ADVANCING APPROACHES TO RESOURCE RECOVERY FROM WASTE MATERIALS

Abstract

The rising volume and complexity of municipal, agricultural, and industrial waste streams place considerable pressure on conventional waste management systems and create an opportunity to reframe these streams as feedstock sources for water, energy, and chemical recovery. Dairy facilities generate large volumes of flushed manure that drive on-site water demand and contribute to greenhouse gas emissions when stored or treated in lagoons, while the organic fraction of municipal solid waste (MSW), including food waste, paper and paperboard, and agricultural residues such as grass, together with agricultural residues such as feedlot manure, represents a large and diverse feedstock supply for biological conversion. This dissertation addresses barriers to resource recovery from these waste streams through studies that evaluate (1) a pilot-scale integrated vibrating screen (V-SC) and vibrating reverse osmosis (V-RO) membrane system for treating flushed dairy manure and (2) a laboratory-scale upflow leach bed reactor (ULBR) platform for converting diverse organic feedstocks into carboxylic acids for generation of high value volatile fatty acids (VFAs). Dairy facilities have a high demand for water, and treatment of on-site flushed dairy manure is challenging due to high organic and solids content. Pressure-driven membrane systems for dairy manure face severe fouling at high solids and high organic loading, requiring extensive pretreatment to maintain stable flux and recovery. To address barriers to the wide-scale implementation of water and energy recovery from flushed dairy manure, the performance of an integrated V-SC and V-RO membrane system was evaluated at pilot scale. Flushed dairy manure with influent concentrations of 46.2 ± 5.4 g/L total suspended solids (TSS), 47.9 ± 6.4 g/L chemical oxygen demand (COD), 2.4 ± 0.6 g/L total nitrogen (TN), and 0.8 ± 0.2 g/L total phosphorus (TP) was first pre-screened using V-SC at 125 µm, then treated by V-RO at 400 PSI and 25 °C. The V-RO unit achieved removals of 98.7 ± 0.6% of COD, 92.1 ± 1.7% of TN, and 84.1 ± 3.9% of ammonia, with complete elimination of TSS and TP, while maintaining an average flux of 6 GFD at approximately 70% recovery. Physical cleaning every 30 minutes and alkaline chemical cleaning approximately every 24 hours of operation were sufficient to sustain stable performance. The V-RO permeate generally met irrigation and livestock drinking water standards, although TN and 5-day biochemical oxygen demand occasionally exceeded surface water discharge limits in selected states. The combined V-SC and V-RO solids supported a biochemical methane potential of 227.8 ± 10.8 mL CH₄/g VS, exceeding values typically reported for flushed dairy manure. These results showed that a vibration-assisted membrane configuration can recover reusable water and concentrated solids from high-strength dairy waste without extensive pretreatment. Conventional anaerobic digestion of organic waste streams generates methane, but converting organic matter in these wastes into VFAs, particularly medium-chain carboxylic acids (MCCAs), through arrested anaerobic digestion (AAD) offers a route to higher-value chemical precursors for bioplastics, biofuels, and animal feed additives. AAD systems require reliable methanogenesis suppression, which is typically achieved through extreme operating conditions or chemical inhibitors that add cost and operational complexity. The influence of feedstock composition, hydraulic retention time (HRT), and reactor configuration on AAD performance in upflow leach bed reactors (ULBRs) is not well understood, and strategies that eliminate the need for external chemical addition while maintaining stable methanogenesis suppression remain underdeveloped. To address the methanogenesis suppression barrier in AAD, methanogenesis inhibition via lactic acid addition, which can be generated on-site as a byproduct, was evaluated in 10 L laboratory-scale ULBRs across three feedstocks representative of common U.S. organic waste streams: food waste, food waste co-digested with feedlot manure (FM), and alfalfa hay. All three feedstocks were tested at pH control setpoints of 5.25, 5.50, and 5.75 over 26-day batch runs. Food waste reactors maintained leachate pH below the lowest setpoint without intervention and fractional methane yield (%YCH4,theor) indistinguishable from zero throughout the experiment. On the other hand, both FM and alfalfa hay required lactic acid addition to maintain AAD condition. Among the three setpoints tested, pH 5.75 provided equivalent methanogenesis suppression with the lowest lactic acid demand in FM system. A %YCH4,theor threshold of 5% was introduced as a dosing criterion that more appropriately reflects methanogenic threat than gas-phase methane concentration alone, particularly in slow-degrading feedstocks. These results provided the first systematic evaluation of pH-controlled AAD in a ULBR system. To examine HRT as an operational parameter for controlling carboxylic acid (CA) production in ULBRs, the effect of HRT on VFA titers, cumulative CA yields, acid chain-length distribution, and methanogenic stability was evaluated across food waste (HRTs of 5, 8, and 11 days), grass (HRTs of 8, 31, and 78 days), and paper and paperboard mixed with feedlot manure (PPB + manure; HRTs of 14, 28, and 71 days). For food waste, total carboxylic acid (TCA) yield was largely insensitive to HRT (0.39 to 0.43 g COD/g VS), but longer HRTs provided increased C4-C7 fractions, with C6-C7 acids accounting for up to 21% of late-phase TCA at HRT 11 days. For grass, cumulative TVFA yield fell with increasing HRT (0.27 to 0.10 g COD/g VS across 8 to 78 days), while caproic and heptanoic acids remained observable across all conditions. For PPB + manure, lactic acid addition was required at HRTs of 14 and 28 days to restore methanogenesis suppression, while HRT 71 days needed none. Dosing at HRT 14 days resulted in desirable C4-C7 acid products, while HRT 71 days, though self-sufficient, produced very low overall CA yields. Correlation analysis confirmed that feedstock biodegradability, protein, and fat content drove total acid yield, while leachate pH and lignocellulosic content governed chain-length distribution. These findings showed that HRT is an important parameter for controlling the CA spectrum but does not override the dominant role of feedstock composition. To eliminate the need for external chemical addition in AAD, a series-fed leachate ULBR configuration was evaluated in which food waste and PPB + manure were physically layered with physical separation within a single reactor to mimic a two-reactor in-series configuration. Previous research conducted by our research group conducted in batch reactors indicated feeding products generated from food waste AAD to manure feedstocks can control methanogenesis and increase carbon chain length of products. Leachate was recirculated upward between the two feedstocks to assess the benefits observed at batch scale in an operational ULBR. The series-fed configuration produced a TCA yield of 0.323 g COD/g VS over 35 days at a 7-day HRT, exceeding the volatile solids-weighted theoretical baseline calculated from standalone food waste and standalone PPB + manure operation by 33.3% (p = 0.007), and supported a strong selectivity shift toward longer-chain CAs in the C4-C7 range without any lactic acid addition. %YCH4,theor in the series-fed configuration reached 1.4 ± 0.8%, a 69.3% reduction relative to PPB + manure standalone operation (4.6 ± 1.2%) despite the standalone receiving 203 g VS of external lactic acid. The ULBR in series therefore outperformed standalone operation of either feedstock for AAD targeting longer-chain CAs (C4-C7), delivering higher TCA yields and stronger methanogenesis suppression without lactic acid addition. These findings established the ULBR in series configuration as a more effective and operationally simpler approach than standalone ULBRs for converting mixed food waste and lignocellulosic-rich feedstocks into C4-C7 CAs.