By James C. Young and Madan Arora
Tulare’s Industrial Wastewater Treatment Plant (IWWTP) in 2006 received 26,000 m3/d (6.9 MGD) of low pH wastewater, primarily from industries that processed milk for various purposes. The raw industrial wastewater had the following anticipated characteristics: COD = 2,500 to 4,000 mg/L, NH3-N = 50 mg/L, TSS = 435 mg/L, VSS = 390 mg/L, FOG = 200 mg/L, EC = 694 µS/cm, pH 3-5. Further review of chemical use records for each contributing industry revealed releases of chemicals, specifically quaternary ammonium compounds and peroxyacetic acid, that can be toxic to methanogens in anaerobic pretreatment processes and nitrification in aerobic processes. The industries were encouraged to eliminate the use of these chemicals and change to hypochlorite or chlorine that could be neutralized with reducing agents such as sodium sulfite.
The system existing at the time of the design review consisted in part of a low-rate anaerobic Bulk Volume Fermenter (BVF, EVOQUA/ADI Systems Inc, Fredericton, NB Canada) that was used to pretreat 15,000 m3/d (4 MGD) of wastewater (Figure 1). The effluent from this anaerobic reactor was combined with the remainder of the raw industrial wastewater, and the mixture was treated further by using a four-stage aerobic/facultative lagoon system that was aerated using mechanical electric-driven surface aerators. Effluent from the aerobic lagoons was blended with 22,300 m3/d (5.9 MGD) of secondary effluent from the co-located Domestic Wastewater Treatment Plant (DWWTP) before it was discharged to ponds that were used for percolation and as a source of irrigation water for local farmers.
Discharge limits for the upgraded municipal/industrial effluent were: BOD <40 mg/L; Electrical Conductivity (EC) <710 µS/cm and Total Inorganic Nitrogen (TIN) < 10 mg/L. These limits were dictated in large part by use of the blended effluent for irrigation in water-short Tulare and in part because the percolated flow enters local aquifers. Lack of compliance with these standards was threatened because high organic loads stressed the existing treatment system and the facilities that existed in 2006 were not designed for nitrogen removal. Additionally, magnesium and sodium hydroxide chemicals added for pH control in the anaerobic pretreatment reactor contributed significantly to the EC so that the final effluent was approaching compliance limits at the time this study was initiated.
New Plant Design
After consideration of potential growth and economic factors, the City decided to upgrade and expand the plant capacity to 45,400 m3/day (12 MGD) and 180,000 kg COD/day (400,000 lb/day). After intensive review of the existing facilities and expansion objectives, the decision was made to upgrade the existing system using a Sequencing Batch Reactor (SBR) type activated sludge process as shown schematically in Figure 2.
The decision to use SBRs was based in large part on the limited site area available for expansion and the fact that SBRs do not need clarifiers or sludge recycle pumps. The existing low-rate anaerobic BVF reactor would continue to pre-treat about 15,000 m3/d (4 MGD) of industrial wastewater, thereby removing approximately 45,000 kg COD/day (100,000 lb/d) and producing 14,000 m3/day (500,000 ft3/d) of biogas that would be used for power production. A Dissolved Air Flotation (DAF) unit was included to remove Fats, Oil, and Grease (FOG) along with some of the suspended solids and COD in the part of the industrial wastewater that is not pretreated anaerobically. The DAF unit was expected to give around 75% removal of FOG and 40% reduction in COD. The first-stage cells of the existing lagoon system, with existing aerators, were used for aerobic pretreatment of the blended raw IWW/anaerobic effluent. Provisions were included to use only the number of lagoon cells needed to optimize overall treatment performance.
The pre-treatment lagoons were expected to remove 50% of the COD load. Aerobic SBR reactors would treat the effluent from the aerobic cells to remove additional COD and to achieve nitrification and partial denitrification. Denitrification filters were included to reduce the residual nitrate nitrogen to below 5 mg/L.
Electrical Conductivity (EC) was an important factor in selecting the above wastewater treatment train. In California’s Central Valley, the State had set an electrical conductivity (EC) limit for wastewater discharges of 500 µS/cm (µmhos/cm) above a local groundwater background value, which in 2007 averaged around 210 µS/cm. Because EC is the combined effect of all dissolved ionized chemicals in the wastewater, any addition of chemicals by industrial contributors to the treatment plant would increase the EC. Chemicals used for pH control in the anaerobic pretreatment reactor also would contribute significantly to the EC of the final effluent.
The existing anaerobic pretreatment process used either magnesium hydroxide [Mg(OH)2] or sodium hydroxide [NaOH] for pH control, but these chemicals significantly increased the EC. The only option for EC control when using these chemicals was Reverse Osmosis (RO), but the cost for RO treatment was found to be prohibitive primarily because of the cost of transporting the reject brine to a disposal location. Ammonium hydroxide [NH4OH] can be used for pH control and offered the potential for achieving EC reduction if the downstream aerobic process provided complete nitrification and denitrification. In this case, the ammonium hydroxide converts to ammonium bicarbonate in the anaerobic reactor. The ammonium bicarbonate leaving the anaerobic system will nitrify to form nitric acid in the downstream aerobic SBR process. This reaction proceeds, in simplified form, as follows:
The nitric acid (HNO3) and carbon dioxide produced during nitrification will consume alkalinity and lead to pH reduction. One of the requirements for the planned mode of operation was that no chemicals be added to control pH in the aerobic SBR process. Denitrification of the nitrate produced in the aerobic process removes the nitric acid as follows, again in simplified form:
where H2 represents electron donors such as organic compounds. This reaction produces an increase in alkalinity and pH.
The above analysis shows that little change in the EC of the treated wastewater would occur with nitrification alone. But complete nitrification and denitrification would remove cations (NH4+) and anions (NO3-) from solution and thereby reduce EC in the treated effluent.
Because of the complexity of the proposed treatment system, a laboratory-scale study was conducted to verify that the anticipated reduction in EC would occur through nitrification and denitrification. This system was designed to simulate operation of the new full-scale plant when operating at 15,000 m3/d (8 MGD). This is a conservative condition; EC would decrease as raw wastewater flow increases due to a smaller fraction of the industrial wastewater being pre-treated in the BVF reactor. A twelve-day Solids Retention Time (SRT) was selected by the design team as the lowest practical SRT for achieving nitrification with the industrial wastewater at Tulare. The primary objective of the authors’ study was to verify that the anaerobic/ aerobic/ anoxic biological system proposed for treating the industrial wastewater at Tulare would provide an effluent meeting the EC discharge limit in addition to limits for conventional parameters such as BOD5, TSS and nitrogen. The laboratory-scale system consisted of a 1-L aerobic pretreatment reactor followed by a 1-L aerobic SBR reactor. The pretreatment reactor received a blend of raw industrial wastewater and BVF reactor effluent and was operated at a 3-d Hydraulic Retention Time (HRT). Table 1 lists the characteristics of these two samples. The aerobic Sequencing Batch Reactor (SBR) was operated at a Solids Retention Times (SRT) of 12 days and was fed pretreated industrial wastewater. Trace minerals were added to ensure that adequate amounts were present to support biological growth.
Bench-Scale Pre-treatment Reactor
The batch test reactors were operated for 77 days with measurement of sCOD, NH3-N, NO3-N, NO2-N, MLSS, and MLVSS at approximately 7-day intervals. The feedstock COD to the pre-treatment reactor was 1,317 mg/L. Total COD concentrations in the effluent from the pretreatment reactor averaged 646 mg/L (data not shown). The pre-treatment reactor also showed almost complete nitrification after 13 days of operation with complete conversion to nitrate after 25 days of operation.
Bench-Scale SBR Reactor
The bench-scale SBR reactor received a mixture of 50% pre-treated effluent plus 50% BVF/IWW mixture using the samples described in Table 1. Because denitrification was incomplete when operating with an anoxic time of 4 hours, the anoxic time was increased to 6 hours on day 30 and eventually to 12 hours on day 40. Thereafter, the sCOD in the SBR reactor remained at concentrations around 20 mg/L with nitrate concentrations stabilizing at 21 mg/L through the 77th day of operation (Table 2).
Nitrification of the ammonia-N added with the blended feed wastewater was completed easily in the SBR reactor in less than six hours of aeration so that TKN and ammonia nitrogen conversion exceeded 99% for the total system (Table 2). One characteristic of SBR systems is that some nitrate is discharged with the effluent during decanting at the end of the aeration phase of the operating cycle. The SBR mode of operation resulted in an effluent nitrate-N concentration of 21 mg/L. Additional removal was achieved in the full-scale treatment system by using denitrification filters with methanol serving as a carbon source to give an effluent nitrate concentration averaging 5 mg/L.
Respirometers were used to measure denitrification rates. In this case, the wastewater fed to the respirometer vessels was varied to evaluate the impact of bypassing some of the raw IWW mixtures around the first-stage pre-treatment cells. Test results showed that the denitrification rate was lowest when using 100% first-stage effluent and improved as the amount of raw IWW was increased (Table 3). These rates were within the range normally seen in full-scale denitrification systems, and clearly showed the benefit of bypassing some of the raw IWW around the first-stage aerobic pre-treatment reactor. The optimum amount of bypass needed for the full-scale system will depend on several operating factors, including soluble COD, amount of nitrate to be denitrified, temperature, and SBR cycle time. Therefore, the capability to bypass raw IWW to the SBR process was incorporated into the design of the full-scale treatment system.
Overall Cycle Time
The best cycle time in the laboratory tests, when using a 50% BVF/IWW bypass, included an anoxic denitrification time of twelve hours and a nitrification time of six hours. Under these conditions and at steady-state operation, the MLVSS concentration in the lab-scale reactors averaged 973 mg/L (Table 2). Scaling the cycle time to full-scale conditions of 3,000 mg VSS/L by using a constant F/M ratio gave a denitrification time of 3.9 hr and a nitrification time of 1.9 hr. These reaction times, especially when considering that denitrification can occur during the filling operation, supported the use of an eight-hour total cycle time for sizing the full-scale SBR system.
As stated above, one objective of the test program was to verify the anticipated change in EC as the ammonia nitrogen was degraded through nitrification and denitrification. Measurements of Electrical Conductivity (EC) were made at various points throughout the bench-scale treatment system with the results listed in Table 4. The 852 μS/cm EC associated with this residual nitrate will be reduced by an additional 24 μS/cm by the downstream denitrification filters. The low pH in the effluent from the pre-treatment reactor, as shown in Table 4, reflects the impact of almost complete nitrification in that reactor. The pH increased through the SBR process because of alkalinity release during denitrification and stripping of carbon dioxide during aeration.
Unique Design Features of New Treatment System
The upgraded full-scale system described in this article has been in operation since November 2009. Unique design features for this project include:
- Continued use of the original first-stage lagoons for additional pretreatment with options to using the number of cells needed for the optimization of performance.
- Continued use of the existing anaerobic BVF reactor for partial removal of 25% to 30% of the total COD load with the production of biogas for co-generation with fuel cells.
- Provision for bypassing a portion of the raw industrial wastewater flow around the pretreatment lagoons to provide carbon needed to support denitrification, and thereby reduce the amount of methanol consumption.
- Use of ammonium hydroxide for pH control in the anaerobic BVF reactor with subsequent nitrification and denitrification in the SBR system to eliminate the associated EC.
- Return of filtrate from the sludge drying process to the BVF reactor influent to reduce the amount of ammonium hydroxide needed for pH control (not discussed in this article).
- A laboratory-scale model of the two-stage pre-treatment/SBR system proposed for Tulare’s IWWTP provided 98.5% COD removal and 99.5% nitrification when operating at a solids retention time of 12 days in the SBR unit. The resulting average effluent COD concentration was 20 mg/L.
- Adjusting laboratory-scale nitrification and denitrification times to full-scale conditions indicated that the total SBR cycle should be around eight hours when operating with an MLVSS concentration of 3,000 mg/L and using the fill time for denitrification.
- The nitrate-N concentration in the decanted SBR effluent averaged 21 mg/L. This nitrate-N concentration was reduced to less than 5 mg/L after further treatment in denitrification filters.
- Nitrification and denitrification in the SBR system reduced the Electrical Conductivity (EC) from 1,210 to 852 μS/cm. Completion of denitrification in the downstream denitrification filters was expected to reduce the effluent EC to around 828 μS/cm.
- Blending 8 MGD of industrial effluent at an EC of 828 μS/cm with 6 MGD of domestic effluent at an EC of 530 μS/cm would give an anticipated blended effluent EC of 700 μS/cm. Operating at 12 MGD would give lower ECs because the fraction of raw wastewater pretreated by the BVF reactor would be lower. Therefore, the use of ammonium hydroxide for pH control in the BVF reactor followed by nitrification and denitrification of the mixed BVF effluent and residual raw industrial wastewater in the SBR system would allow Tulare to meet an EC discharge standard of 710 μS/cm.
- Construction of a full-scale industrial treatment plant capable of treating 45,400 m3/d (12 MGD) began in 2007 with completion in November of 2009. Startup challenges have been documented in other publications.
About the Authors
James Young Ph.D., P.E., BCEE received his BS in civil engineering and his MS in environmental engineering from New Mexico State University; and his Ph.D. in environmental engineering from Stanford University. He was a professor and researcher for 41 years and recently retired from the University of Arkansas, where he taught courses in industrial wastewater treatment and conducted research on biological treatment processes.
He has served as a consultant on projects involving industrial wastewater treatment and has published numerous papers on this topic in journals and conference proceedings. Dr. Young is the 2015 recipient of the W.W. Eckenfelder Industrial Water Quality Lifetime Achievement Award from the Water Environment Federation.
Madan Arora, Ph.D., P.E., BCEE, a Technical Director at Parsons (a global engineering and construction company with its corporate offices in Centreville, Virginia, USA) has over 50 years of experience in wastewater treatment and water reuse. Dr. Arora has M.S. and Ph.D. degrees in Environmental Engineering from Iowa State University in Ames, Iowa and is a registered engineer in the State of California and a life member of the Academy of Environmental Engineers and Scientists.
He is also a life member of the Water Environment Federation (WE&F). Dr. Arora has published extensively in national and international journals and spoken at conferences on water reclamation and reuse.
© Smart Water & Waste World. Send us your editorial contributions at email@example.com