ambient temperature. Water Res. 34 (3), Borowski, S., Szopa, J.S., 27. Experiences with the dual digestion of municipal sewage sludge. Bioresour. Technol. 98 (6), Chen, G.-H., An, K.-J., Saby, S., Brois, E., Djafer, M., 23. Possible cause of excess sludge reduction in an oxic-settling-anaerobic activated sludge process (OSA process). Water Res. 37 (16), Chon, D-H., Rome, M.cN., Kim, Y.M., Park, K.Y., Park, C., 211. Investigation of the sludge reduction mechanism in the anaerobic side-stream reactor process using several control biological wastewater treatment processes. Water Res. 45 (18), Chudoba, P., Chevalier, J.J., Chang, J., Capdeville, B., Effect of anaerobic stabilization of activated sludge on its production under batch conditions at various S /X ratios. Water Sci. Technol. 23 (4 6), Chudoba, P., Chudoba, J., Capdeville, B., The aspect of energetic uncoupling of microbial growth in the activated sludge process-osa system. Water Sci. Technol. 26 (9 11), Gujer, W., Henze, M., Mino, T., van Loosdrecht, M., Activated sludge model No. 3. Water Sci. Technol. 39 (1), Khursheed, A., Kazmi, A.A., 211. Retrospective of ecological approaches to excess sludge reduction. Water Res. 45 (15), Novak, J.T., Chon, D.H., Curtis, B.-A., Doyle, M., 27. Biological solids reduction using the Cannibal process. Water Environ. Res. 79 (12), Saby, S., Djafer, M., Chen, G.-H., 23. Effect of low ORP in anoxic sludge zone on excess sludge production in oxic-settling-anoxic activated sludge process. Water Res. 34 (1), Tchobanoglous, G., Burton, F.L., Stensel, H.D., 23. Wastewater engineering: treatment and reuse, 4th ed. Metcalf and Eddy Inc. McGraw-Hill Higher Education, New York, USA. Troiani, C., Eusebi, A.L., Battistoni, P., 211. Excess sludge reduction by biological way: from experimental experience to a real full scale application. Bioresour. Technol. 12 (22), van Loosdrecth, M.C.M., Henze, M., Maintenance, endogenous respiration, lysis, decay and predation. Water Sci. Technol. 39 (1),

7 M. Coma et al. / Bioresource Technology 129 (213) Wei, Y., Van Houten, R.T., Borger, A.R., Eikelboom, D.H., Fan, Y., 23. Minimization of excess sludge production for biological wastewater treatment. Water Res. 37, Yang, S.-S., Guo, W.-Q., Zhou, X.-J., Meng, Z.-H., Liu, B., Ren, N.-Q., 211. Optimization of operating parameters for sludge process reduction under alternating aerobic/ oxygen-limited conditions by response surface methodology. Bioresour. Technol. 12 (21), Ye, F.-X., Zhu, R.-F., Li, Y., 28. Effect of sludge retention time in sludge holding tank on excess sludge prosuction in the oxic-settling-anoxic (OSA) activated sludge process. J. Chem. Technol. Biotechnol. 83 (1), Ye, F., Li, Y., 21. Oxic-settling-anoxic (OSA) process combined with 3,3,4,5- tetrachlorosalicyanilide (TCS) to reduce excess sludge production in the activated sludge system. Biochem. Eng. J. 49 (2),