Fermenting Organic Wastes to Produce Volatile Fatty Acids (VFAs) as a Carbon Sources or Alternate High Value Product

Municipal organic wastes, such as sewage sludge and food wastes, are typically stabilized in conventional anaerobic digesters producing biogas and biosolids. These two products are tried and true recoverable resources that have been captured and reused across the world. This study proposes adding volatile fatty acids, VFAs, to this list of recoverable resources via fermentation. This proposal consists of modifying existing digesters to two-phase digesters that includes a thermophilic acid digester followed by a gas mesophilic digester that would produce higher valued VFAs, biogas, and potentially Class A biosolids. Figure 1 depicts the conventional solids handling system in NYC WRRFs and the proposed two-phased digestion system. All experiments were conducted in batch reactors at the City College of New York (CCNY). Gravity thickened combined primary and secondary sludge were collected from the Wards Island WRRF, municipal food waste slurry from the Newtown Creek WRRF, and food waste from the University cafeteria. Each of the two feedstocks were fermented in separate acid reactors for a 2-day SRT. Despite the practice at Newtown Creek WRRF combining thickened sludge with food waste slurry prior to anaerobic digestion, the two feedstocks were fermented in separate acid reactors because food waste produces higher concentrations of VFAs with lower concentrations of ammonia. Thus, the two feedstocks were fermented separately with different VFA/ammonia ratios which has the potential to replace glycerol in a BNR facility, or to reroute a recovery process for alternate byproducts. This study was conducted in three phases: (1) characterization of fermentation products from the two feedstocks, sewage sludge and food waste, (2) comparison of the VFAs produced to glycerol as sources carbon to support denitrification, and (3) preliminary approach for VFA recovery. Phase 1: Characterization of Fermentation Byproducts from Sewage Sludge and Food Wastes Sewage sludge from Wards Island WRRF and food waste are fermented in lab scale batch reactors for 2 days contact time at 38°C at the following two conditions: (1) maintain a pH of 7 with reagent-grade sodium hydroxide (NaOH) addition, and (2) without pH control. Table 1 to Table 3 shows the characteristics of fermented combined sludge and fermented food waste when fermented with and without pH control. Results from experiments fermenting combined sludge and CCNY food waste show VFA production increased when pH was maintained at 7 for both organic wastes. In the case of food waste, VFA production per unit volatile solids fed (VFA/VS) increased eight-fold from 100 mg CODVFAs/g VS to 830 mg CODVFAs/g VS when maintaining the pH at 7 while with no pH control, the pH dropped from 5.0 to 3.6, causing an insignificant VFA production due to inhibition. Further indication of inhibition was the concentration of soluble COD (sCOD) remaining constant with the control. In contrast, sewage sludge fermentation retained the pH constant at 5.6 while the sCOD increased 8-fold to 8,000 mg/L because of active fermentation and significant release of ammonia from the combined sludge. The additional ammonia contributed to alkalinity, thereby increasing the buffering capacity and resisting acidification caused by VFA production and accumulation, as shown in Table 2. The quantity of ammonia released from fermenting sewage sludge compared to food waste was measured at 37-40 mg N/g VS compared with 12 mg N/g VS, respectively. Table 3 shows the Newtown Creek food slurry behaved more like the fermentation of Wards Island combined sludge, namely, that pH control to 7 had minimum impact on further production of sCOD. This may be a result of a more diverse food waste that was combined in the slurry and significant fermentation that had occurred prior to reaching Newtown Creek WRRF. Phase 2: Denitrification Studies Comparing VFAs from Fermented Sewage Sludge and Glycerol Batch experiments were performed to determine the efficacy of VFAs in the fermented organic waste filtrate satisfying carbon demand for denitrification. The efficacy of fermented sewage sludge and food waste were determined by comparing specific denitrification rates (SDNR) and CODconsumed/Nremoved ratio to glycerol specifically formulated for denitrification. The experiments were conducted at CCNY in 2-liter jars equipped with paddle mixers and polystyrene floats as a barrier between the sample and atmosphere. Mixed liquor from Wards Island WRRF BNR tanks were dosed with 20 mg-N/L sodium nitrate (NaNO3) and 200 mg/L COD from (1) fermented sewage sludge filtrate, (2) fermented food waste filtrate, or (3) glycerol. The carbon sources were prepared by centrifugation followed by filtering the supernatant through a 0.45 um pore size filter. A control of mixed liquor was included and dosed with 20 mg-N/L of NaNO3 without adding carbon source. Samples were taken at specific intervals to measure the reduction of nitrate (NO3-N) and nitrite (NO2-N), collectively as NOx-N, over a 2-hour period. The data from the duplicate batch experiments are plotted in Figure 2. Figure 2-A shows the profile of NOx-N throughout the duration of the experiment. A regression analysis was performed using the linear part of the profile, within the first 40 minutes. The slopes of the linear regressions developed in Figure 2-A were used to calculate SDNR values and were shown in Figure 2-B. The SDNR based on the fermented sewage sludge filtrate, fermented food waste filtrate, and glycerol carbon source were determined to be on average 5.45, 5.95, and 4.83 mg N/g VSS/hour, respectively, indicating that fermented sewage sludge and fermented food waste supports denitrification as effectively as the specifically formulated glycerol. The carbon demands as COD/N ratio of fermented sewage sludge, fermented food waste, and glycerol were 9.66, 6.75, and 12.2 g COD/g N, respectively. COD/N ratio indicates that VFAs from fermentation of combined sludge and food waste achieved similar denitrification capacity as glycerol but at a lower dosage rate. Glycerol consumption averages 13,000 lbs COD/day (~1600 gal/day) for mainstream flow of the Wards Island WRRF (as reported by DEP). Wards Island WRRF generates 185,000 lbs TS/day of sewage sludge solids which when fermented could potentially produce 26,000 lbs COD/day of VFAs. Therefore, 50% of VFAs produced could replace the glycerol consumed at Wards Island WRRF. Considering food waste fermentation, only 21,000 lbs TS/day is required to satisfy the same glycerol consumption. Hence, there will be excess availability of VFAs that could be repurposed for alternate high value products. Phase 3: Preliminary Approach for VFA Recovery Bench scale VFA extraction from the fermented sludge is conducted based on the modified two-step flash evaporation and vacuum distillation process in a batch system. This process would make VFA a higher value product by concentrating and purifying it for efficient transport. The extraction process also isolates ammonia from the fermented sludge, which would remove the nitrogen load when applied for denitrification. The bench scale VFA extraction apparatus depicted in Figure 3 consists of a 2-L pressure vessel that carries the centrate from thickened fermented sludge. The vessel is heated to 65oC using a heating mantle at atmospheric conditions. When the target temperature is achieved, the vessel is sealed from the atmosphere and a vacuum is applied at 100 mbara from the vacuum pump. At these temperatures and pressures, the solution in the pressure vessel is heated beyond the boiling point, thus causing rapid boiling that mimics flash evaporation. The sudden boiling volatilizes both VFAs and water from centrate which is diverted and bubbled through the alkaline scrubber containing NaOH solution. The scrubber solution is also subjected to the temperatures and vacuum as the pressure vessel to prevent water vapor from condensing and diluting the scrubbing solution. The high pH of the scrubber solution would bind the VFAs in the scrubbing solution while the water is evaporated out of the scrubber solution, hence concentrating VFAs within the scrubber solution. NaOH addition is included to maintain the high pH. The water that escapes the alkaline scrubber is condensed in a water-cooled condenser and the condensate is collected while any residual gases not removed by the scrubber and condenser are exhausted out the vacuum pump. This VFA extraction system is in the preliminary phase and requires further work to determine its efficacy. Various operational parameters and their impacts on recovery efficiency are to be studied such as fermentation centrate characteristics (VFA speciation and concentration, pH), and system temperature and pressures. Other methods for VFA condensation are to be studied such as cold-column condensation and mechanical vapor recompression to manipulate the temperatures that can affect condensation.