This article was downloaded by: [UPM] On: 24 December 2014, At: 01:43 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Environmental Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tent20
Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal a
a
b
c
Zhen Zhou , Can Xing , Zhichao Wu , Fei Tong & Junru Wang
c
a
College of Environmental and Chemical Engineering, Shanghai University of Electric Power, Shanghai 200090, China b
State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China c
Shanghai Chentou Wastewater Treatment Co., Ltd., Shanghai 201203, China Published online: 25 Oct 2013.
Click for updates To cite this article: Zhen Zhou, Can Xing, Zhichao Wu, Fei Tong & Junru Wang (2014) Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal, Environmental Technology, 35:6, 766-772, DOI: 10.1080/09593330.2013.850519 To link to this article: http://dx.doi.org/10.1080/09593330.2013.850519
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Environmental Technology, 2014 Vol. 35, No. 6, 766–772, http://dx.doi.org/10.1080/09593330.2013.850519
Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal Zhen Zhoua∗ , Can Xinga , Zhichao Wub , Fei Tongc and Junru Wangc a College of Environmental and Chemical Engineering, Shanghai University of Electric Power, Shanghai 200090, China; b State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China; c Shanghai Chentou Wastewater Treatment Co., Ltd., Shanghai 201203, China
Downloaded by [UPM] at 01:43 24 December 2014
(Received 18 December 2012; final version received 25 September 2013 ) The Unitank process combines the advantages of traditional continuous-flow activated sludge processes and sequencing batch reactors, and has been extensively employed in many wastewater treatment plants (WWTPs) in China. Biological phosphorus removal (BPR) of a full-scale Unitank WWTP was optimized by increasing anaerobic time from 80 to 120 min in an operation cycle of 360 min and reducing solid retention time (SRT) from 21.3 to 13.1 d. The BPR efficiency of the full-scale Unitank system increased from 63.8% (SRT of 21.3 d) to 83.2% for a SRT of 13.1 d. When the anaerobic time increased from 80 to 120 min, the net anaerobic phosphorus release amount increased from 0.25 to 1.06 mg L−1 , and sludge phosphorus content rose from 13.8 to 15.0 mgP·(gSS)−1 . During half an operation cycle, the average specific phosphorus release rate increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP·(gVSS·h)−1 in 40–60 min. Reducing SRT and increasing anaerobic time account for 84.6% and 15.4% in the total increment of phosphorus removal of 1.15 mg L−1 . Keywords: wastewater treatment; Unitank process; biological phosphorus removal; optimization; solid retention time
Introduction To prevent eutrophication in waters, several biological phosphorus (P) removal (BPR) processes, such as anaerobic/anoxic/aerobic (AAO),[1–3] anoxic/anaerobic/ aerobic,[4] University of Cape Town,[3,5] the sequencing batch reactor (SBR) [6,7] and oxidation ditch,[8] have been developed and applied in full-scale wastewater treatment plants (WWTPs). As an integrated biological wastewater treatment process combine the characteristics of continuous-flow activated sludge processes and the SBR, the Unitank process was developed in the 1990s and provided with advantages of compact structure, small footprint and low cost. The Unitank process has been extensively employed in many WWTPs, especially in Pearl River Delta [9–11] and Yangtse River Delta [12,13] of China. In recent years, researchers reported that the Unitank process was difficult to form a typical anaerobic P release environment and thus caused lower phosphorus removal efficiency.[14,15] Total phosphorus (TP) concentrations in the effluent of some full-scale Unitank WWTPs were usually higher than 1.0 mg L−1 . Therefore, auxiliary technologies such as adding independent anaerobic zone [10,11] or chemical phosphorus removal [14] were employed to decrease the TP in the effluent. However, few studies have been reported on the operational characteristics and optimization methods of the BPR in the Unitank process. ∗ Corresponding
author. Email: [email protected]
© 2013 Taylor & Francis
To achieve the objectives of this study, a full-scale Unitank WWTP in Shanghai was surveyed to collect operational data related to BPR, and then optimized by increasing wastage sludge and prolonging the anaerobic time. A phosphorus mass balance model was also put forward to evaluate the contributions of these two optimization measures. The results are expected to provide sound understandings of the BPR characteristics and optimization methods for the Unitank process. Materials and methods The full-scale Unitank WWTP The full-scale Unitank WWTP in Shanghai was built in 2004 and receives the wastewater of approximately 700,000 inhabitant equivalents from a combined sewer system. The WWTP, with a designed capacity of 400,000 m3 ·d−1 , includes a rotational flow grit chamber, 12 Unitank tanks and a sludge retention tank. Twelve Unitank tanks were divided into four groups (labelled 1#, 2#, 3# and 4#) with independent waste sludge discharge system and effluent channel. The waste activated sludge is drawn off from the sludge retention tank and sent to sludge thickener and dewatering machines. Each Unitank process is divided into three tanks (shown in Figure 1) with the same dimension of 35 × 35 × 7 m3 and
Environmental Technology
767 Effluent
Influent
Side tank A Figure 1.
Side tank C
Schematic diagram of the Unitank process.
Table 1.
Operation matrix of the full-scale Unitank WWTP.
Time (min)
Downloaded by [UPM] at 01:43 24 December 2014
Middle tank B
0–40 40–135 135–155 155–180 180–220 220–315 315–335 335–360
Side tank A
Middle tank B
Side tank C
Feeding and agitation Feeding and aeration Aeration Settling Settling, draining and sludge discharging
Aeration Aeration Feeding and aeration Feeding and aeration Aeration Aeration Feeding and aeration Feeding and aeration
Settling, draining and sludge discharging
effective depth of 6 m. The designed sludge load and sludge concentration are 0.11 kg BOD5 ·(kgSS·d)−1 and 4.0 g·L−1 , respectively. The designed hydraulic retention time (θ ) is 15.9 h, and the hydraulic load of side tank (used as secondary clarifier) with inclined plates is 1.13 m3 ·(m2 ·h)−1 . The dissolved oxygen (DO) in the aerobic stage was controlled at about 2.0–3.0 mg L−1 . The operation matrix of the full-scale Unitank process before optimization is shown in Table 1. Optimization and analytical method Operation data from 25 December 2011 to 7 February 2012 (stage 1, S1) were collected as control data for optimization. The optimization study was commenced on 8 February 2012 in the full-scale Unitank WWTP. From the commence day to 23 March 2012 (stage 2, S2), the wastage sludge discharge was increased to reduce the solid retention time (SRT), and the time for feeding and agitation (in Table 1), namely anaerobic time, was prolonged from 80 to 120 min in a cycle. During stage 3 (S3) from 24 March 2012 to 7 May2012, the wastage sludge discharge was further enhanced, and the anaerobic time was kept at 120 min at a cycle. The average temperatures of mixed liquor in the Unitank system during the three stages were, respectively, 14.1 ± 1.1◦ C, 13.2 ± 0.8◦ C and 18.8 ± 2.4◦ C due to seasonal variation. Twenty-four-hour composite samples were regularly collected every day to analyse pollutants in influent and effluent of the Unitank WWTP. Activated sludge samples were regularly collected from 36 Unitank tanks at aeration stage to determine mixed liquor suspended solid (MLSS) and mixed volatile suspended solids (MLVSS). To analyse
Feeding and agitation Feeding and aeration Aeration Settling
in-process variations of phosphorus and nitrogen during the agitation stage, grab samples were regularly collected at the point of 3 m below the water level in side tank of the Unitank process at 0, 10, 20, 30, 40, 50 and 60 min. Samples were filtered and taken to determine relevant compositions. Measurements of chemical oxygen demand (COD), 5-day biological oxygen demand (BOD5 ), ammonium nitrogen (AN), oxidized nitrogen (nitrite nitrogen + nitrate nitrogen, NOx -N), total nitrogen (TN), TP, suspended solids (SS), MLSS and MLVSS were performed according to Chinese NEPA standard methods.[16] Activated sludge samples for sludge P content measurement were collected at the end of aerobic stage when sufficient P uptake by phosphate-accumulating organisms (PAOs) occurred, and TP concentrations of mixed liquor and filtrate from samples were measured. The difference between two TP concentrations divided by the MLSS yields the sludge P content. Continuous stirred tank reactor (CSTR) models To obtain net P release and uptake amount in the Unitank process, the contribution of TP in the influent must be deducted by mass balance equation calculation. The side tank can be approximated as a CSTR, and the mass balance equation can be described as V
dC = Q0 C0 − Q0 C − VR, dt
(1)
where Q0 is the influent flow rate (m3 ·d−1 ), C0 and Ce are, respectively, the pollutant concentrations in the influent and
768
Z. Zhou et al.
effluent (mg L−1 ), V is the CSTR volume (m3 ), R is the reaction rate (mg·(Ld)−1 ), C is the pollutants concentration in the reactor (mg L−1 ) and t is the operation time (d). In the CSTR, Ce is equal to C; therefore, if the reaction item is not considered, Equation (1) can be deformed as dC (2) = Q0 C0 − Q0 C. dt Because θ is equal to V /Q0 , Equation (2) can be integrated as (3) C = C0 (1 − e−t/θ ) + m, V
Downloaded by [UPM] at 01:43 24 December 2014
where m is a constant and can be calculated by the measured pollutant concentrations at t = 0. Results and discussion Carbon source supply and SRT of the Unitank WWTP As shown in the operation matrix of Table 1, in an operation cycle of 360 min, wastewater was fed into side tank under anaerobic environment during 0–40 and 180–220 min, and into the side tank or middle tank under aerobic environment in the other time. This illustrates that 22.2% of the influent evolved through the anaerobic–aerobic (AO) process, and 77.8% of the influent were only treated by the aerobic process, causing high-level consumption of COD in the influent by oxygen. Therefore, only a small fraction of COD in the influent was utilized as carbon source for anaerobic P release, and thus the Unitank process is difficult to reach satisfying BPR efficiency. The increase of anaerobic time from 80 to 120 min could enhance the proportion of wastewater treated by AO process from 22.2% to 33.3%. In the full-scale Unitank system, the middle tank was always aerated and thus considered as a reaction tank, whereas the two side tanks were used as settling tank for 205 min during a cycle of 360 min. Since volumes of the three tanks were the same, the reaction volume accounted for 62.0% of the total volume. Taking both wastage sludge discharge and effluent SS into consideration, the SRT can be calculated as shown in the following equation: SRT =
0.62VX , Q e Xe + Q w X w
(4)
where V is the total volume of the Unitank system (m3 ), X , Xe and Xw are, respectively, SS in the reactor, effluent and Table 2.
Stage 2 Stage 3
Pollutants removal of the full-scale Unitank process Table 2 shows the average concentrations of pollutants in the influent and effluent of the full-scale Unitank WWTP during three stages. As shown in Table 2, the average COD and BOD5 in the effluent after optimization were slightly higher than those in the effluent in stage 1, and increment of COD and BOD5 were below 2.5 and 0.2 mg L−1 , respectively. Compared with stage 1, the average AN concentrations in the effluent rose by about 0.6 in both stage 2 and stage 3. The increase of both organic pollutants and AN in the effluent might be attributed to the decrease of 20 min aeration time; however, higher influent COD and BOD5 could also lead to the increase of COD and BOD5 in the effluent. Furthermore, the decrease of SRT after optimization probably caused the increase of organic pollutants and AN in the effluent.[17] Fluctuations of TP and TN in the influent and effluent of the full-scale Unitank WWTP during three stages are illustrated in Figure 2. Compared with stage 1, the average TN and TP in the effluent both decreased by about 24.0% both in stage 2, and respectively, 35.4% and 39.0% in stage 3. As shown in Figure 2, a relatively high platform of TP was observed in the period between 20th and 25th February during stage 2 owing to the malfunction of the aeration system. Without considering this period, AN and TP in the effluent were 1.54 ± 0.88 and 0.85 ± 0.23 mg L−1 in stage 2, respectively. From stage 1 to stage 3, COD and BOD5 in the influent gradually increased, while influent TN decreased by about 3 mg L−1 . This caused the gradual increase of C/N (COD/TN) ratio from 6.73 in stage 1 to 8.44 in stage 3. C/P ratios (COD/TP) were 87.06, 96.15 and 86.62 for the three stages, respectively. With the increase of C/N ratio, more nitrogen was removed and average TN in the effluent dropped from 12.55 mg L−1 at stage 1 to 8.11 mg L−1 at stage 3. Wu et al.,[2] who worked with an AAO system under different C/N ratios, also found that higher C/N ratio decreased the TN in the effluent. In the lab-scale UniFed SBR reactor reported by Zhao et al. [7] at C/N ratios of
Average concentration of pollutants in influent and effluent of the full-scale Unitank WWTP [in (mg L−1 )].
Pollutants Stage 1
wastage sludge (mg L−1 ) and Qe and Qw are the flow rates of the effluent and wastage sludge (m3 ·d−1 ). The calculated SRT of the Unitank WWTP during three stages were 21.3 ± 3.5, 14.5 ± 2.0 and 13.1 ± 1.8 d, respectively.
Influent Effluent Influent Effluent Influent Effluent
COD
BOD5
AN
SS
TN
TP
325 ± 92 41.2 ± 6.9 349 ± 124 43.5 ± 6.2 381 ± 106 42.5 ± 5.6
170 ± 56 4.3 ± 1.2 186 ± 67 4.5 ± 0.9 206 ± 65 4.5 ± 1.0
35.10 ± 5.65 1.23 ± 1.27 30.92 ± 6.63 1.87 ± 1.22 35.88 ± 8.41 1.82 ± 1.09
222 ± 59 16.7 ± 2.6 238 ± 67 15.5 ± 3.1 261 ± 59 16.7 ± 2.4
48.33 ± 9.65 12.55 ± 3.13 45.46 ± 9.38 9.55 ± 1.60 45.12 ± 9.71 8.11 ± 1.72
3.73 ± 0.87 1.23 ± 0.22 3.63 ± 1.24 0.93 ± 0.28 4.40 ± 1.04 0.75 ± 0.31
Environmental Technology 4.0
50
3.5
45 40
TN (mg·L -1)
3.0
TP (mg·L -1)
769
2.5 2.0 1.5
35 30 25 20 15
1.0
10
0.5
5
0.0 12/23 1/12
2/1
2/21
3/12
4/1
4/21
0 12/23
5/11
1/12
2/1
2/21
Influent/1.6
Figure 2.
Influent/1.6
Effluent
4/21
5/11
Effluent
(b) 22 20
50
Effluent SS (mg·L-1)
Effluent COD (mg·L-1)
55
Downloaded by [UPM] at 01:43 24 December 2014
4/1
TP and TN in the influent and effluent of the full-scale Unitank WWTP during process operation. (a)
45 40
18 16 14 12
35
10
30
8 1#
2# Stage 1
3# Stage 2
4#
Total
1#
(d) 2.0
4.0
1.8
3.5
1.6
Effluent TP (mg·L-1)
(c) 4.5 Effluent AN (mg·L-1)
2# Stage 1
Stage 3
3.0 2.5 2.0 1.5 1.0
3# Stage 2
4#
Total
Stage 3
1.4 1.2 1.0 0.8 0.6 0.4
0.5
0.2
0.0
0.0
1#
2# Stage 1
Figure 3.
3/12
Date
Date
3# Stage 2
4#
Total
1#
Stage 3
2# Stage 1
3# Stage 2
4#
Total
Stage 3
Average COD, SS, AN and TP in the effluent of each Unitank group.
6.5 and above, phosphate could not be detected in the effluent and complete phosphorus removal had been achieved. Yagci et al. [18] also reported a complete removal of phosphorus at a C/P ratio of 20. Nevertheless, the full-scale Unitank system with relative high C/N and C/P ratios only yields a BPR efficiency ranging from 67.0% to 83.0%, which was probably attributed to the low proportion of carbon source utilized for anaerobic P release. Comparison of effluent pollutants among Unitank groups Average COD, AN, TP and SS in the effluent of four Unitank groups during three stages were illustrated in Figure 3.
As shown in Figure 3(a), the average COD concentrations in effluents of the four groups during three stages all fluctuated between 41.1 and 45.1 mg L−1 , whereas the average SS concentrations ranged from 15.0 to 17.3 mg L−1 . There was no obvious difference in COD and SS in the effluent among four groups during three stages, indicating that the enhancement of anaerobic time and the increase of wastage sludge discharge had insignificant influence on COD degradation and solid–liquid removal efficiency. Compared with stage 1, the average AN concentrations in effluents of the four groups all increased in stage 2 (Figure 3(c)). Since the temperature in stage 2 was almost the same as that in stage 1, the slight increase of AN in the effluent is probably a result of reducing SRT. Average
Downloaded by [UPM] at 01:43 24 December 2014
Z. Zhou et al.
Influence of anaerobic time on BPR of the Unitank process The TP and NOx -N variations in side tank of Unitank group 3# during anaerobic phase (0–60 min) in stage 3 are shown in Figure 4. By increasing the anaerobic time from 40 to 60 min in half a cycle, NOx -N in the filtrate decreased from 3.97 to 3.25 mg L−1 owing to the increase of carbon source supply for denitrification. During the 60 min anaerobic stage, TP in the filtrate obviously increased from 0.27 to 1.95 mg L−1 . Nevertheless, owing to continuous feeding of wastewater in the Unitank process, the contribution of influent TP should be taken into consideration to calculate the net released P under anaerobic environment. The average influent TP of the Unitank system is 3.60 mg L−1 , and θ of the side tank is 5.3 h. The anaerobic stage is started at 0 min with initial TP of 0.27 mg L−1 . Substituting the above data to Equation (3), the theoretical TP variation during anaerobic stage is shown in the following equation: CTP = 3.87 − 3.60e−t/318 (5) The measured TP concentrations in Figure 4 subtracting calculated values of Equation (5) can obtain the net
2.0
5.5 TP Net released TP
1.6
5.0
NOx-N 1.2
4.5
0.8
4.0
0.4
3.5
0.0
NOx -N (mg·L-1)
AN concentrations in effluents of groups 1#, 2#, 3# and 4# in stage 3 were 1.45, 0.87, 0.42 and 0.50 mg L−1 higher than those in stage 1, and increments of average AN concentrations in groups 3# and 4# were lower than those in groups 1# and 2#, respectively. In fact, AN concentrations in mixed liquor of the side tank was below 1.0 mg L−1 after 75 min aeration. These results suggest that the enhancement of anaerobic time does not have obvious negative impact on nitrification efficiency, although this optimization method leads to the loss of 40 min aeration time. Variations of TP in effluents of groups 1# and 2# can be used to analyse the influence of SRT on BPR removal, and data in stages 1 and stage 3 were employed due to their similar C/P ratios. The average BPR efficiency of groups 1# and 2# increased from 63.8% for a SRT of 21.3 to 83.2% for a SRT of 13.1 d. The finding was in agreement with the study of Lee et al. [17] who reported that the BPR efficiency of an AAO SBR decreased from 47.1% (SRT of 5.9 d) to 31.0% for a SRT of 16.2 d. Average TP concentrations of effluents in groups 1# and 2# in stage 3 were 0.55 and 0.67 mg L−1 lower than that in stage 1, while the drop of average TP concentrations in groups 3# and 4# were 0.75 and 0.47 mg L−1 , respectively. As shown in Figure 3(d), the average TP concentrations in effluents of groups 1#, 2# and 3# were close (1.34, 1.36 and 1.40 mg L−1 , respectively), and higher than that of group 4# (1.15 mg L−1 ). Therefore, the drop in group 4# was lower than the other three groups. With similar effluent TP in stage 1, the drop in group 3# was higher than those in group 1# and group 2#, indicating that the enhancement of anaerobic time was able to improve the BPR of the Unitank process.
TP (mg·L -1)
770
3.0 0
10
20 30 40 Anaerobic time (min)
50
60
Figure 4. Variations of TP and NOx -N in side tank of the full-scale Unitank system.
anaerobic P release amount, as shown in Figure 4. After the anaerobic time increased from 80 to 120 min, the net anaerobic P release amount increased from 0.25 to 1.06 mg L−1 . In half of an operation cycle, the average specific phosphorus release rate (SPRR) increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP·(gVSS·h)−1 in 40–60 min. The anaerobic P release rate was enhanced obviously by prolonging the anaerobic time. The SPRR was significantly lower than the initial SPRR reported by Panswad et al. [19,20] which was because the reported SPRR was measured in the environment rich in carbon source. Although high C/P ratio was provided in the WWTP, the presence of NOx -N (Figure 4) and low distribution of COD in the anaerobic stage resulted in a phosphorus release process that probably limited by carbon source supply. Activated sludge samples were collected for sludge P content measurement from side tank of each Unitank group at the end of aerobic stage (155 or 335 min) in stage 3. The measured sludge P contents were illustrated in Figure 5. The average sludge P content of groups 3# and 4# with anaerobic time of 120 min was 15.0 mgP·(gSS)−1 , which was higher than that of groups 1# and 2# of 13.8 mgP·(gSS)−1 . These results suggest that the enhancement of anaerobic time was conducive to the accumulation of phosphorus in activated sludge. In two full-scale AAO WWTPs in Shanghai with C/P ratios above 60, measured sludge P contents were 27.40 ± 0.53 and 27.84 ± 0.56 mgP·(gSS)−1 (n = 3). In the AAO process, wastewater are first fed into the anaerobic stage and utilized as carbon source for phosphorus release; therefore, its BPR environment is easier to form than the Unitank process. Lee et al. [21] reported a lab-scale anaerobicintermittent aeration process with sludge P content of 44.2–59.5 mgP·(gSS)−1 at similar SRT (15 d) but lower C/P ratio (30). In the lab-scale AO process studied by Panswad et al. [20] the sludge P content was 53 mgP·(gVSS)−1 at C/P ratio of 50. Although enhanced to 15.0 mgP·(gSS)−1 by the improvement of anaerobic P release environment,
Environmental Technology
Sludge P content [mgP·(gSS)-1 ]
18
increasing sludge P content (CP,EPC ) can be calculated as
15.4 14.6
17
14.1 15
13.6
14 13 12 11 1#
Downloaded by [UPM] at 01:43 24 December 2014
MW (XP2 − XP1 ) . (9) Q0 According to the influent and effluent data in Table 2, the TP removed in stage 1 and stage 3 was 2.50 and 3.65 mg L−1 , and thus the increment of phosphorus removed by the two optimization methods was 1.15 mg L−1 . If the effect of SRT fluctuation on sludge P content is ignored, the average sludge P content of groups 1# and 2# can be regarded as background data to evaluate contributions of two optimization measures. Based on the background sludge P content of 13.8 mgP·(gSS)−1 and wastage sludge discharge in stage 1 and stage 3, the calculated CP,SRT was 0.99 mg·L−1 . When sludge P content rose from 13.8 mgP·(gSS)−1 to 15.0 mgP·(gSS)−1 by the enhancement of anaerobic time, the increment of P removed (CP,EPC ) was 0.18 mg L−1 . The sum of CP,SRT and CP,EPC was 1.17 mg L−1 , which was very close to the measured increment of P removed (1.15 mg L−1 ). In the total increment of phosphorus removal, phosphorus removed by reducing SRT contributes 84.6%, while that removed by the enhancement of anaerobic time accounts for 15.4%. These results suggest that the improvement of BPR efficiency was mainly the result of increasing wastage sludge discharge. CP,EPC =
16
Figure 5. (n = 6).
771
2# 3# Unitank group No.
4#
Sludge P contents in four Unitank groups in stage 3
the sludge P content was still significantly lower than the reported data. These results indicate that the typical BPR environment is difficult to form owing to the configuration and operation of the Unitank process.[10] Model-based evaluation of BPR optimization methods In BPR systems, phosphorus in wastewater was transferred to activated sludge by excess phosphate uptake of PAOs and anabolism of other biomass, and then removed by wastage sludge discharge. Therefore, the mass balance equation can be formulated as Q0 CP0 = Qe CPe + Qw Xw XP
(6)
Conclusions The BPR of a full-scale Unitank WWTP was optimized by increasing anaerobic time from 80 to 120 min (for group 3# and 4#) and reducing SRT from 21.3 to 13.1 d, and the main conclusions of the study can be summarized as follows.
(8)
(1) In the Unitank process, relatively low fraction of anaerobic time (feeding and agitation) in the total time of an operation cycle usually limits anaerobic P release, and then leads to relatively low BPR efficiency. (2) The increase of wastage sludge discharge can improve BPR effect. The BPR efficiency of the Unitank system increased from 63.8% (SRT of 21.3 d) to 83.2% for a SRT of 13.1 d. (3) In the Unitank process, the increase of anaerobic time can enhance anaerobic phosphorus release amount, and thus increase sludge P content. The average SPRR increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP· (gVSS·h)−1 in 40–60 min in half a cycle. (4) Reducing SRT and increasing anaerobic time account for 84.6% and 15.4% in the total increment of phosphorus removal of 1.15 mg L−1 .
If the sludge P content rises from XP1 to XP2 by optimization measures (such as increasing anaerobic time in the Unitank process), the increment of phosphorus removed by
Funding This work was supported by Chinese National 863 Program [2012AA063403]; Shanghai Chenguang Program
where CP0 and CPe are, respectively, TP concentrations in the influent and effluent (mg L−1 ) andXP is P content of wastage activated sludge [mgP·(gSS)−1 ]. The wastage sludge discharge flow rate is significantly lower than influent flow rate; therefore, the effluent flow rate could be assumed as influent flow rate. Equation (6) can be deformed as Qw Xw XP Mw X P CP0 − CPe = = , (7) Q0 Q0 where Mw is the wastage sludge discharge (g·d−1 ). The phosphorus removed by different optimization measures can be calculated by Equation (7). When the wastage sludge discharge increases from Mw1 to Mw2 , the increment of phosphorus removed by reducing SRT (CP,SRT ) can be calculated as CP,SRT =
(MW2 − MW1 )XP . Q0
772
Z. Zhou et al.
[2011CG60]; Innovation Program of Shanghai Municipal Education Commission [12YZ137]; 085 Project Program ‘Energy Storage Technologies of Intelligence Network’.
Downloaded by [UPM] at 01:43 24 December 2014
References [1] Ma Y, Peng Y, Wang X, Wang S. Nutrient removal performance of an anaerobic–anoxic–aerobic process as a function of influent C/P ratio. J Chem Technol Biotechnol. 2005;80:1118–1124. [2] Wu C, Peng Y, Wan X, Wang S. Performance and microbial population variation in a plug-flow A2 O process treating domestic wastewater with low C/N ratio. J Chem Technol Biotechnol. 2011;86:461–467. [3] Makinia J, Swinarski M, Dobiegala E. Experiences with computer simulation at two large wastewater treatment plants in northern Poland. Water Sci Technol. 2002;45: 209–218. [4] Guerrero J, Tayà C, Guisasola A, Baeza JA. Understanding the detrimental effect of nitrate presence on EBPR systems: effect of the plant configuration. J Chem Technol Biotechnol. 2012;87:1508–1511. [5] Zhou Z, Wu Z, Wang Z, Tang S, Gu G, Wang L, Wang Y, Xin Z. Simulation and performance evaluation of the anoxic/anaerobic/aerobic process for biological nutrient removal. Korea J Chem Eng. 2011;28:1233–1240. [6] Zhu R, Wu M, Yang J. Effect of sludge retention time and phosphorus to carbon ratio on biological phosphorus removal in HS-SBR process. Environ Technol. 2013;34: 429–435. [7] Zhao C, Peng Y, Wang S, Takigawa A. Effects of influent C/N ratio, C/P ratio and volumetric exchange ratio on biological phosphorus removal in UniFed SBR process. J Chem Technol Biotechnol. 2008;83:1587–1595. [8] Lim SH, Ko KB, Rho KE. Biological nutrient removal using an IACOD process: determining the combined effects of low temperature and long solids retention time. Environ Technol. 2006;27:467–475. [9] Zhang FG, Liu JX, Sui J. Sludge concentration dynamic distribution and its impact on the performance of UNITANK. J Environ Sci. (China) 2007;19:141–147.
[10] Li L, Lei M, Tao T, Chen Y. Study on anaerobic environment and optimization of UNITANK process. China Water Wastewater. 2006;22:86–89 (in Chinese). [11] Wang L, Miao H, Zhou Y. Commissioning and operation of UNITANK process in Dali wastewater treatment plant. China Water Wastewater. 2008;24:92–94 (in Chinese). [12] Dong S, Wang X, Wang W, Xu Z, Yao C. Non-steady stage of ortho-phosphate concentration in a Unitank process. Environ Chem. 2007;26:1–4 (in Chinese). [13] Zhang H, Huang W, Cheng L, Qiang X, Xu H, Jin J. Application of modified UNITANK process in extension project of Zhuji WWTP. China Water Wastewater. 2008;24:67–71 (in Chinese). [14] Qiu Y, Li Y, Zhou S, Wu S, Du H, Li Y. Analysis of operation effect of UNITANK process. Environ Sci Technol. 2010;33:124–127 (in Chinese). [15] Lei M, Li L, Su X, Li Y, Zhang Y. In-depth cognition on conventional Unitank process. Environ Eng. 2006;24:35–36 (in Chinese). [16] Chinese NEPA. Water and wastewater monitoring methods. Beijing, China: Chinese Environmental Science Publishing House; 2012. [17] Lee JK, Choi CK, Lee KH, Yim SB. Mass balance of nitrogen, and estimates of COD, nitrogen and phosphorus used in microbial synthesis as a function of sludge retention time in a sequencing batch reactor system. Bioresourc Technol. 2008;99:7788–7796. [18] Yagci N, Artan N, Cokgor EU, Randall CW, Orhon D. Metabolic model for acetate uptake by a mixed culture of phosphate and glycogen-accumulating organisms under anaerobic conditions. Biotechnol Bioengin. 2003;84:359– 373. [19] Panswad T, Doungchai A, Anotai J. Temperature effect on microbial community of enhanced biological phosphorus removal system. Water Res. 2003;37:409–415. [20] Panswad T, Tongkhammak N, Anotai J. Estimation of intracellular phosphorus content of phosphorus-accumulating organisms at different P:COD feeding ratios. J Environ Manag. 2007;84:141–145. [21] Lee D, Kim M, Chung J. Relationship between solid retention time and phosphorus removal in anaerobic-intermittent aeration process. J Biosci Bioengin. 2007;103:338–344.
Environmental Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tent20
Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal a
a
b
c
Zhen Zhou , Can Xing , Zhichao Wu , Fei Tong & Junru Wang
c
a
College of Environmental and Chemical Engineering, Shanghai University of Electric Power, Shanghai 200090, China b
State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China c
Shanghai Chentou Wastewater Treatment Co., Ltd., Shanghai 201203, China Published online: 25 Oct 2013.
Click for updates To cite this article: Zhen Zhou, Can Xing, Zhichao Wu, Fei Tong & Junru Wang (2014) Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal, Environmental Technology, 35:6, 766-772, DOI: 10.1080/09593330.2013.850519 To link to this article: http://dx.doi.org/10.1080/09593330.2013.850519
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Environmental Technology, 2014 Vol. 35, No. 6, 766–772, http://dx.doi.org/10.1080/09593330.2013.850519
Optimization of a full-scale Unitank wastewater treatment plant for biological phosphorus removal Zhen Zhoua∗ , Can Xinga , Zhichao Wub , Fei Tongc and Junru Wangc a College of Environmental and Chemical Engineering, Shanghai University of Electric Power, Shanghai 200090, China; b State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China; c Shanghai Chentou Wastewater Treatment Co., Ltd., Shanghai 201203, China
Downloaded by [UPM] at 01:43 24 December 2014
(Received 18 December 2012; final version received 25 September 2013 ) The Unitank process combines the advantages of traditional continuous-flow activated sludge processes and sequencing batch reactors, and has been extensively employed in many wastewater treatment plants (WWTPs) in China. Biological phosphorus removal (BPR) of a full-scale Unitank WWTP was optimized by increasing anaerobic time from 80 to 120 min in an operation cycle of 360 min and reducing solid retention time (SRT) from 21.3 to 13.1 d. The BPR efficiency of the full-scale Unitank system increased from 63.8% (SRT of 21.3 d) to 83.2% for a SRT of 13.1 d. When the anaerobic time increased from 80 to 120 min, the net anaerobic phosphorus release amount increased from 0.25 to 1.06 mg L−1 , and sludge phosphorus content rose from 13.8 to 15.0 mgP·(gSS)−1 . During half an operation cycle, the average specific phosphorus release rate increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP·(gVSS·h)−1 in 40–60 min. Reducing SRT and increasing anaerobic time account for 84.6% and 15.4% in the total increment of phosphorus removal of 1.15 mg L−1 . Keywords: wastewater treatment; Unitank process; biological phosphorus removal; optimization; solid retention time
Introduction To prevent eutrophication in waters, several biological phosphorus (P) removal (BPR) processes, such as anaerobic/anoxic/aerobic (AAO),[1–3] anoxic/anaerobic/ aerobic,[4] University of Cape Town,[3,5] the sequencing batch reactor (SBR) [6,7] and oxidation ditch,[8] have been developed and applied in full-scale wastewater treatment plants (WWTPs). As an integrated biological wastewater treatment process combine the characteristics of continuous-flow activated sludge processes and the SBR, the Unitank process was developed in the 1990s and provided with advantages of compact structure, small footprint and low cost. The Unitank process has been extensively employed in many WWTPs, especially in Pearl River Delta [9–11] and Yangtse River Delta [12,13] of China. In recent years, researchers reported that the Unitank process was difficult to form a typical anaerobic P release environment and thus caused lower phosphorus removal efficiency.[14,15] Total phosphorus (TP) concentrations in the effluent of some full-scale Unitank WWTPs were usually higher than 1.0 mg L−1 . Therefore, auxiliary technologies such as adding independent anaerobic zone [10,11] or chemical phosphorus removal [14] were employed to decrease the TP in the effluent. However, few studies have been reported on the operational characteristics and optimization methods of the BPR in the Unitank process. ∗ Corresponding
author. Email: [email protected]
© 2013 Taylor & Francis
To achieve the objectives of this study, a full-scale Unitank WWTP in Shanghai was surveyed to collect operational data related to BPR, and then optimized by increasing wastage sludge and prolonging the anaerobic time. A phosphorus mass balance model was also put forward to evaluate the contributions of these two optimization measures. The results are expected to provide sound understandings of the BPR characteristics and optimization methods for the Unitank process. Materials and methods The full-scale Unitank WWTP The full-scale Unitank WWTP in Shanghai was built in 2004 and receives the wastewater of approximately 700,000 inhabitant equivalents from a combined sewer system. The WWTP, with a designed capacity of 400,000 m3 ·d−1 , includes a rotational flow grit chamber, 12 Unitank tanks and a sludge retention tank. Twelve Unitank tanks were divided into four groups (labelled 1#, 2#, 3# and 4#) with independent waste sludge discharge system and effluent channel. The waste activated sludge is drawn off from the sludge retention tank and sent to sludge thickener and dewatering machines. Each Unitank process is divided into three tanks (shown in Figure 1) with the same dimension of 35 × 35 × 7 m3 and
Environmental Technology
767 Effluent
Influent
Side tank A Figure 1.
Side tank C
Schematic diagram of the Unitank process.
Table 1.
Operation matrix of the full-scale Unitank WWTP.
Time (min)
Downloaded by [UPM] at 01:43 24 December 2014
Middle tank B
0–40 40–135 135–155 155–180 180–220 220–315 315–335 335–360
Side tank A
Middle tank B
Side tank C
Feeding and agitation Feeding and aeration Aeration Settling Settling, draining and sludge discharging
Aeration Aeration Feeding and aeration Feeding and aeration Aeration Aeration Feeding and aeration Feeding and aeration
Settling, draining and sludge discharging
effective depth of 6 m. The designed sludge load and sludge concentration are 0.11 kg BOD5 ·(kgSS·d)−1 and 4.0 g·L−1 , respectively. The designed hydraulic retention time (θ ) is 15.9 h, and the hydraulic load of side tank (used as secondary clarifier) with inclined plates is 1.13 m3 ·(m2 ·h)−1 . The dissolved oxygen (DO) in the aerobic stage was controlled at about 2.0–3.0 mg L−1 . The operation matrix of the full-scale Unitank process before optimization is shown in Table 1. Optimization and analytical method Operation data from 25 December 2011 to 7 February 2012 (stage 1, S1) were collected as control data for optimization. The optimization study was commenced on 8 February 2012 in the full-scale Unitank WWTP. From the commence day to 23 March 2012 (stage 2, S2), the wastage sludge discharge was increased to reduce the solid retention time (SRT), and the time for feeding and agitation (in Table 1), namely anaerobic time, was prolonged from 80 to 120 min in a cycle. During stage 3 (S3) from 24 March 2012 to 7 May2012, the wastage sludge discharge was further enhanced, and the anaerobic time was kept at 120 min at a cycle. The average temperatures of mixed liquor in the Unitank system during the three stages were, respectively, 14.1 ± 1.1◦ C, 13.2 ± 0.8◦ C and 18.8 ± 2.4◦ C due to seasonal variation. Twenty-four-hour composite samples were regularly collected every day to analyse pollutants in influent and effluent of the Unitank WWTP. Activated sludge samples were regularly collected from 36 Unitank tanks at aeration stage to determine mixed liquor suspended solid (MLSS) and mixed volatile suspended solids (MLVSS). To analyse
Feeding and agitation Feeding and aeration Aeration Settling
in-process variations of phosphorus and nitrogen during the agitation stage, grab samples were regularly collected at the point of 3 m below the water level in side tank of the Unitank process at 0, 10, 20, 30, 40, 50 and 60 min. Samples were filtered and taken to determine relevant compositions. Measurements of chemical oxygen demand (COD), 5-day biological oxygen demand (BOD5 ), ammonium nitrogen (AN), oxidized nitrogen (nitrite nitrogen + nitrate nitrogen, NOx -N), total nitrogen (TN), TP, suspended solids (SS), MLSS and MLVSS were performed according to Chinese NEPA standard methods.[16] Activated sludge samples for sludge P content measurement were collected at the end of aerobic stage when sufficient P uptake by phosphate-accumulating organisms (PAOs) occurred, and TP concentrations of mixed liquor and filtrate from samples were measured. The difference between two TP concentrations divided by the MLSS yields the sludge P content. Continuous stirred tank reactor (CSTR) models To obtain net P release and uptake amount in the Unitank process, the contribution of TP in the influent must be deducted by mass balance equation calculation. The side tank can be approximated as a CSTR, and the mass balance equation can be described as V
dC = Q0 C0 − Q0 C − VR, dt
(1)
where Q0 is the influent flow rate (m3 ·d−1 ), C0 and Ce are, respectively, the pollutant concentrations in the influent and
768
Z. Zhou et al.
effluent (mg L−1 ), V is the CSTR volume (m3 ), R is the reaction rate (mg·(Ld)−1 ), C is the pollutants concentration in the reactor (mg L−1 ) and t is the operation time (d). In the CSTR, Ce is equal to C; therefore, if the reaction item is not considered, Equation (1) can be deformed as dC (2) = Q0 C0 − Q0 C. dt Because θ is equal to V /Q0 , Equation (2) can be integrated as (3) C = C0 (1 − e−t/θ ) + m, V
Downloaded by [UPM] at 01:43 24 December 2014
where m is a constant and can be calculated by the measured pollutant concentrations at t = 0. Results and discussion Carbon source supply and SRT of the Unitank WWTP As shown in the operation matrix of Table 1, in an operation cycle of 360 min, wastewater was fed into side tank under anaerobic environment during 0–40 and 180–220 min, and into the side tank or middle tank under aerobic environment in the other time. This illustrates that 22.2% of the influent evolved through the anaerobic–aerobic (AO) process, and 77.8% of the influent were only treated by the aerobic process, causing high-level consumption of COD in the influent by oxygen. Therefore, only a small fraction of COD in the influent was utilized as carbon source for anaerobic P release, and thus the Unitank process is difficult to reach satisfying BPR efficiency. The increase of anaerobic time from 80 to 120 min could enhance the proportion of wastewater treated by AO process from 22.2% to 33.3%. In the full-scale Unitank system, the middle tank was always aerated and thus considered as a reaction tank, whereas the two side tanks were used as settling tank for 205 min during a cycle of 360 min. Since volumes of the three tanks were the same, the reaction volume accounted for 62.0% of the total volume. Taking both wastage sludge discharge and effluent SS into consideration, the SRT can be calculated as shown in the following equation: SRT =
0.62VX , Q e Xe + Q w X w
(4)
where V is the total volume of the Unitank system (m3 ), X , Xe and Xw are, respectively, SS in the reactor, effluent and Table 2.
Stage 2 Stage 3
Pollutants removal of the full-scale Unitank process Table 2 shows the average concentrations of pollutants in the influent and effluent of the full-scale Unitank WWTP during three stages. As shown in Table 2, the average COD and BOD5 in the effluent after optimization were slightly higher than those in the effluent in stage 1, and increment of COD and BOD5 were below 2.5 and 0.2 mg L−1 , respectively. Compared with stage 1, the average AN concentrations in the effluent rose by about 0.6 in both stage 2 and stage 3. The increase of both organic pollutants and AN in the effluent might be attributed to the decrease of 20 min aeration time; however, higher influent COD and BOD5 could also lead to the increase of COD and BOD5 in the effluent. Furthermore, the decrease of SRT after optimization probably caused the increase of organic pollutants and AN in the effluent.[17] Fluctuations of TP and TN in the influent and effluent of the full-scale Unitank WWTP during three stages are illustrated in Figure 2. Compared with stage 1, the average TN and TP in the effluent both decreased by about 24.0% both in stage 2, and respectively, 35.4% and 39.0% in stage 3. As shown in Figure 2, a relatively high platform of TP was observed in the period between 20th and 25th February during stage 2 owing to the malfunction of the aeration system. Without considering this period, AN and TP in the effluent were 1.54 ± 0.88 and 0.85 ± 0.23 mg L−1 in stage 2, respectively. From stage 1 to stage 3, COD and BOD5 in the influent gradually increased, while influent TN decreased by about 3 mg L−1 . This caused the gradual increase of C/N (COD/TN) ratio from 6.73 in stage 1 to 8.44 in stage 3. C/P ratios (COD/TP) were 87.06, 96.15 and 86.62 for the three stages, respectively. With the increase of C/N ratio, more nitrogen was removed and average TN in the effluent dropped from 12.55 mg L−1 at stage 1 to 8.11 mg L−1 at stage 3. Wu et al.,[2] who worked with an AAO system under different C/N ratios, also found that higher C/N ratio decreased the TN in the effluent. In the lab-scale UniFed SBR reactor reported by Zhao et al. [7] at C/N ratios of
Average concentration of pollutants in influent and effluent of the full-scale Unitank WWTP [in (mg L−1 )].
Pollutants Stage 1
wastage sludge (mg L−1 ) and Qe and Qw are the flow rates of the effluent and wastage sludge (m3 ·d−1 ). The calculated SRT of the Unitank WWTP during three stages were 21.3 ± 3.5, 14.5 ± 2.0 and 13.1 ± 1.8 d, respectively.
Influent Effluent Influent Effluent Influent Effluent
COD
BOD5
AN
SS
TN
TP
325 ± 92 41.2 ± 6.9 349 ± 124 43.5 ± 6.2 381 ± 106 42.5 ± 5.6
170 ± 56 4.3 ± 1.2 186 ± 67 4.5 ± 0.9 206 ± 65 4.5 ± 1.0
35.10 ± 5.65 1.23 ± 1.27 30.92 ± 6.63 1.87 ± 1.22 35.88 ± 8.41 1.82 ± 1.09
222 ± 59 16.7 ± 2.6 238 ± 67 15.5 ± 3.1 261 ± 59 16.7 ± 2.4
48.33 ± 9.65 12.55 ± 3.13 45.46 ± 9.38 9.55 ± 1.60 45.12 ± 9.71 8.11 ± 1.72
3.73 ± 0.87 1.23 ± 0.22 3.63 ± 1.24 0.93 ± 0.28 4.40 ± 1.04 0.75 ± 0.31
Environmental Technology 4.0
50
3.5
45 40
TN (mg·L -1)
3.0
TP (mg·L -1)
769
2.5 2.0 1.5
35 30 25 20 15
1.0
10
0.5
5
0.0 12/23 1/12
2/1
2/21
3/12
4/1
4/21
0 12/23
5/11
1/12
2/1
2/21
Influent/1.6
Figure 2.
Influent/1.6
Effluent
4/21
5/11
Effluent
(b) 22 20
50
Effluent SS (mg·L-1)
Effluent COD (mg·L-1)
55
Downloaded by [UPM] at 01:43 24 December 2014
4/1
TP and TN in the influent and effluent of the full-scale Unitank WWTP during process operation. (a)
45 40
18 16 14 12
35
10
30
8 1#
2# Stage 1
3# Stage 2
4#
Total
1#
(d) 2.0
4.0
1.8
3.5
1.6
Effluent TP (mg·L-1)
(c) 4.5 Effluent AN (mg·L-1)
2# Stage 1
Stage 3
3.0 2.5 2.0 1.5 1.0
3# Stage 2
4#
Total
Stage 3
1.4 1.2 1.0 0.8 0.6 0.4
0.5
0.2
0.0
0.0
1#
2# Stage 1
Figure 3.
3/12
Date
Date
3# Stage 2
4#
Total
1#
Stage 3
2# Stage 1
3# Stage 2
4#
Total
Stage 3
Average COD, SS, AN and TP in the effluent of each Unitank group.
6.5 and above, phosphate could not be detected in the effluent and complete phosphorus removal had been achieved. Yagci et al. [18] also reported a complete removal of phosphorus at a C/P ratio of 20. Nevertheless, the full-scale Unitank system with relative high C/N and C/P ratios only yields a BPR efficiency ranging from 67.0% to 83.0%, which was probably attributed to the low proportion of carbon source utilized for anaerobic P release. Comparison of effluent pollutants among Unitank groups Average COD, AN, TP and SS in the effluent of four Unitank groups during three stages were illustrated in Figure 3.
As shown in Figure 3(a), the average COD concentrations in effluents of the four groups during three stages all fluctuated between 41.1 and 45.1 mg L−1 , whereas the average SS concentrations ranged from 15.0 to 17.3 mg L−1 . There was no obvious difference in COD and SS in the effluent among four groups during three stages, indicating that the enhancement of anaerobic time and the increase of wastage sludge discharge had insignificant influence on COD degradation and solid–liquid removal efficiency. Compared with stage 1, the average AN concentrations in effluents of the four groups all increased in stage 2 (Figure 3(c)). Since the temperature in stage 2 was almost the same as that in stage 1, the slight increase of AN in the effluent is probably a result of reducing SRT. Average
Downloaded by [UPM] at 01:43 24 December 2014
Z. Zhou et al.
Influence of anaerobic time on BPR of the Unitank process The TP and NOx -N variations in side tank of Unitank group 3# during anaerobic phase (0–60 min) in stage 3 are shown in Figure 4. By increasing the anaerobic time from 40 to 60 min in half a cycle, NOx -N in the filtrate decreased from 3.97 to 3.25 mg L−1 owing to the increase of carbon source supply for denitrification. During the 60 min anaerobic stage, TP in the filtrate obviously increased from 0.27 to 1.95 mg L−1 . Nevertheless, owing to continuous feeding of wastewater in the Unitank process, the contribution of influent TP should be taken into consideration to calculate the net released P under anaerobic environment. The average influent TP of the Unitank system is 3.60 mg L−1 , and θ of the side tank is 5.3 h. The anaerobic stage is started at 0 min with initial TP of 0.27 mg L−1 . Substituting the above data to Equation (3), the theoretical TP variation during anaerobic stage is shown in the following equation: CTP = 3.87 − 3.60e−t/318 (5) The measured TP concentrations in Figure 4 subtracting calculated values of Equation (5) can obtain the net
2.0
5.5 TP Net released TP
1.6
5.0
NOx-N 1.2
4.5
0.8
4.0
0.4
3.5
0.0
NOx -N (mg·L-1)
AN concentrations in effluents of groups 1#, 2#, 3# and 4# in stage 3 were 1.45, 0.87, 0.42 and 0.50 mg L−1 higher than those in stage 1, and increments of average AN concentrations in groups 3# and 4# were lower than those in groups 1# and 2#, respectively. In fact, AN concentrations in mixed liquor of the side tank was below 1.0 mg L−1 after 75 min aeration. These results suggest that the enhancement of anaerobic time does not have obvious negative impact on nitrification efficiency, although this optimization method leads to the loss of 40 min aeration time. Variations of TP in effluents of groups 1# and 2# can be used to analyse the influence of SRT on BPR removal, and data in stages 1 and stage 3 were employed due to their similar C/P ratios. The average BPR efficiency of groups 1# and 2# increased from 63.8% for a SRT of 21.3 to 83.2% for a SRT of 13.1 d. The finding was in agreement with the study of Lee et al. [17] who reported that the BPR efficiency of an AAO SBR decreased from 47.1% (SRT of 5.9 d) to 31.0% for a SRT of 16.2 d. Average TP concentrations of effluents in groups 1# and 2# in stage 3 were 0.55 and 0.67 mg L−1 lower than that in stage 1, while the drop of average TP concentrations in groups 3# and 4# were 0.75 and 0.47 mg L−1 , respectively. As shown in Figure 3(d), the average TP concentrations in effluents of groups 1#, 2# and 3# were close (1.34, 1.36 and 1.40 mg L−1 , respectively), and higher than that of group 4# (1.15 mg L−1 ). Therefore, the drop in group 4# was lower than the other three groups. With similar effluent TP in stage 1, the drop in group 3# was higher than those in group 1# and group 2#, indicating that the enhancement of anaerobic time was able to improve the BPR of the Unitank process.
TP (mg·L -1)
770
3.0 0
10
20 30 40 Anaerobic time (min)
50
60
Figure 4. Variations of TP and NOx -N in side tank of the full-scale Unitank system.
anaerobic P release amount, as shown in Figure 4. After the anaerobic time increased from 80 to 120 min, the net anaerobic P release amount increased from 0.25 to 1.06 mg L−1 . In half of an operation cycle, the average specific phosphorus release rate (SPRR) increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP·(gVSS·h)−1 in 40–60 min. The anaerobic P release rate was enhanced obviously by prolonging the anaerobic time. The SPRR was significantly lower than the initial SPRR reported by Panswad et al. [19,20] which was because the reported SPRR was measured in the environment rich in carbon source. Although high C/P ratio was provided in the WWTP, the presence of NOx -N (Figure 4) and low distribution of COD in the anaerobic stage resulted in a phosphorus release process that probably limited by carbon source supply. Activated sludge samples were collected for sludge P content measurement from side tank of each Unitank group at the end of aerobic stage (155 or 335 min) in stage 3. The measured sludge P contents were illustrated in Figure 5. The average sludge P content of groups 3# and 4# with anaerobic time of 120 min was 15.0 mgP·(gSS)−1 , which was higher than that of groups 1# and 2# of 13.8 mgP·(gSS)−1 . These results suggest that the enhancement of anaerobic time was conducive to the accumulation of phosphorus in activated sludge. In two full-scale AAO WWTPs in Shanghai with C/P ratios above 60, measured sludge P contents were 27.40 ± 0.53 and 27.84 ± 0.56 mgP·(gSS)−1 (n = 3). In the AAO process, wastewater are first fed into the anaerobic stage and utilized as carbon source for phosphorus release; therefore, its BPR environment is easier to form than the Unitank process. Lee et al. [21] reported a lab-scale anaerobicintermittent aeration process with sludge P content of 44.2–59.5 mgP·(gSS)−1 at similar SRT (15 d) but lower C/P ratio (30). In the lab-scale AO process studied by Panswad et al. [20] the sludge P content was 53 mgP·(gVSS)−1 at C/P ratio of 50. Although enhanced to 15.0 mgP·(gSS)−1 by the improvement of anaerobic P release environment,
Environmental Technology
Sludge P content [mgP·(gSS)-1 ]
18
increasing sludge P content (CP,EPC ) can be calculated as
15.4 14.6
17
14.1 15
13.6
14 13 12 11 1#
Downloaded by [UPM] at 01:43 24 December 2014
MW (XP2 − XP1 ) . (9) Q0 According to the influent and effluent data in Table 2, the TP removed in stage 1 and stage 3 was 2.50 and 3.65 mg L−1 , and thus the increment of phosphorus removed by the two optimization methods was 1.15 mg L−1 . If the effect of SRT fluctuation on sludge P content is ignored, the average sludge P content of groups 1# and 2# can be regarded as background data to evaluate contributions of two optimization measures. Based on the background sludge P content of 13.8 mgP·(gSS)−1 and wastage sludge discharge in stage 1 and stage 3, the calculated CP,SRT was 0.99 mg·L−1 . When sludge P content rose from 13.8 mgP·(gSS)−1 to 15.0 mgP·(gSS)−1 by the enhancement of anaerobic time, the increment of P removed (CP,EPC ) was 0.18 mg L−1 . The sum of CP,SRT and CP,EPC was 1.17 mg L−1 , which was very close to the measured increment of P removed (1.15 mg L−1 ). In the total increment of phosphorus removal, phosphorus removed by reducing SRT contributes 84.6%, while that removed by the enhancement of anaerobic time accounts for 15.4%. These results suggest that the improvement of BPR efficiency was mainly the result of increasing wastage sludge discharge. CP,EPC =
16
Figure 5. (n = 6).
771
2# 3# Unitank group No.
4#
Sludge P contents in four Unitank groups in stage 3
the sludge P content was still significantly lower than the reported data. These results indicate that the typical BPR environment is difficult to form owing to the configuration and operation of the Unitank process.[10] Model-based evaluation of BPR optimization methods In BPR systems, phosphorus in wastewater was transferred to activated sludge by excess phosphate uptake of PAOs and anabolism of other biomass, and then removed by wastage sludge discharge. Therefore, the mass balance equation can be formulated as Q0 CP0 = Qe CPe + Qw Xw XP
(6)
Conclusions The BPR of a full-scale Unitank WWTP was optimized by increasing anaerobic time from 80 to 120 min (for group 3# and 4#) and reducing SRT from 21.3 to 13.1 d, and the main conclusions of the study can be summarized as follows.
(8)
(1) In the Unitank process, relatively low fraction of anaerobic time (feeding and agitation) in the total time of an operation cycle usually limits anaerobic P release, and then leads to relatively low BPR efficiency. (2) The increase of wastage sludge discharge can improve BPR effect. The BPR efficiency of the Unitank system increased from 63.8% (SRT of 21.3 d) to 83.2% for a SRT of 13.1 d. (3) In the Unitank process, the increase of anaerobic time can enhance anaerobic phosphorus release amount, and thus increase sludge P content. The average SPRR increased from 0.097 mgP·(gVSS·h)−1 in 0–40 min to 0.825 mgP· (gVSS·h)−1 in 40–60 min in half a cycle. (4) Reducing SRT and increasing anaerobic time account for 84.6% and 15.4% in the total increment of phosphorus removal of 1.15 mg L−1 .
If the sludge P content rises from XP1 to XP2 by optimization measures (such as increasing anaerobic time in the Unitank process), the increment of phosphorus removed by
Funding This work was supported by Chinese National 863 Program [2012AA063403]; Shanghai Chenguang Program
where CP0 and CPe are, respectively, TP concentrations in the influent and effluent (mg L−1 ) andXP is P content of wastage activated sludge [mgP·(gSS)−1 ]. The wastage sludge discharge flow rate is significantly lower than influent flow rate; therefore, the effluent flow rate could be assumed as influent flow rate. Equation (6) can be deformed as Qw Xw XP Mw X P CP0 − CPe = = , (7) Q0 Q0 where Mw is the wastage sludge discharge (g·d−1 ). The phosphorus removed by different optimization measures can be calculated by Equation (7). When the wastage sludge discharge increases from Mw1 to Mw2 , the increment of phosphorus removed by reducing SRT (CP,SRT ) can be calculated as CP,SRT =
(MW2 − MW1 )XP . Q0
772
Z. Zhou et al.
[2011CG60]; Innovation Program of Shanghai Municipal Education Commission [12YZ137]; 085 Project Program ‘Energy Storage Technologies of Intelligence Network’.
Downloaded by [UPM] at 01:43 24 December 2014
References [1] Ma Y, Peng Y, Wang X, Wang S. Nutrient removal performance of an anaerobic–anoxic–aerobic process as a function of influent C/P ratio. J Chem Technol Biotechnol. 2005;80:1118–1124. [2] Wu C, Peng Y, Wan X, Wang S. Performance and microbial population variation in a plug-flow A2 O process treating domestic wastewater with low C/N ratio. J Chem Technol Biotechnol. 2011;86:461–467. [3] Makinia J, Swinarski M, Dobiegala E. Experiences with computer simulation at two large wastewater treatment plants in northern Poland. Water Sci Technol. 2002;45: 209–218. [4] Guerrero J, Tayà C, Guisasola A, Baeza JA. Understanding the detrimental effect of nitrate presence on EBPR systems: effect of the plant configuration. J Chem Technol Biotechnol. 2012;87:1508–1511. [5] Zhou Z, Wu Z, Wang Z, Tang S, Gu G, Wang L, Wang Y, Xin Z. Simulation and performance evaluation of the anoxic/anaerobic/aerobic process for biological nutrient removal. Korea J Chem Eng. 2011;28:1233–1240. [6] Zhu R, Wu M, Yang J. Effect of sludge retention time and phosphorus to carbon ratio on biological phosphorus removal in HS-SBR process. Environ Technol. 2013;34: 429–435. [7] Zhao C, Peng Y, Wang S, Takigawa A. Effects of influent C/N ratio, C/P ratio and volumetric exchange ratio on biological phosphorus removal in UniFed SBR process. J Chem Technol Biotechnol. 2008;83:1587–1595. [8] Lim SH, Ko KB, Rho KE. Biological nutrient removal using an IACOD process: determining the combined effects of low temperature and long solids retention time. Environ Technol. 2006;27:467–475. [9] Zhang FG, Liu JX, Sui J. Sludge concentration dynamic distribution and its impact on the performance of UNITANK. J Environ Sci. (China) 2007;19:141–147.
[10] Li L, Lei M, Tao T, Chen Y. Study on anaerobic environment and optimization of UNITANK process. China Water Wastewater. 2006;22:86–89 (in Chinese). [11] Wang L, Miao H, Zhou Y. Commissioning and operation of UNITANK process in Dali wastewater treatment plant. China Water Wastewater. 2008;24:92–94 (in Chinese). [12] Dong S, Wang X, Wang W, Xu Z, Yao C. Non-steady stage of ortho-phosphate concentration in a Unitank process. Environ Chem. 2007;26:1–4 (in Chinese). [13] Zhang H, Huang W, Cheng L, Qiang X, Xu H, Jin J. Application of modified UNITANK process in extension project of Zhuji WWTP. China Water Wastewater. 2008;24:67–71 (in Chinese). [14] Qiu Y, Li Y, Zhou S, Wu S, Du H, Li Y. Analysis of operation effect of UNITANK process. Environ Sci Technol. 2010;33:124–127 (in Chinese). [15] Lei M, Li L, Su X, Li Y, Zhang Y. In-depth cognition on conventional Unitank process. Environ Eng. 2006;24:35–36 (in Chinese). [16] Chinese NEPA. Water and wastewater monitoring methods. Beijing, China: Chinese Environmental Science Publishing House; 2012. [17] Lee JK, Choi CK, Lee KH, Yim SB. Mass balance of nitrogen, and estimates of COD, nitrogen and phosphorus used in microbial synthesis as a function of sludge retention time in a sequencing batch reactor system. Bioresourc Technol. 2008;99:7788–7796. [18] Yagci N, Artan N, Cokgor EU, Randall CW, Orhon D. Metabolic model for acetate uptake by a mixed culture of phosphate and glycogen-accumulating organisms under anaerobic conditions. Biotechnol Bioengin. 2003;84:359– 373. [19] Panswad T, Doungchai A, Anotai J. Temperature effect on microbial community of enhanced biological phosphorus removal system. Water Res. 2003;37:409–415. [20] Panswad T, Tongkhammak N, Anotai J. Estimation of intracellular phosphorus content of phosphorus-accumulating organisms at different P:COD feeding ratios. J Environ Manag. 2007;84:141–145. [21] Lee D, Kim M, Chung J. Relationship between solid retention time and phosphorus removal in anaerobic-intermittent aeration process. J Biosci Bioengin. 2007;103:338–344.
