Assess the denitrification potential of fermented biosolids based on their specific denitrification rate (sdnr)

Journal of Science and Technology 54 (2A) (2016) 112-119 112 ASSESS THE DENITRIFICATION POTENTIAL OF FERMENTED BIOSOLIDS BASED ON THEIR SPECIFIC DENITRIFICATION RATE (SDNR) Phung Anh Duc 1, * , Phung Chi Sy 2 1 Royal Melbourne Institute of Technology, 124 La Trobe Street, Melbourne, Victoria, Australia 2 Environmental Technology Center (ENTEC), 439A9 Phan Van Tri Street, Ward 5, Go Vap District, Ho Chi Minh City, Vietnam * Email: anhducphung1988@gmail.com Received: 1 April 2016; Accepted for publication: 15 June 2016 ABSTRACT The study examined the potential of fermented and dark-fermented biosolids as external carbon sources for denitrification improvement. This was done by up-scaling the selected two (out of seven) sludge fermentation conditions from past studies, carrying out ammonia stripping pre-treatment to fix the C/N ratio, before finding their specific denitrification rate (SDNR) using SDNR experiment set-up. The gotten SDNR were then compared to the SDNR of other substances gotten from both previous studies and literature, to weight the denitrification potential of fermented biosolids as a substance. The results found that with an initial COD of 607-704 mgCOD/L, the SDNR of the two fermented biosolids and dark fermented biosolids were found to be 8.35 ± 0.41 and 8.56 ± 0.71 respectively. This was much higher than the 1.53 - 2.57 for sucrose and 1.29 ± 0.21 for wastewater found in earlier study using the same methodology; and comparable to the denitrification potential value for the well-studied methanol Keywords: fermented sludge, dark fermentation, denitrification potential, SDNR. 1. INTRODUCTION The lack of organic carbon available for denitrification in anoxic zones of municipal wastewater treatment plants (WWTP) have always been one of the biggest issue within the industry. As organic carbon often was the limiting substrates that prevent complete denitrification to be achieved, subsequently resulted in high concentration of nitrates in post- treatment effluent. Denitrification-specialized process like Modified Ludzack-Ettinger (MLE), or Bardenpho were specifically designed to address this issue, however they were unable to completely eliminate the problem around the globes [6]. Because firstly reconstructing and modifying an existing wastewater treatment plant to improve the system efficiency was not always ideal or even possible, due to the need for high capital investment, high operating cost and sometime simply due to the lack of space/land available to implement extra treatment zones. And secondly even for a well-designed system, if the readily biodegradable COD (rbCOD) Assess the denitrification potential of fermented biosolids 113 inside municipal wastewater influent was not inadequate, low specific denitrification rate (SDNR) and overall ineffective denitrification could still happen. In contrast, adding an external carbon sources into post- or pre-anoxic zone to improve the denitrification would be much assured and easier to implement (due to low required modifications to an existing WWTPs). This was especially ideal for short-term solution, but also viable for long term if a cheap and abundant external carbon sources could be secured. Hence, finding an alternative cost-effective external carbon substance was listed amongst the priorities of the wastewater treatment industry during WEF 2005 [8] and have been going on for the past two decades [12]. The focus was mostly on local industrial waste products/ by-products that is rich in carbon, where they were addressed to be ‘highly recommended especially if one WWTP can has the access’[12]. Many materials rich in carbon that has been studied in the past include industrial wastewater, corn starch, reject water [1], syrup from distillery waste product food industries [12]. Previous to this study, batch tests on various conditions of fermented and dark fermented Wasted Activated Sludge (WAS) and anaerobic digested biosolids (biosolids) were carried out to find the optimum conditions and to assess their potential as external carbon sources. While none met the minimum required standard at first due to the high ammonia concentration. After being treated by ammonia-stripping, their initial characteristic (mostly C/N ratio) all showed to be viable for further testing. Two of the fermented sludges (amongst seven) were picked out for further study here in this paper. Because the resulted characteristics of all fermented sludges were not decisively different; the selection process was based on how complicate/resources-consuming to prepare for each batch tests. Hence the most basic fermented and dark-fermented biosolids were picked (in opposing to the others that required mixed stream and/or addition of cellulose carbons). 1. MATERIALS AND METHODS 2.1. Chemicals and Inoculums Similar to earlier published study on sludge fermentation [9], the inoculum for this fermentation experiment was also taken from local anaerobic digestion plant rather than from standard WWTP; as this would be best to simulate the condition of a continuous running fermenter condition. The biosolids were taken from the influent and effluent drains from the same anaerobic digester. Both of the feeds and inoculum were sieved to remove larger particles. The used biosolids samples had similar characteristic to the one in earlier published papers with Total Solid (TS) of 28,342 ± 1319 mg/L and Volatile Solid (VS) of 20,755 ± 1152mg/L. The biggest difference was the high undiluted soluble COD in the inoculum using in this paper was 4919 ± 392 mg/L. Despite saying that, this was generally still low within the expected range due the measured VS of over 20,000 mg/L. However, considering it’s higher than the one found in earlier published paper (of 1247 ± 79 mg/L), hence this would be taken into account when assessing the final soluble COD. pH for fermentation and dark-fermentation set-up were controlled using diluted citric acid. Potassium nitrate, ammonia chloride, potassium hydrogen orthophosphorus for the SDNR batch tests were all bought from Science Supply Australia Company. Phung Anh Duc, Phung Chi Sy 114 2.2. Analytical Method Similar to earlier study and literature [9], COD, TN, Ammonia, Nitrate, Volatile Fatty Acid were analysed using HACH standard methods for the DR 5000 (Methods 8000, 10072, 10031, 10020, and 8196 respectively). The rbCOD in this case was measured by the filtration of soluble fermented biosolids through 0.45µm membrane filtration according to Melcer (5), before being tested with COD reagent kit. This is because most of the particulate soluble COD would have been absorbed into the sludge, and the measured COD would be very close to actual rbCOD (difference of <5 mg/L). pH and DO were tested by using the Mettler Toledo S20 Seveneasy pH meter and YSI 5100 dissolved oxygen meter. 2.3. Experimental set-up a) Biosolids fermentation and ammonia stripping The ratio of fermented and dark fermented biosolids to inoculum were similar to earlier published study (inoculum: feed of 80:20) before each reactor pH was then adjusted using diluted citric acid, or/and milli-Q water. Each fermentation reactor was purged with nitrogen gas before being placed into shakers for continuous mixing for 5 days. This 5 days number is the optimum period for fermentation based from both literature and past experimental results. Generated gas was released daily from the reactor, except for the first few days where it was released twice a day to prevent pressure building up inside the reactor The pH for fermentation reactors were set to be 6.6±0.1 according to results from earlier batch tests results and literature [4], while the pH chosen for dark-fermented was 5.6±0.1 [14]. Both set of reactors were operated under 55 °C, the reported ideal conditions for thermophilic organisms (thermophile). The two selected conditions were tested with earlier published experiments and found that a C/N ratio of more than 20:1 could be gotten for them after ammonia stripping pre-treatment. After fermentation process, for ammonia stripping, pH of the two fermented biosolids would be adjusted to pH 10 using calcium carbonate as this pH was reported to be ideal for ammonia stripping. The solid matter was then filtered out of the two fermented biosolids sample to eliminate any chance that ammonia reduction would be caused by nitrification. The ammonia stripping set-up was simple: air was set to bubble through the 2L beakers of fermented biosolids using an air stone. Ammonia was measured every 2 hours until the target >65% ammonia removal was reached. b) SDNR experiments The SDNR experiments methodology was the same method used in earlier published paper [11], The sludge used were from the laboratory SBRs (Sequencing Batch reactors) after it was acclimated with the two fermented sludge samples for a month. The average initial MLVSS of all reactors (include the duplicates) is 908 ± 29 mg/L. The initial COD is 704±2.2 mg/L for fermented biosolids reactors and 607±3.7 mg/L for dark fermented biosolids reactors. The range of 600-700 was based on the literature to ensure that excessive carbon sources are available during the whole experiment. COD, NH4-N, Volatile Acid (VA) concentration were measured at the beginning and at the end of the experiments. Nitrate concentration was monitored every 30 minutes for over 2.5 Assess the denitrification potential of fermented biosolids 115 hours. The SDNR (also the theoretical maximum denitrification rate of tested carbons) was calculated from the constant slope of removed NO3-N concentration. Potassium nitrate, ammonia chloride, potassium hydrogen orthophosphorus were added into the sludge slurry to maintain a concentration of ~30mgNO3-N/L, ~60mgNH4-N/L and ~9mgPO4-P/L, respectively in each reactor. This should provide excessive nutrient sources for denitrification process to reach the maximum rate [2]. 2. RESULTS AND DISCUSSION 3.1. Biosolids fermentation and ammonia stripping The full results of fermentation and dark fermentation batch tests are showed in Table 1 below, with the C/N ratio for the two sets of experiments to be 10.34 ± 0.17 and 10.14 ± 0.12 respectively. They are slightly higher than the 5.8–9.3 gotten from earlier study, most likely due to the initial higher soluble COD in used inoculum as noted in Section 2.1. However the difference is not as significant, because for both sets, the results were still much lower than the minimum required C/N ratio of 20:1 (or ideally of 30:1) to be considered as an effective external carbon sources. [3] Note that the C/N ratio or rbCOD/NH4 + is the most important independent variables in this experiment. Because the fermented biosolids would be added back to the anoxic zone as external carbon source, hence a lower C/N ratio would means: with the same amount of added carbon, more nitrogen would be introduced back into the system. That would undermine the nitrogen removal capacity of the whole system. Also, the N component in C/N theoretically supposed to be the Total Nitrogen instead of NH4 + . However soluble organic nitrogen and nitrates were tested for the filtered samples and the concentrations of these two were found to be insignificant. Especially when comparing to the high concentration of soluble NH4 + . The method used for NH4 + analysis also is much more reliable and produced much less error in comparing to TN measurement. Note that even in unexpected circumstance where higher organic nitrogen and nitrates were found in the fermented biosolids; the impact it have would still be minimal. Because once added into the anoxic zone as external carbon sources, the nitrate and/or organic nitrogen would be quickly denitrified and/or absorbed into the sludge mass respectively; making those two parameters even less relevant. Once the fermented and dark-fermented biosolids samples were generated, collected and filtered, ammonia stripping was then carried out as pre-treatment to fix the C/N ratio to the minimum required 20:1 ratio. The results are showed in Figure 1. As seen, it took 4 hours of constant aeration to reduce the NH4 + of fermentation and dark fermentation by an average of 65.5% and 67.3% respectively. This results in a C/N ratio of 30.0 ± 0.7 and 31.0 ± 0.6, much higher than the minimum required 20:1 and is around the ideal C/N ratio to be considered as an effective external carbon sources. Note that, this removal efficiency was from a simple lab-scaled aeration set up and can still be further optimized if needed. Commercial ammonia-stripping units were widely reported to reach a value as high as 99% for example Phung Anh Duc, Phung Chi Sy 116 Figure 1. The ammonia stripping performance as pre-treatment to fix the fermented biosolids C/N ratio 3.2. SDNR of the two fermented biosolids As the C/N ratio of these two fermented biosolids samples have showed it to be suitable as external carbon source. The next step is to calculate the specific denitrification rate (SDNR) of these two types of external carbon sources. And they were done by setting two sets of SDNR batch tests to theoretically calculate each carbon sources SDNR. The results for fermented and dark-fermented biosolids (including duplicate reactors) after adjusted for MLVSS (due to the MLVSS in each reactor maybe different) can be showed in Figure 2. Figure 2. Nitrate profile of the two set of SDNR batch test (included duplicates) The data for each reactor and its duplicate was then fed into R statistics to calculate the slopes (SDNR), the R^2, the adjusted R^2, the Significant Error (SE) and the 95% Confidence Interval (95% CI). The detail results are presented in Tables 1, 2 below. Assess the denitrification potential of fermented biosolids 117 Table 1. The full set of data for all fermentation experiments Day 0 results Day 5 results No Types Feed types Planned pH Ini pH °C Ini VS pH rbCOD NH4 C/N VFA VFA% 1 Fermentation Biosolid 6.5 6.63 ± 0.09 55 5 20,755 ± 1152 7.36 ± 0.01 11,730 ± 102 1,135 ± 25 10.34 ± 0.17 1,702 ± 68 14.5 ± 0.6 2 Dark Fermentation 5.5 5.59 ± 0.09 6.89 ± 0.04 11,257 ± 275 1,110 ± 11 10.14 ± 0.12 1,158 ± 5 10.3 ± 0.3 Table 2. The SDNR results of the two fermented and dark-fermented biosolids COD (mg/L) VA (mg/L) NH4 (mg/L) MLVSS SDNR Results mgCOD consumed per mgNO3 removed Initial Final Initial Final Initial Final SDNR R^2 R^2 (adjusted) SE SDNR (95%CI) Fermented Biosolids 704±2.2 442±1.0 102±0.3 2.0±1.0 68±0.1 64±0.8 930±20 8.35 0.9876 0.9851 0.41 7.50-9.19 7.57±0.13 Dark Fermented Biosolids 607±3.7 460±7.5 62±0.4 2.0±0.5 68±0.0 67±0.3 885±15 8.56 0.9733 0.9667 0.71 7.04 - 10.07 7.43±0.58 As seen on figure 2, after omitted the first 30 minutes due to the microorganisms need some time to acclimate to the reactors, both set of reactors showed quite consistent trends over most of the 2.5 hours of testing time. The reason why the first 30 minutes were omitted, was due to earlier studies found that even between repeated reactors of the same substance with same SDNR results, the amount of nitrate removal during the first half of hours could be varied greatly Phung Anh Duc, Phung Chi Sy 118 For dark-fermented biosolid, the SDNR rate showed slowing down and in fact almost flat out toward the end. This was quite unexpected as the final rbCOD concentration in dark- fermented biosolids of the SDNR batch tests was still quite high (460 ± 7.5 mg/L) indicating it was not because of the lack of rbCOD. Upon checking the Volatile Acid (VA) for both set of SDNR reactors, the VA was found to be almost depleted. This could be explained based on literature [3, 5] , where they stated that when fermented sludge is used as external carbon sources, VA rather than other non-VA rbCOD will be the preferred carbon source for denitrification [13], means depletion of VA may end up affect the denitrification rate of the system toward the end. However more studies will be needed to verify this, as it could simply be caused by errors. On the SDNR value of 8.35 ± 0.41 for fermented biosolids and 8.56 ± 0.71 for dark- fermented biosolids. These values were much higher than the 1.53-2.57 for sucrose and 1.29 ± 0.21 for wastewater that were gotten from earlier study using the same methodology [11]. The SDNR values for fermented biosolids were also comparable to the 6.1-10.0 SDNR value for methanol [2, 7, 10], the most studied and commonly used external carbon in US. This shows some genuine potential of fermented and dark-fermented biosolids as external carbon sources 3. CONCLUSION There are some conclusions that can be drawn from this experiment: 1. The larger scale of the same conditions of earlier fermentation and dark-fermentation batch tests, despite produced a higher C/N ratio of 10.34 ± 0.17 and 10.14 ± 0.12; however once again confirmed that even at optimum condition, the C/N ratio of fermented biosolids would still fall way short to the minimum C/N ratio of 20:1 (or ideally of 30:1). This was due to the high concentration of ammonia in the biosolids that simply could only increase during the fermentation process 2. An upscaled of earlier ammonia stripping tests, confirmed that even without optimisation, a 65% removal rate of ammonia stripping could be achieved within 4 hours of simply aerating through the 2L solutions. This pre-treatment helped to bring the C/N ratio of fermented biosolids to the ideal value of roughly 30:1 3. The denitrification potential of the two fermented biosolids was assessed using the SDNR experiments. The results showed some genuine potential for fermented sludge as external carbon sources due to their high SDNR value of 8.35 ± 0.41 and 8.56 ± 0.71 for fermented and dark-fermented biosolids (and with a fairly consistent 95% CI of 7.50-9.19 and 7.04 - 10.07 respectively) 4. Despite rbCOD was still abundant, the abrupt drop in denitrification rate toward the end of the SDNR batch tests for dark-fermented biosolids aligned with published literature; on how the Volatile Acid component in rbCOD is more crucial for denitrification than total rbCOD. However more study needed to be done to verify this. Acknowledgements. The research was possible thanked to the helps and support from the technical staffs and students from RMIT. Assess the denitrification potential of fermented biosolids 119 REFERENCES 1. Cherchi C, Onnis-Hayden A, El-Shawabkeh I, Gu AZ. Implication of using different carbon sources for denitrification in wastewater treatments. Water Environment Research. 2009;81(8):788-99. 2. Dold P, Takács I, Mokhayeri Y, Nichols A, Hinojosa J, Riffat R, et al. Denitrification with carbon addition - kinetic considerations. Water Environment Research. 2008;80(5):417-27. 3. Feng L, Chen Y, Zheng X. Enhancement of Waste Activated Sludge Protein Conversion and Volatile Fatty Acids Accumulation during Waste Activated Sludge Anaerobic Fermentation by Carbohydrate Substrate Addition: The Effect of pH. Environ Sci Technol. 2009;43(12):4373-80. 4. Lu S-g, Imai T, Ukita M, Sekine M. Start-up performances of dry anaerobic mesophilic and thermophilic digestions of organic solid wastes. Journal of Environmental Sciences. 2007;19(4):416-20. 5. Melcer H. Methods for Wastewater Characterization in Activated Sludge Modelling. London, UK: Water Environment research Foundation; 2003 01 Nov 2004. 6. Meyer RL, Zeng RJ, Giugliano V, Blackall LL. Challenges for simultaneous nitrification, denitrification, and phosphorus removal in microbial aggregates: Mass transfer limitation and nitrous oxide production. FEMS Microbiology Ecology. 2005;52(3):329-38. 7. Mokhayeri Y, Riffat R, Takacs I, Dold P, Bott C, Hinojosa J, et al. Characterizing denitrification kinetics at cold temperature using various carbon sources in lab-scale sequencing batch reactors. 2008. p. 233-8. 8. Oleszkiewicz JA, Kalinowska, E., Dold, P., Barnard, J. L., Bieniowski, M., Ferenc, Z., Jones, R., Rypina, A. & Sudol, J. Feasibility studies and pre-design simulation of Warsaw’new wastewater treatment plant. Environment Technololgy. 2004;26:1405-11. 9. Park J-H, Lee S-H, Yoon J-J, Kim S-H, Park H-D. Predominance of cluster I Clostridium in hydrogen fermentation of galactose seeded with various heat-treated anaerobic sludges. Bioresour Technol. 2014;157(0):98-106. 10. Phung DY, J. Effect of sucrose on denitrification through simulation, lab-scaled batch tests and pilot plants. Science & Technology Development Journal. 2014;17:22-32. 11. Phung DY, J. The Effect of sucrose on specific denitrification rates of sucrose. Science & Technology Development Journal. 2014;17:86-94. 12. Sage M, Daufin G, Gésan-Guiziou G. Denitrification potential and rates of complex carbon source from dairy effluents in activated sludge system. Water Res. 2006;40(14):2747-55. 13. Tong J, Chen Y. Recovery of nitrogen and phosphorus from alkaline fermentation liquid of waste activated sludge and application of the fermentation liquid to promote biological municipal wastewater treatment. Water Res. 2009;43(12):2969-76. 14. Yuan Q, Sparling R, Oleszkiewicz JA. Waste activated sludge fermentation: Effect of solids retention time and biomass concentration. Water Res. 2009;43(20):5180-6