Biofiltration of Methane Emission from Abandoned Oil and Gas Wells A Thesis Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Master of Science by Ziqi Li May 2025 © 2025 Ziqi Li ABSTRACT Abandoned oil and gas wells are significant sources of methane (CH4) emissions to the atmosphere. There are millions of abandoned wells in the United States, so there is a critical need for low-cost and sustainable approaches for mitigating CH4 emissions from these sources. In this study, we designed and tested methanotrophic biofilters as passive tools to mitigate CH4 emissions from abandoned wells. We first assessed the biokinetics of CH4 oxidation in batch tests with enrichments from wetland sediments as well as biofilms from membrane bioreactors with Methylosinus trichosporium OB3b, Methycystis parvus OBBP and Methylomicrobium album BG8 and two Methylomonas sp. at a range of temperatures spanning seasonal variation in air temperature in the Appalachian region, where a high density of abandoned wells are located. Cultures with the highest maximum reaction velocities (Vmax) were then tested in laboratory-scale flow-through bio-trickling filter configurations with packing materials including gravel, sponge, and biofilm carriers. Batch and column studies were used to parameterize biokinetic models to describe CH4 removal efficiencies in flow-through systems under field-relevant temperatures, flow rates, and CH4 concentrations. Modeling indicates that a 1 m³ sponge-packed biofilter can achieve from 19% to 34% methane removal at 22°C, and from 4% to 7% at 5°C in the Appalachian region assuming first order kinetic iii BIOGRAPHICAL SKETCH Ziqi Li, born and raised in Harbin, China, earned her bachelor’s degree in Environmental Science from Nanjing University in 2023. In August of the same year, she joined the Department of Civil and Environmental Engineering at Cornell University to start her Master of Science program under the guidance of Prof. Matthew C. Reid and Prof. Ruth E. Richardson. iv ACKNOWLEDGMENTS First and foremost, I need to express my deepest gratitude to my advisor, Prof. Matthew C. Reid, for his invaluable guidance and support over the past two years. He has instilled in me a serious and rigorous approach to scientific research, while his optimism, patience, and passion for environmental science have profoundly influenced me. Whenever I encountered challenges in my research, his guidance and encouragement became my motivation to keep moving forward. His mentorship will continue to inspire me throughout my academic and professional journey. The spring this year is very warm, so is his impression in my heart. The time I spent in Prof. Reid’s lab over the past two years will remain a cherished memory for a lifetime. I am also deeply grateful to my minor advisor, Prof. Ruth E. Richardson, whose course on Microbial Environmental Engineering has been instrumental in strengthening my foundation in environmental research. Her support and invaluable insights have been an essential part of my academic growth. Her enthusiasm, dedication, and patience have left a lasting impression on me, and I truly appreciate the opportunities to learn from her. Additionally, I would like to extend my heartfelt gratitude to my research group members, Michael, Hyun, Yi, Zihao, Yizhuo and Liz, who have provided academic support and valuable discussions throughout my research. I am also sincerely thankful to Yuan and Yanfei, whose help has been indispensable to my work. Their encouragement and companionship have made this experience even more meaningful. v Finally, I want to express my deepest appreciation to my family and friends in China for their unwavering love and care. Their steadfast support and encouragement have given me strength through every step of this journey. This experience has been both challenging and rewarding, and I am deeply grateful to everyone who has been a part of it. vi TABLE OF CONTENTS ABSTRACT .................................................................................................................................... ii BIOGRAPHICAL SKETCH ......................................................................................................... iii ACKNOWLEDGMENTS ............................................................................................................. iv CHAPTER 1 Introduction ............................................................................................................... 1 1.1 Background and Significance ............................................................................................... 1 1.2 Methanotrophic Biofiltration and Methane Oxidizing bacteria ............................................ 3 CHAPTER 2 Materials and Methods ............................................................................................. 9 2.1 Inoculum ............................................................................................................................... 9 2.2 Packing Materials ................................................................................................................ 13 2.3 Biofilter Set-up .................................................................................................................... 15 2.4 Modeling ............................................................................................................................. 19 2.4.1 Modeling of first-order kinetics and methane oxidation rate ....................................... 19 2.4.2 Modeling of zero-order kinetics and methane oxidation rate ...................................... 21 2.4.3 Modeling framework ................................................................................................... 23 2.5 Field investigation at Allegany State Park .......................................................................... 25 2.5.1 Study Site ..................................................................................................................... 25 2.5.2 Field Measurement Strategy ........................................................................................ 26 CHAPTER 3 Results and Discussion ........................................................................................... 28 vii 3.1 Inoculum ............................................................................................................................. 28 3.2 Influence of chemical and physical factors on biofilm attachment .................................... 31 3.3 Biofiltration test .................................................................................................................. 34 3.4 Methanotrophic activity test of materials ........................................................................... 42 3.5 Modeling the flow-through column experiment ................................................................. 43 3.5.1 The first-order kinetics and methane removal efficiency ............................................ 44 3.5.1 The zero-order kinetics and methane removal efficiency ............................................ 48 3.6 Comparison of packing material ......................................................................................... 49 3.7 Modeling the real-world conditions .................................................................................... 51 3.7.1 The zero-order kinetics and methane removal efficiency ............................................ 53 3.7.2 The first-order kinetics and methane removal efficiency ............................................ 57 3.8 Field investigation at Allegany State Park .......................................................................... 60 Chapter 5 Future Work ................................................................................................................. 63 REFERENCES ............................................................................................................................. 64 1 CHAPTER 1 Introduction 1.1 Background and Significance Methane is a powerful greenhouse gas and ranks as the second most significant contributor to global warming caused by human-generated emissions(1). Its global warming potential(GWP) is estimated to be 26–29 times greater than that of carbon dioxide (CO₂) over 100 years(2). The current concentration of CH₄ stands at 1.8 ppmv and has steadily risen since the onset of the industrial revolution(3). Abandoned oil and gas wells are significant sources of methane emissions to the atmosphere(4). There are estimated 4 million abandoned oil and gas (AOG) wells in the United States, and fugitive emissions of methane (CH4) from these wells are estimated to be equivalent to 8.2 million metric tons (MMT) of carbon dioxide equivalents (CO2eq) per year (95% confidence interval of 1.4 to 25.1 MMT CO2eq) (5). In Pennsylvania, abandoned oil and gas wells are estimated to account for 4–8% of the state's annual anthropogenic methane emissions(4). Most (58%) of abandoned wells are unplugged(5), providing a low-resistance pathway for the migration of hydrocarbon fluids from subsurface reservoirs to the atmosphere(6). Typical methane emission factors range from <1 to >20 CH4 g/h per well depending on the location and plugging status(7). The White House's 2022 report, U.S. Innovation to Meet 2050 Climate Goals: Assessing Initial R&D Opportunities, highlighted abandoned wells as a priority for achieving net-zero emissions by 2050, emphasizing the need for cost-effective and efficient methods to measure, reduce, or capture their emissions. Reflecting the importance of this issue, the 2021 Bipartisan Infrastructure Law (BIL) allocated $4.7 billion for plugging, remediating, and restoring orphaned well sites. 2 There are a series of challenges facing emerging efforts for large-scale remediation of abandoned wells. First, the locations of many abandoned wells are unknown, owing to the age of the wells (many dating to the late 19th or early 20th century) and poor record-keeping at the time of installation of older “legacy” wells(8). Second, CH₄ emissions from abandoned wells exhibit significant variability(4) and uncertainty, making it challenging to determine which wells should be prioritized for remediation to maximize cost-effectiveness and achieve the greatest reductions in CH₄ emissions. Measuring emissions from abandoned wells directly using chamber-based methods is labor-intensive, and it has been challenging to establish a clear relationship between emissions and well characteristics that can be observed at the wellhead(9, 10). The traditional mitigation strategy for abandoned oil and gas wells is to plug them, which is very expensive, costing about $20,000 to $40,000 per well(11). This cost is likely to limit the number of wells that can be realistically rehabilitated in the short term. A significant factor in these costs is that abandoned wells are often located in difficult-to-access areas, such as forests in rugged terrain, such as the Appalachian Mountains, where there are a large fraction of abandoned old oil wells(Fig.1) (10). Meanwhile, Some wells that are plugged but have vents have been found to emit exceptionally high levels of methane(12). And cement itself has a very large carbon footprint(13), raising questions about the sustainability of using cement materials for greenhouse gas reduction efforts. Therefore, there is clearly a critical need for more cost-effective, sustainable and easily deployable remediation methods. This approach will be tested in the Allegany State Park in the Appalachian region of western New York State (NYS), a site with a high density of abandoned oil and gas wells. 3 Fig.1. A large fraction of abandoned old oil wells are located in Pennsylvania (10) 1.2 Methanotrophic Biofiltration and Methane Oxidizing bacteria This study aims to develop cost-effective methane-oxidizing biofilters to mitigate methane emissions from abandoned wells. Methanotrophic biofilters are mostly passive systems, designed to operate with low energy consumption and maintain long-term stability(14). These biofilters typically use forced convection to mix exhaust gas with humid air, ensuring efficient methane oxidation under controlled conditions(14). And the biofilm layer which develops on the surface of packing material converts methane to carbon dioxide and water under aerobic conditions(15). Many studies have focused on the use of biofiltration to mitigate methane emissions in various environments, such as wastewater treatment plants, animal husbandry facilities, and landfill-cover soils (16-19). Under real-world operating conditions, these systems have achieved methane removal efficiencies of up to 85%(17). However, relatively little attention has been given to applying this approach for mitigating methane emissions from abandoned wells, which include higher CH4 concentrations than in other applications as well as the need to sustain efficient CH4 4 removal in cold weather. Unlike landfill cover soils, where CH₄ transport occurs primarily through diffusion, CH₄ emissions from abandoned wells are driven by advective fluxes, which means short empty bed contact times (EBCTs) and CH₄ oxidation rates must be fast to achieve effective removal. This is similar to the context of biofiltration of exhaust from animal husbandry operations, where ventilation fans induce advective airflows, leading to relatively short EBCTs (7–80 minutes in one study(17)). Despite these short EBCTs, removal efficiencies ranging from 20% to 85% have been reported (17). Importantly, CH₄ oxidation rates in biofilters increase linearly with CH₄ concentrations in the exhaust air, as higher CH₄ loadings supply more substrate for methanotrophs, promoting more biomass growth and higher oxidation capacity (17). This observation suggests that the higher CH₄ concentrations and/or flow rates associated with emissions from abandoned wells, often viewed as a treatment challenge, may instead promote faster oxidation kinetics and greater treatment capacity. CH₄ removal rates of up to 9 g CH₄ m⁻³ h⁻¹ have been reported in methanotrophic biofilters treating animal husbandry exhaust (17). This rate compares favorably with mean CH₄ emission factors for unplugged wells in the Appalachian region, reported at 9.6 g CH₄ m⁻³ h⁻¹. These findings indicate that biofilters with a volume of approximately 1 m³ could effectively mitigate CH₄ emissions from abandoned wells characteristic of the Appalachian region. The packing material used in methanotrophic biofilters are classified as organic and inorganic materials. Organic materials (e.g., compost, landfill cover soil) naturally contain nutrients and help maintain moisture content(20, 21). However, they are easy to compaction and degradation over time, which leads to a relatively short operational lifespan and increases operational costs(22). In contrast, inorganic materials (e.g., gravel(23), stone(24) and polyethylene rings(25)) contribute to the homogenous distribution of gas by enhancing the structural stability of the filter bed and 5 avoiding compaction. Furthermore, inorganic materials have been proven to achieve higher removal efficiency than compost based-bed biofilter(26). Therefore, this study aims to select an inorganic material that demonstrates a high methane elimination capacity and supports the growth of methanotrophic biomass effectively. Methanotrophic microorganisms oxidize methane to carbon dioxide to obtain metabolic energy and assimilate carbon into microbial biomass under aerobic conditions. Theoretically, the complete oxidation of 1 mol of methane requires 2 mol of oxygen, but methanotrophs usually require slightly less than 2 mol of oxygen, as part of the carbon is incorporated into cellular biomass or other organic compounds instead of being oxidized completely(27). Methanotrophs are classified into Type I, Type II group depending on the phospholipid fatty acid composition, carbon assimilation pathways and methane monooxygenase (MMO) type (15). Bacterial methanotrophs use methane monooxygenase (MMO) enzymes to initiate methane oxidation. Two primary forms of MMO have been identified: membrane-bound particulate methane monooxygenase (pMMO) and intracellular soluble methane monooxygenase(sMMO). The pMMO enzyme is associated with the specialized intracellular membrane structures characteristic of most methanotrophic bacteria and has a high affinity for methane. In contrast, the sMMO enzyme is localized within the cytoplasm and is present in only a subset of methanotrophic species(28). Thus, promoting pMMO expression allows for higher oxidation rates at low CH4 concentrations, which also means having smaller half- saturation constants (KS,CH4). The expression of the two forms of MMO is regulated by the ratio of copper to biomass(29). Under conditions of low copper level, the expression of sMMO is upregulated, whereas high copper level favors the expression of pMMO(30). Most methanotroph bacteria are mesophiles with the optimum operating temperature ranging from 20 to 37°C(31), 6 though methanotrophic communities can adapt to temperatures varying between 0 and 55°C(31, 32). Type I methanotrophs will be more active at low temperatures(5-10°C) than Type II methanotrophs(33, 34). In contrast, Type II methanotrophs are less sensitive in methane-limited conditions. Previous researches have demonstrated that mixed methanotrophic cultures often outperform pure cultures, particularly under fluctuating environmental conditions. (14, 19, 35) Therefore, a mixed methanotrophic culture may enhance biofilter performance, enabling effective operation in more variable and extreme environmental conditions. There are key differences between environments where methanotrophic biofiltration has been previously studied (e.g., animal husbandry) and abandoned wells emissions, with important implications for selection of methanotrophic biofilm communities. First, CH4 concentrations in gas flows from abandoned wells (thought to be >10% CH4, though measurements are scarce(4)) are probably two or more orders of magnitude higher than concentrations in animal husbandry exhaust (~0.1-2% CH4(17)). The biofiltration applications focused on animal husbandry exhaust or other related applications (e.g., removal of CH4 from atmospheric air(36)) have therefore focused on cultivating biofilm communities with a high affinity for CH4 (or a low KS,CH4) rather than optimizing communities for high CH4 concentrations, which involves increasing the maximum reaction velocity (Vmax). We will ensure adequate Cu supply through a recirculating liquid Cu-containing media (10 µM). In this study, we will focus on cultivating methanotrophic communities with maximum Vmax due to the high methane concentration in abandoned well gas flows. Second, since biofilters need to function throughout the year in Appalachian region (temperature varies from <-3°C to >28.3°C over year), we need to cultivate a more diverse methanotroph community that is more resilient to changing environmental conditions. Some 7 previous studies have investigated the performance of methanotrophic biofilters under low- temperature conditions, providing insight into their potential functionality in cold environments. We will use the Q10 model to describe the changes in Vmax and select the communities that are least sensitive to temperature as the inoculum for the biofilter. Previous studies have primarily focused on methane concentrations below 500 ppmv(36-38), whereas this study targets a higher concentration of 3% (30,000 ppmv) to reflect conditions more relevant to abandoned oil and gas wells. In addition, this study examines the temperature sensitivity of methane oxidation rates to evaluate the feasibility of biofiltration under varying seasonal temperatures. The aims of this study were to develop methanotrophic biofilters for the sustainable and cost-effective mitigation of CH4 emissions from abandoned wells. Given the high CH₄ concentrations and continuous flow rates characteristic of these wells, this study focuses on designing and optimizing biofilters capable of sustaining efficient CH₄ removal under real-world conditions, including cold weather scenarios. To achieve this, methanotrophic biofilm communities were enriched and selected based on their maximum reaction velocity (Vmax) and temperature sensitivity of biokinetic constants, ensuring effective methane oxidation across varying environmental conditions. Laboratory-scale biofilters were designed and tested with different packing materials, including gravel, sponge, and biofilm carriers, to assess the impact of packing material selection and gas loading rate on CH₄ oxidation rates and removal efficiencies. Additionally, mathematical models were developed to describe CH₄ removal kinetics under field- relevant conditions and predict the performance of large-scale biofilters in real-world applications. By integrating experimental and modeling approaches, this study provides a scientific foundation 8 for deploying methanotrophic biofilters as a practical and scalable solution for mitigating methane emissions from abandoned oil and gas wells. 9 CHAPTER 2 Materials and Methods 2.1 Inoculum Before conducting the biofiltration experiments, batch tests were performed to determine the optimal methanotroph biofilm communities. The selection was based on Vmax and the temperature- sensitivity of biokinetic constants. Two mixed cultures were enriched and tested. The first one (Culture A) was derived from wetland sediment collected from Beebe Lake located on Cornell University campus. The second culture (Culture B) was a mixture of aerobic methanotroph species , including Methylocystis parvus OBBP (Type II, pMMO and sMMO), Methylosinus trichosporium OB3b(Type II, pMMO), Methlyomicrobium album BG8(Type I, pMMO) and two Methylomonas sp.(Type I, pMMO), provided by Egidio Francisco Tentori(39). Methanotroph communities were enriched using nitrate mineral salts (NMS) media (Table 1)(40) containing 10 µM CuSO4 in 1L sealed media bottles with 500ml of culture and 500ml of headspace under constant shaking at 100 rpm. The bottles were fed with CH4 and air daily into the headspace to achieve a headspace concentration of 10% CH4 and 20% O2 (balance N2). 250 ml NMS media was refreshed every two weeks to maintain optimal growth conditions for the methanotrophic communities. Component Amount (per liter) MgSO4•7H2O 1.0 g KNO3 1.0 g CaCl2•H2O 0.2 g CuCl2.2H2O 1.4 mg (for 10 µM) 10 3.8% (w/v) solution Fe-EDTA 0.1 ml 0.1% (w/v) NaMo•4H2O 0.5 ml FeSO4•7H2O 0.5 mg ZnSO4•7H2O 0.4 mg MnCl2•7H2O 0.02 mg CoCl2•6H2O 0.05 mg NiCl2•6H2O 0.01 mg H3BO3 (boric acid) 0.015 mg EDTA 0.25 mg KH2PO4 0.26 g Na2HPO4•7(H2O) 0.62 g Biotin 0.02 mg Folic acid 0.02 mg Thiamine HCl 0.05 mg Ca pantothenate 0.05 mg Vitamin B12 0.001 mg Riboflavin 0.05 mg Nicotiamide 0.05 mg Table 1. Nitrate Mineral Salts (NMS) media composition 11 The enrichment time of incubation (2 months) was determined based on the methanotroph activity tests from batch experiments with liquid culture, ensuring the presence of active biomass for biofilters incubation. The methane oxidation rate was determined at 20°C in the triplicate batch experiments in a 132 ml continuously shaking vial containing 20 ml of culture by regularly measuring the decrease of methane concentration in the headspace from 2% to 0% with a gas chromatograph (Hewleft Packard 5890; column: Supelco Custom Column 052300 detector: FID) in intervals of 30 minutes. The methane oxidation rate was calculated based on the time-dependent uptake of the CH4 contained in the headspace. 𝑣 = (∆𝑝 𝑉!"#𝑀)/(𝑅 𝑇 ∆𝑡 ) (1) where v is the methane oxidation rate, Dp is the variation in partial pressure of methane in the headspace, Vgas is the volume of headspace, M is methane relative molecular mass (16 g/ mol), R is the universal gas constant (0.0821 L atm/ mol K), T is temperature, Dt is time interval. The kinetic parameters of these two liquid enrichment cultures were determined using triplicate batch experiments. A 5 mL aliquot of the liquid enrichment culture was transferred into 25 mL stoppered glass tubes, with initial methane concentrations of 0.5%, 1%, 5%, and 10% in the headspace. The tubes were incubated under shaking conditions at 100 rpm. Methane concentrations were continuously measured using gas chromatography (GC) at 30-minute intervals for a total of three measurements. Methanotrophs require sufficient O2 to oxidize methane aerobically, following a stoichiometric ratio of 2:1 (O₂:CH₄). When the CH₄ concentration is 10% (v/v), the O₂:CH₄ ratio in an air mixture is approximately 2.1:1, indicating a slight excess of oxygen. Therefore, when CH₄ is diluted to concentrations below 10% (v/v), oxygen is present in stoichiometric excess, ensuring that methanotrophs are not limited by oxygen availability.The 12 Vmax and Ks were determined using the linear transform of the Michaelis- Menten equation(37, 41) (assuming O2 is in excess): 𝑐 = 𝑉$ / % & 0 − 𝐾# (2) 𝑉$ = 𝑉$"'𝑋((3) where c is the initial methane concentration in the liquid phase and Xf is the biomass concentration in liquid culture. Statistical analysis of the linear regression was performed by the built-in function “fitlm” in MATLAB. The biomass concentration in liquid media was determined using the BCA protein assay. For protein extraction, 3-4 ml liquid enrichment culture sample was placed into 50 mL centrifuge tubes in triplicate. To each tube, 1 g of 1 mm borosilicate glass beads and 5 mL of cell lysis buffer (containing 1% SDS, 40 mM Tris-HCl,5 mM EDTA, PH adjusted to 8) were added. The glass beads facilitated mechanical disruption of microbial cells during vortexing. A blank control tube containing glass beads but no sample was prepared to check for contamination. The tubes were vortexed three times to facilitate cell disruption and then centrifuged at 14,000 g for 10 minutes. The supernatant (containing solubilized proteins) was carefully collected with a pipette. Protein concentration was measured following the BAC protein assay protocol,and estimated by assuming that protein represents 55% of dry weight biomass. Assuming equilibrium gas and liquid-phase concentrations, c was calculated by Henry’s law(42), the Henry’s law constants at different temperature were calculated by the van 't Hoff equation(42): 𝑐 = 𝑃)*!𝐾*,(4) 𝐾* = 𝐾*+ 𝑒𝑥𝑝 7−𝐶 ( , - − , -" )9 (5) 13 where PCH4 is the partial pressure of methane in the atmosphere, T is any given temperature, Tq is the standard state temperature of 303 K (30 °C) , KH is the Henry’s law constant at Tq, KHq is the Henry's law constant at the given temperature, C is a constant with dimension of kelvins(1600 K for methane)(42). Temperature(K) KH(mol m−3 Pa−1) 276 7.7 × 10−6 288 9.8 × 10−6 298 1.2 × 10−5 303 1.3 × 10−5 Table 2. Henry's law constants (CH4 in water)(37, 43, 44) The temperature sensitivity of biokinetic constants was assessed with incubation performed at 5, 15, and 25°C in temperature control incubators under constant shaking at 100 rpm. The temperature sensitivity of Vmax is described using Q10 models(45). 𝑄,. = (/#$%& /#$%' ) '( )&*)' (6) where Vmax2 and Vmax1 are the maximum reaction rates at the different temperatures T2 and T1(45). The Q10 value was performed by the built-in function “nlinfit” in MATLAB. 2.2 Packing Materials Three inorganic materials, including sponge (bulk density: 6.7 g/ L, porosity: 98.35%), plastic biofilm carriers (bulk density: 131.1 g/ L, particle size: 0.8-1.1cm, porosity: 67.5%) and cylindrical pieces of gravel material (bulk density: 1268.5 g/ L, particle size: 0.2-2.1 cm, porosity: 39.0%), 14 were used as the packing media. Before being packed into the biofilters, the three selected materials were incubated separately in 1 L bottles containing 500 ml liquid culture under constant shaking at 100 rpm at 25°C to facilitate uniform attachment of biofilm onto the surfaces of packing materials for 2 weeks. The porosity of packing materials (sponge, gravel, and biofilm carrier) was determined using a graduated cylinder water displacement method. First, a known volume of water was added to a graduated cylinder, and the initial water level was recorded. For sponge samples, a cube with a side length of 5 cm was cut and used for the measurement. For gravel and biofilm carrier, a known quantity of material was directly added to the graduated cylinder to determine the total material volume. The materials were gently added to the cylinder containing water, and the final water level was recorded. The volume of water displaced (the difference between the initial and final readings) represents the volume of water occupying the void spaces within the material. The porosity (ε) of the material was then calculated using the following equation: 𝜀 = /+$,-. /,/,$0 (7) Where Vwater is the volume difference between the two readings, Vtotal is the total volume of the added material. The density of the packing materials was determined based on their dry weight and volume. The dry mass of each sample was measured using an analytical balance. The volume of each material sample was measured following the same procedure described for porosity determination. The density(𝜌) was calculated using the following equation: 𝜌 = 0 /,/,$0 (8) 15 Where M is the dry mass of the material. During the inoculation of packing materials, we observed a slow rate of biofilm attachment on the materials surface. For the sponge material, the inoculation process involved immersing the sponge in a liquid enrichment culture for a week. However, after this period, the methane oxidation rate was only 0.01 mg CH₄ g⁻¹ sponge h⁻¹, indicating limited biofilm development on the sponge surface.Therefore, batch tests were conducted to assess the influence of chemical and physical factors on biofilm formation. Three factors were tested: temperature stress, mechanical stress and CH4 concentration. After 2 weeks of inoculating fresh gravel in liquid enrichment culture at room temperature under constant shaking at 100 rpm, 7 g gravel was removed and transferred into 50 ml sealed bottle containing 5ml NMS media. In the temperature stress group, the gravel was cultured at 5°C and 22°C, with the shaker maintained at 100 rpm. 10% methane was added daily to the headspace of the bottles. In the mechanical stress group, the bottles were placed on shakers at 120, 230, and 270 rpm respectively, at room temperature, with 10% methane added to the headspace daily. In the methane concentration group, 0%, 5%, and 10% methane were added to the headspace of three bottles, which were cultured at room temperature with constant shaking at 100 rpm. The culture was maintained for 8 days, and biomass concentration was measured daily from day 5 onward by sampling from the bottles. 2.3 Biofilter Set-up The laboratory scale biofilters were set up using three PVC columns (36 cm high, 7.5 cm inner diameter) packed with three materials previously detailed, as depicted in Fig. 2. For the biofilter with sponge, after about 3 weeks of acclimatization, visual observation of the cross-section 16 revealed that biofilm attachment was more substantial at the edges of the sponge, while the central area exhibited noticeably less biofilm growth. This observation suggests that the liquid media tended to preferentially flow along the edges of the sponge, resulting in uneven distribution of nutrients and microbial colonization. Due to the structural characteristics of the packing materials, such preferential flow was not observed in the biofilters packed with biofilm carrier and gravel. To address this issue, a 3 cm layer of 3 mm diameter glass beads was added to both the top and bottom of the column to ensure that the growth media could flow evenly across the surface of the packing material. The height of filter packing is 30 cm. The inlet gas was supplied via gas cylinder. Considering the explosion limit of CH4 and safety reasons, a gas mixture consisting of 3% CH₄ balanced with air was used as the feed gas. A manifold of rotameters is used to control flow rates to individual columns, and air samples can be collected from valves at the column inlet and outlet. A peristaltic pump is used for countercurrent flow of NMS growth media which is important for ensuring adequate supply of Cu and other nutrients, as well as maintaining moisture and enhancing CH4 transfer into liquid films. 17 Fig.2. Schematic of the experimental set-up: (1) gas cylinder, (2) flow-through bio-trickling filter column, (3) multiplexed peristaltic pump, (4) flow meter, (5) gas sampling ports. At day 0 of the biofiltration test, three biofilters were inoculated with 50 ml liquid enrichment culture at room temperature. The NMS medium was refreshed weekly to avoid nutrient limitation. Inlet and outlet samples from triplicate columns were collected every two days to monitor the CH4 removal efficiency. After a 29-days acclimatization period at 15.7 ml/min gas loading rate, the CH4 removal efficiency stabilized. The liquid media volume was also reduced to 20 ml to test its effect on CH4 transfer into liquid films (46). Every approximately 20 days, once the CH4 removal efficiency reached a steady state, the gas loading rate was increased gradually (from 15.7 ml/min to 31.7 ml/min, from 31.7 ml/min to 55 ml/min, from 55 ml/min to 125.3 ml/min, from 125.3 ml/min to 211.9 ml/min) to test the effects of different gas loading rates on CH4 removal performance (Table 3). 18 Gas flow rate (ml/min) EBCT (min) Sponge Gravel Biofilm carrier 15.7 99.18 39.47 68.31 31.7 49.12 19.54 33.83 55 28.31 11.26 19.50 125.3 12.42 4.94 8.56 211.9 7.34 2.92 5.06 Table 3. Gas flow rate and empty bed contact time (EBCT) for different packing materials The methane removal efficiency (RE) was calculated by mass balance. 𝑅𝐸 = )121)/3, )12 (9) where Cin is the inlet CH4 concentration (µmol /L), Cout is the outlet CH4 concentration (µmol /L). The elimination capacity (EC) for a component in the biofilter system was described as, 𝐸𝐶 = (𝐶23 − 𝐶456) × 1.6 × 1017𝑄/𝑉(10) where EC is the elimination capacity(g m-3h-1), Q is the gas flow rate (L/h), and V is the pore volume of the packed bed(m3). Before switching to the next flow rate condition, 5 ml of packing material were collected triplicate from the top, middle and bottom of three columns separately for the determination of biomass. The media was then replaced with packing material which has been pre-inoculated before moving on to the next flow rate condition. 19 Samples of each packing material were collected to determine the biokinetic parameters of the biomass attached to the media surface. Gravel and sponge samples were collected after 2 weeks of inoculation in liquid enrichment media, while the biofilm cairrier sample was collected from the biofilter in the end of experimental period. The methanotroph activity of the biofilm was assessed using the same procedure as described previously for measuring the biokinetics of the liquid enrichment culture. The maximum specific methane oxidation rate (qmax) and Ks were determined using the linear transform of the Michaelis- Menten equation(18, 37, 41): 𝑞 = 𝑣/𝑀( × 108(11) 𝑐 = 𝑞$"' / % 9 0 − 𝐾#(12) where q is the methane oxidation rate (umol CH4 g-1 biomass h−1); Mf is the total biomass (μg); qmax is the specific methane oxidation rate (umol CH4 g-1 biomass h−1). The determination of biomass followed a similar procedure to that described for liquid media. However, for protein extraction from the packing material, all samples used in each biokinetics batch test was transferred into a 50 mL centrifuge tube. To each tube, 1 g of 1 mm borosilicate glass beads and 5 mL of cell lysis buffer were added. The tubes were then vortexed to facilitate cell lysis and protein extraction. Finally, the total biomass concentration was calculated based on the total protein content extracted from the samples in each batch test vial. 2.4 Modeling 2.4.1 Modeling of first-order kinetics and methane oxidation rate The reaction kinetics of methanotrophic biofilter usually conforms to Michaelis-Menten kinetics. The methane oxidation rate in biofilters is expressed as: 20 𝑣 = 𝑐:24$"## 𝑞 = 𝑐:24$"## 9#$%%0 ;4<%0 (13) where v is the methane oxidation rate in a biofilter system, mg CH4 h−1; q is the methane oxidation rate for biofilm on packing media, mg CH4 g-1 biomass h−1; cl is the dissolved methane concentration in the biofilm, mg m-3; qmax is the maximum specific methane oxidation rate of biofilm on packing media, mg CH4 g biomass-1 h-1, cbiomass is biomass concentration in the biofilter, g m-3; Ks is the half saturation constant, mg m-3. If methane oxidation follows first-order kinetics (𝑐= ≪ 𝐾#). Assuming that the gas and liquid phase are in dynamic equilibrium, and applying Henry’s Law to convert the methane concentration from the gas phase to the liquid phase, it can be derived that: 𝑣 = 𝑐:24$"## 9#$%%0 ;4 = 𝑐:24$"## 9#$% ;4 %5 * (14) Where cg is the methane concentration in gas phase, mg m-3; H is the Henry’s law constant methane at 30 °C, 30 (mg/L)/(mg/L)(36). The first-order rate constant in a biofilter can be calculated as follows: k, = 𝑐:24$"## 𝑞$"' 𝐾# 𝐻 (15) where k1 is the first-order rate constant for methane oxidation, h-1. Assuming first-order kinetics and plug-flow conditions in the biofilter system, it can be derived that: 𝑄 𝑉 = 𝑘, ln (𝑐23/𝑐456) (16) where cin is the inlet methane concentration, g m-3; cout is the outlet methane concentration, g m-3; Q is the volumetric gas flow rate, m3h-1; V is the empty bed volume of the biofilter, m3. 21 The methane elimination capacity (EC) is expressed as(17): 𝑐$,=4! = (𝑐23 − 𝑐456) ln / 𝑐23𝑐456 0 (17) 𝐸𝐶 = (𝑐23 − 𝑐456)𝑄 𝑉 = 𝑘,(𝑐23 − 𝑐456) ln / 𝑐23𝑐456 0 = 𝑘,𝑐$,=4!(18) where cm, log is the logarithmic mean of the inlet and outlet concentration, g m-3; EC is the methane elimination capacity, g m-3 h-1. The methane removal efficiency (𝜂) can be derived as: 𝜂 = 𝐶23 − 𝐶456 𝐶23 = 1 − 𝑒1?'( / A)(19) where 𝜂 is the methane removal efficiency. 2.4.2 Modeling of zero-order kinetics and methane oxidation rate If methane oxidation follows zero-order kinetics (𝑐= ≫ 𝐾#), the zero order rate constant can be derived that: k. = 𝑣 = 𝑐:24$"##𝑞$"'(20) Assuming plug-flow conditions, it can be derived that, 𝑄 𝑉 = 𝑘. 𝑐23 − 𝑐456 (21) The methane elimination capacity (EC) is expressed as, 𝐸𝐶 = (𝑐23 − 𝑐456)𝑄 𝑉 = 𝑘.(22) The emission factor (𝛼) is expressed as: 𝛼 = 𝑄𝑐23(23) The methane removal efficiency (𝜂) can be derived as: 22 𝜂 = 𝐶23 − 𝐶456 𝐶23 = 𝑘.𝑉 𝛼 (24) where v is the methane oxidation rate in a biofilter system, mg CH4 h−1; q is the methane oxidation rate for biofilm on packing media, mg CH4 g-1 biomass h−1; cl is the dissolved methane concentration in the biofilm, mg m-3; qmax is the maximum specific methane oxidation rate of biofilm on packing media, mg CH4 g biomass-1 h-1, cbiomass is biomass concentration in the biofilter, g m-3; Ks is the half saturation constant, mg m-3; k0 is the zero-order rate constant, g m-3h-1; Q is the volumetric gas flow rate, m3h-1; V is the empty bed volume of the biofilter, m3; cin is the inlet methane concentration, g m-3; cout is the outlet methane concentration, g m-3; EC is the methane elimination capacity, g m-3 h-1; 𝛼 is the emission factors, g h-1; 𝜂 is the methane removal efficiency. 23 2.4.3 Modeling framework Fig.3. The modeling framework 24 Methane removal efficiency (𝜼) Condition Lab-scale experiments (k0, k1) Field condition (k0, k1) Gas flow rate (Q, m3/h) 0.012 0.016 ~ 0.15 (22°C) 0.015 ~ 0.13 (5°C) Inlet CH4 concentration (cin, %v/v) 3 10 ~ 90 Total biofilter volume (V, L) 1.5 Temperature (T, °C) 22 5, 22 Table 4. Parameters used in lab-scale and field-scale biofilters modeling For the real-world condition modeling, the following assumptions were made: (1) The average methane emission factor of abandoned oil and gas (AOG) wells is approximately 9.6 g/h per well (7). (2) Due to limited data on methane concentrations in exhaust gases from AOG wells, the inlet methane concentration was assumed to range from 10% to 90% (v/v). (3) Two temperatures, 5°C and 22°C, were selected to evaluate the impact of temperature on the methane removal efficiency of the biofilter under field conditions. These temperatures reflect typical low and moderate temperature scenarios in the field, allowing assessment of the biofilter performance and temperature sensitivity. Assume the biomass concentration and the porosity of packing materials are not changing from lab to the field condition. Samples of each packing material were collected to determine the biokinetic parameters of the biomass attached to the media surface. Specifically, gravel and sponge samples were collected after 2 weeks of inoculation in liquid enrichment media, while the biofilm carrier sample was 25 collected from the flow-through biofilter in the end of experimental period. The biokinetic parameters, including the half-saturation constant (Ks) and maximum reaction rate (Vmax), were determined and subsequently used to model the biofilter performance under both column experiment conditions and real-world scenarios, based on first-order (k₁) and zero-order (k₀) kinetics, using equation 15, 19, 20 and 24. Additionally, in the flow-through column experiments, the first-order reaction rate constant (k₁) was estimated using equation 18. The k₁ was then used to model the methane removal efficiency of lab-scale biofilter using equation 19. Furthermore, to evaluate the temperature sensitivity of methane oxidation, batch experiments were conducted at different temperatures using liquid enrichment cultures. The kinetic parameters and methane removal efficiency under field conditions at low temperature (5°C) was predicted using the Q10 model. 2.5 Field investigation at Allegany State Park 2.5.1 Study Site In this study, Allegany State Park was selected as the study site for preliminary field testing of CH4 emissions from abandoned wells. It also serves as a potential experimental site for the future practical application of biofiltration. Allegany State Park is situated in Cattaraugus County, directly north of the Allegheny National Forest in Pennsylvania. The Area has a high density of abandoned oil and gas wells, many of which are located in forested regions with steep elevation gradients (Fig.3). The methane emissions from these wells vary significantly(4). The climate in Allegany State Park has a wide annual temperature range, with cold winters and mild to warm 26 summers. In winter months, temperatures often range from lows of -10°C to highs of 0°C, while summer temperatures typically vary between 15°C and 25°C. The fieldwork was conducted on November 14, 2024, with temperature ranging from 2.7°C to 3.5°C. 2.5.2 Field Measurement Strategy Based on the well location data recorded in the DECinfo Locator, we used ArcGIS to map the distribution of wells within Allegany State Park, as shown in Fig.4. However, the well location data recorded in the DECinfo Locator are partially derived from old maps, historical drilling reports, or estimated data rather than direct GPS measurements or field surveys. As a result, the reliability of these records is relatively low, necessitating field investigations to further improve data accuracy. The park contains a total of 10 active oil or gas wells, drilled between 1956 and 1993. Additionally, there are 129 plugged and abandoned wells, some of which were drilled over 50 years ago. The field measurement helps to assess potential CH4 emission from these wells. Furthermore, 28 wells are classified as 'unknown status,' with earlier drilling dates, from1910 to 1940. It is also important to conduct field investigation to determine their precise locations, status, and potential CH4 emissions to improve data accuracy. Survey focused on the highest density of wells near the center of the map (Fig.4). The primary access road for driving is highlighted in yellow. For CH4 concentration measurements, we employed the GAS-ROVER II DETECTORS with methane sensor in Survey Mode, which allows for the detection of the full range of gas concentrations (0–100%) (Fig.5). The readings and GPS coordinates were recorded and outputted via DataLink Access to facilitate spatial analysis and emission assessments. Many of these wells were in remote or heavily vegetated areas, making 27 navigation challenging. To improve field efficiency, we used GARMIN etrex 10, a handheld GPS device for outdoor navigation, to pre-mark well locations and assist in navigation for well identification. Fig.4. The locations of documented plugged and unplugged wells in Allegany State Park, with key road for driving in yellow Fig.5. Direct measurement of CH4 concentration in the atmosphere around the well head 28 CHAPTER 3 Results and Discussion 3.1 Inoculum We first screened a methanotroph mixed culture with a high maximum reaction rate (Vmax) and low-temperature sensitivity of biokinetic constants from two mixed cultures, Culture A and Culture B (Table 5). At room temperature (22°C), the maximum reaction velocity (Vmax) of Culture A (54.34 ± 2.74 mmol CH₄ g⁻¹ biomass h⁻¹) and Culture B (58.48 ± 6.35 mmol CH₄ g⁻¹ biomass h⁻¹) were comparable, indicating that both cultures exhibited similar methane oxidation capacities. However, a significant difference was observed in the half-saturation constant (Ks), with Culture A having a lower Ks value (50.82 ± 1.64 μM) compared to Culture B (91.95 ± 10 μM). Since a lower Ks value represents higher methane affinity, Culture A demonstrated a stronger ability to use methane at lower concentrations. The biofilter needs to be operated and exposed to winter conditions, so it is important to use an inoculum that is not sensitive to temperature change. The effect of temperature was described by the Q10-values, which were 2.60 for Culture A and 2.41 for Culture B in the temperature range of 7-22 °C (Fig.6). Previous studies have reported a wide range of Q₁₀ values for methanotrophs in soil environment, for example Szafranek-Nakonieczna et al.(51) reported Q10-values of 1.0-9.67 for temperatures in the range of 10-20 °C and Einola et al.(32) reported Q10-values of 1.08-6.8 for 10-30 °C. Our results fall well within this previously reported range. The Q₁₀ values of Culture A and Culture B are close to each other, suggesting that two cultures exhibit similar temperature sensitivity, and neither culture experiences extreme fluctuations in methane oxidation rates over the tested temperature range. The relatively moderate Q₁₀ values also indicate that while temperature influences methane oxidation, it does not lead to sharp activity reductions, supporting 29 the potential application of these cultures in outdoor biofiltration under varying environmental conditions. Table 5. Comparison of biokinetics parameters of methanotroph mixed cultures under different temperatures(52, 53) 30 Fig. 6. Temperature sensitivity of methane oxidation rates for two cultures modeled using the Q₁₀ equation. 31 Fig.7. The linearized Michaelis–Menten plot for two methanotroph mixed cultures at different temperatures. The y-axis represents the initial dissolved methane concentration (c), while the x- axis represents the ratio of initial dissolved concentration to methane oxidation rate (c/v). The half- saturation constant (Ks) is determined as the negative y-axis intercept of the linear regression, while the maximum reaction velocity (Vmax) is given by the slope of the linear regression, normalized by biomass concentration (mmol CH4 g−1biomass h−1 ). (A) Culture A at 6.5 °C. (B) Culture A at 15 °C. (C) Culture A at 22°C. (D) Culture B at 7 °C. (E) Culture B at 15 °C. (F) Culture B at 22°C. 3.2 Influence of chemical and physical factors on biofilm attachment As shown in Fig.8, gravel incubated at 5°C and 22°C shows clear differences in biomass concentration. After five days of incubation, the biomass concentration at 5°C is higher than at 22°C. Additionally, during the next three days (days 5–8), the biofilm grown at 5°C continues to attach at a faster rate, reaching a peak biomass concentration of 49.91 µg/g at day 7. It can be concluded that decreasing temperature during biofilm development may increase the mass rate of biofilm accumulation. Most methanotroph bacteria are mesophiles with the optimum operating temperature ranging from 20 to 37°C(31). Therefore, the decreased microbial activity at low temperature (5 °C) should have resulted in reduced cell growth rate and biofilm accumulation. However, in the meantime, some studies suggested that under low temperature conditions, microorganisms may enhance their resistance to cold stress by increasing the synthesis of the secretion of extracellular polymers (EPS), thereby improving their survival ability(54, 55). Consistent with these findings, the results of this study indicate that low temperatures promote 32 methanotrophic biofilm accumulation, likely due to increased EPS production enhancing biofilm stability and cell adhesion. Fig.8. Biomass concentration development versus time for two temperatures: (●) 22 °C, (▼) 5 °C. Shaking frequency (Fig. 9) and methane concentration (Fig. 10) do not appear to have a strong influence on biofilm growth within the tested ranges. A unique feature was observed in the experiment, where a sudden decline in biomass concentration occurred after day 7, suggesting potential biofilm detachment (Fig. 8-10). This phenomenon has also been reported in many other studies(56). In Fig. 9, it was observed that under the 270 rpm condition, biofilm formation between days 5–7 occurred at a faster rate compared to the other two conditions, while the detachment rate between days 7–8 was also higher. This finding is consistent with the mechanism proposed by Picioreanu et al.(57), who suggested that under similar hydrodynamic conditions and biofilm strength, faster-growing biofilms have higher detachment rates than slower-growing biofilms. Therefore, a high biofilm growth rate may lead to instability in biofilm accumulation, ultimately resulting in sudden biomass loss or detachment. Fig.9 also suggests that the enhanced biofilm 33 growth rate under the 270 rpm condition may be attributed to the increased shaking frequency, which likely accelerates nutrient transport to the biofilm, thereby promoting biofilm development(58, 59). However, biofilm detachment is a complex process influenced by multiple factors, and further work needs to be done to determine the optimal shaking rate and methane concentration conditions for stable biofilm formation. Fig.9. Biomass concentration development versus time for three shaking frequencies: (●) 120 rpm, (▼) 230 rpm, (○) 270 rpm. 34 Fig.10. Biomass concentration development versus time for three methane concentrations: (●) 10%, (▼) 5%, (○) 0% 3.3 Biofiltration test The biofilters operated under ambient temperature conditions for a total of 145 days, with an approximately 40-day acclimatization process. During acclimatization period, it was operated at a relatively low gas flow rate of 15.7 mL/min to facilitate adaptation to the system, promote biofilm growth, and enhance the capacity of oxidizing methane loading. The gas loading rate was increased gradually. From Day 0 to Day 40 (Phase 1), the gas flow rate was maintained at 15.7 mL/min. Between Days 41 and 57 (Phase 2), the flow rate was increased to 31.7 mL/min. From Day 58 to Day 80 (Phase 3), the flow rate was further raised to 55.0 mL/min. During Days 81 to 131 (Phase 4), the gas flow rate was more than doubled to 125.3 mL/min. It should be noted that during this phase, the reactors were temporarily shut down between Day 92 and Day 117, and resumed operation on Day 118. Finally, during Days 131 to 145 (Phase 5), the flow rate was increased to 35 211.9 mL/min. The corresponding Empty Bed Contact Times (EBCT) for the different packing materials under each flow rate condition are summarized in Table 3. Fig.11 shows the time-course of methane removal efficiency in three flow-through bio-trickling filter columns packed with different packing materials: gravel, sponge and biofilm carrier, while Fig.12 shows the corresponding methane elimination capacity over time. Take the sponge-packed biofilter as an example. In the sponge-packed column, the removal efficiency stabilized at ~48- 50% after about one month of acclimatization (Phase 1) (Fig.11). On day 40, additional sponge medium was added after material samples collection, leading to a marked increase in methane removal efficiency to ~79-83% in Phase 2, even under a double gas flow rate condition. Fig.11. Methane removal efficiency over time for different packing materials 36 Fig.12. Methane elimination capacity over time for different packing materials In Phase 3 (starting day 57), the removal efficiency decreased from 82.71% to 58.30%, representing a one-third reduction, while the gas flow rate was increased 1.7-fold. Higher gas velocity shortens the empty bed contact time(EBCT), reducing the overall methane oxidation, but it also increases turbulence and gas-film mass transfer coefficients, which can improve methane mass transfer from gas phase into biofilm (60, 61). Given methane’s very low aqueous solubility in water (Henry’s law constant H ≈ 28 in dimensionless form at 25 ºC), biofiltration systems often face mass transfer limitations(62, 63). An increased gas flow can thus help overcome mass transfer limitation and promote methane transport into biofilm. 37 Similar trends occurred in Phase 4 and Phase 5. In Phase 4 (day 88), the methane removal efficiency decreased from 47.88% to 38.75%, representing a one-eighth reduction, when the gas flow rate was raised from 55 to 125.3 mL/min (approximately halving the EBCT). In Phase 5 (day 131), roughly doubling the gas flow rate (from 125.3 to 211.3 mL/min) further reduced the removal efficiency from 34.40% to 23.49%, which was also a decrease of less than one-half. Fig. 12 confirms this behavior: at the onset of Phases 3, 4, and 5, the methane elimination capacity (EC) exhibits a sharp increase when the gas flow rate is increased. Similar to our results, N. Josiane et al. also found that in a methane biofiltration system with an inlet methane concentration ranging from 0.13% to 1%, increasing the gas flow rate led to a decrease in methane conversion efficiency while simultaneously increasing the EC(60). The increase in EC was not proportional to the increase in gas flow rate(60). In the gravel-packed column and biofilm carrier-packed column, similar behavior was also observed. Fig.13 illustrates the relationship between methane removal efficiency and EBCT, while Fig.14 shows the relationship between methane elimination capacity (MEC) and EBCT. Overall, in all three biofilters, as EBCT increased (gas flow rate decreased), MEC showed a decreasing trend. However, there was no clear linear relationship between methane removal efficiency and EBCT. 38 Fig.13. Methane removal efficiency as a function of EBCT of different materials-packed biofilters: (A) sponge (y = 43.503x + 35.15, R² = 0.23) (B) gravel (y = 39.22x + 17.973, R² = 0.17) (C) biofilm carrier (y = -5.274x + 18.83 R² = 0.057) Fig.14. Methane elimination capacity (MEC) as a function of EBCT of different materials- packed biofilters: (A) sponge (y=-102.68x+70.552, R² =0.70) (B) gravel (y=-430.02x+98.201, R² =0.56) (C) biofilm carrier (y=-397.1x+80.896, R² =0.75) To determine whether methane removal was primarily driven by the attached methanotrophic biofilm on the surface of packing materials, or by the suspended culture in the recirculating liquid phase, the liquid medium was drained from the biofilters on day 29 (Fig.15). After draining the 39 liquid, the EC of sponge increased from 11.45 g CH4 m-3 h-1 to 11.87 g CH4 m-3 h-1, the EC of gravel increased from 17.20 g CH4 m-3 h-1 to 18.17 g CH4 m-3 h-1, and the EC of biofilm carrier increased from 6.05 g CH4 m-3 h-1 to 7.37 g CH4 m-3 h-1. This improvement in EC indicates that the attached methanotrophic biofilm was actively oxidizing methane, confirming that methane removal did not rely on the methanotrophs in the liquid phase. Such behavior aligns with prior studies showing that biofilters achieve higher methane removal rates when methane-oxidizing bacteria are immobilized on support media rather than suspended in liquid cultures(64). These results demonstrate that the developed biofilm had a significant capacity for methane oxidation, independent of the liquid enrichment culture. Fig.15. Methane elimination capacity (EC) before and after draining liquid media Fig.16 illustrates the biomass concentration of packing materials determined at the end of each phase to monitor how biomass accumulates during reactor operation. The concentration is 40 normalized by the dry weight of packing materials (mg biomass/ g packing material). Fig.17 shows the total biomass in the entire biofilter columns at the end of each phase. Due to sponge (bulk density: 6.7 g/ L, porosity: 98.35%) higher porosity and significantly lower bulk density compared to gravel (bulk density: 1268.5 g/L, porosity: 39.0%) and biofilm carrier (bulk density: 131.1 g/L, porosity: 67.5%), the biomass concentration of sponge, normalized by the dry weight of the packing materials, is considerably higher than that of the other two materials, while the total biomass within a given volume remains on the same order of magnitude. The total biomass of sponge and gravel in the reactor is higher than that of biofilm carrier. The biomass concentration of gravel stabilizes during Phases 4 and 5, with a peak value of 0.67 ± 0.008 mg/g material at the end of Phase 4. Other studies have reported that methane biofilters based on gravel can reach biomass concentrations of 0.23–1.17 mg/g in a stable state, which is consistent with the values obtained in this study (65). The biomass of the biofilm carrier continues to grow throughout the reactor operation, ultimately reaching a concentration of 0.14 ± 0.01 g/L. The biomass concentration of sponge gradually increases during Phases 1 to 3 and stabilizes during Phases 3 to 5. At the end of the operational process, the biomass concentration of sponge reached 0.30 ± 0.01 g/L. Compared to previous studies, the biomass measured for sponge and biofilm carrier in this study is relatively low. Sponge has a large surface area and porosity, which can retain moisture and nutrients for microbial growth. When fully colonized, it can achieve a biomass concentration of 10 mg/L(66). Biofilm carriers, due to their large specific surface area (often 100–800 m²/m³), can achieve a few grams of biomass per liter of packing volume(66). Therefore, the biofilm on the biofilm carrier in this study is still in a state of growing at the end of Phase 5 and can reach a higher biomass concentration. 41 Fig.16. Biomass concentration of different packing materials at the end of each phase Fig.17. Total biomass in bio-trickling columns at the end of each phase 42 3.4 Methanotrophic activity test of materials Fig.18 shows the linearized Michaelis–Menten plots for three packing materials. The maximum specific methane oxidation rate (qmax) and half-saturation constant (Ks) for materials were as follows: (1) qmax =157.26 ± 14.78 mg CH4 g biomass-1 h-1 and Ks = 49.41± 11.12 μM for gravel. (2) qmax =83.30 ± 9.95 mg CH4 g biomass-1 h-1 and Ks = 34.18 ± 12.65 μM for biofilm carrier. (3) qmax =254.4 ± 42.83 mg CH4 g biomass-1 h-1 and Ks = 22.6 ± 12.16 μM for sponge (Table 6). The qmax and Ks values of the three materials are all in the same order of magnitude as the kinetics parameters of methanotrophic biofilm in other materials, other studies’results show that qmax ranges from 0.04 mg CH4 g biomass-1 h-1 in samples from a biofilter integrated into a landfill cover system to 500 mg CH4 g biomass-1 h-1 in pure cultures, while qmax ranges from 16 mg m-3 in pure cultures to 1157 mg m-3 in methanotrophic cultures enriched from activated sludge and composted leaves(18, 67). Packing material Ks (μM ) qmax(mg CH4 g biomass-1 h-1 ) Sponge 22.6 ± 12.16 254.4 ± 42.83 Gravel 49.41± 11.12 157.26 ± 14.78 Biofilm carrier 34.18 ± 12.65 83.30 ± 9.95 Table 6. The biokinetic parameters of different packing materials Compared to other studies, the methanotrophic biofilm in the biofilters of this study has a relatively low methane affinity and a high methane oxidation rate. Among the three materials, the methanotrophic biofilm from the sponge has the highest methane oxidation rate and the highest 43 methane affinity. The biofilm on gravel has the second-fastest methane oxidation rate but a lower affinity for methane. This may suggest that the sponge maintains better fluid dynamic properties and its enormous surface area and internal pore network promote biofilm growth. Fig.18. The linearized Michaelis–Menten plot for three packing materials. The y-axis represents the initial dissolved methane concentration (c), µmol /L, while the y-axis represents the ratio of initial dissolved concentration to methane oxidation rate (c/q), g biomass h L-1. The half-saturation constant (Ks) is determined as the negative y-axis intercept of the linear regression (µM), while the maximum methane oxidation rate (qmax) is given by the slope of the linear regression (µmol CH4 g−1biomass h−1 ). (A) gravel. (B) biofilm carrier. (C) sponge. 3.5 Modeling the flow-through column experiment The inlet methane concentration was set at 3% (v/v), corresponding to an equilibrium methane concentration in the liquid phase of 39.01 μM (Table 7). When comparing this concentration to the half-saturation constants (Ks) of the different packing materials listed in Table 5, it is evident 44 that the liquid-phase methane concentration and Ks values are of the same order of magnitude. Therefore, it is difficult to clearly determine whether methane oxidation in the laboratory-scale biofilter followed zero-order or first-order kinetics under these conditions. To address this uncertainty, both zero-order and first-order kinetic models were applied to evaluate the potential methane removal efficiency in the biofilter system. Methane concentration in the headspace (cg, %) Methane equilibrium concentration in the liquid phase (ceq, μM ) 3 39.01 10 130.04 90 1170.35 Table 7. Equilibrium methane concentration in the liquid phase at different headspace methane concentrations 3.5.1 The first-order kinetics and methane removal efficiency Assuming the methane oxidation follows the first-order kinetics, Fig.19 shows the measured methane elimination capacity (EC) versus cm, log for sponge-packed biofilter. In Fig.19, the k1 value is estimated to be 4.63 ± 1.00 h-1, indicates that indeed first-order kinetics occur in sponge. This k1 value corresponds to the range of 0.3−6.6 h-1 that can be calculated from the data of other studies at cm, log from 0.4 to 22 g m-3(48, 68, 69). During the operation of the reactor, as time progresses, the data points move in the direction indicated by the arrow. This suggests that, with the gas flow rate held constant, both the methane elimination capacity (EC) and removal efficiency 45 (RE) increase over time. Fig.16 and Fig.17 shows there is significant biomass accumulation on both gravel and biofilm carrier from Phase 1 to Phase 2. Therefore, During the operational process of reactors, the total biomass in gravel-packed and biofilm carrier-packed biofilters was changing, and the first-order rate constants fluctuate significantly over time. As a result, when the measured methane elimination capacity and cm, log are plot, no obvious linear relationship is observed. Fig.19. Dependence of methane elimination capacity (EC) on concentration for sponge packed biofilter The first order kinetics (k1) in biofilters also can be calculated using the biokinetics parameters of methanotrophic biofilm in packing media and the total biomass using equation 15. The k1 for the biofilter with sponge is 9.07 ± 4.91 h-1, for gravel is 2.56 ± 0.62 h-1, and for biofilm carrier is 1.07 ± 0.42 h-1. This result is consistent with Fig.11, showing the removal efficiency of the sponge- packed biofilter outperforms than the other biofilters, indicating a higher methane oxidation rate 46 for the sponge. A comparison between the two methods of calculating the k1 for sponge shows that while the values are in the same order of magnitude, the k1 calculated using the biokinetic parameters is higher than the value obtained from measurements. This discrepancy indicates that the methane transfers process into the liquid film in the biofilter system encounters a certain degree of mass transfer resistance. Studies have shown that higher methane concentrations in the gas phase can reduce mass transfer limitations(70). In practical applications, the methane concentration in the exhaust gases from abandoned oil wells is probably higher than the inlet concentration (3%) used in this study. This would alleviate mass transfer resistance and enhance the reaction rate by increasing the concentration gradient and, consequently, the driving force for mass transfer. Fig. 20 shows the simulated methane removal efficiency for the laboratory-scale sponge-packed biofilter, using the k1 value from the filter in the model. According to the model, the 1.5L biofilter achieves 31.2% methane removal in the steady state, which is slightly higher than the measured methane removal efficiency of 23%. Using the k1 value estimated from the biokinetic batch experiments, the model predicts that the 1.5L biofilter will achieve 53.1% methane removal when packed with sponge, 11.7% with gravel, and 9.9% with biofilm carrier (Fig. 21). These results are close to the measured values of 14.7% methane removal for gravel and 13.6% for biofilm carrier. 47 Fig. 20. Simulated methane removal efficiency in the laboratory-scale sponge-packed biofilters, calculated using the k1 value estimated from the flow-through column (4.63 ± 1.00 h-1). The y- axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the total biofilter volume (V), m3 . Fig.21. Simulated methane removal efficiency in the laboratory-scale biofilters, calculated using the k1 values estimated from batch tests. The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. 48 3.5.1 The zero-order kinetics and methane removal efficiency Assuming the methane oxidation follows the zero-order kinetics, the zero order kinetics (k0) in biofilters were calculated using the biokinetics parameters of methanotrophic biofilm in packing media and the total biomass (equation 20). The k0 for the biofilter with sponge is 0.113 ±0.019 g m-3h-1, for gravel is 0.072 ± 0.006 g m-3h-1, and for biofilm carrier is 0.016 ± 0.002 g m-3h-1. Fig. 22 shows the simulated methane removal efficiency for the laboratory-scale biofilters, using the k0 values estimated from the biokinetic batch experiments. The model predicts that the 1.5L biofilter will achieve 100% methane removal when packed with sponge, 51.9% with gravel, and 19.9% with biofilm carrier. These predicted values are significantly higher than the actual methane removal efficiencies measured in the flow-through column experiments. Therefore, by comparing the prediction results of the first-order kinetics model, the zero-order kinetics model, and the experimental data, it can be concluded that the first-order model provides a better fit to the results. This suggests that, under the laboratory-scale biofilter conditions with an inlet methane concentration of 3% (v/v), methane oxidation behavior is more likely to follow first- order kinetics. 49 Fig.22. Simulated methane removal efficiency in the laboratory-scale biofilters, calculated using the k0 values estimated from batch tests. The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. 3.6 Comparison of packing material Fig.11 shows the methane removal efficiency of the biofilter packed with sponge consistently outperformed the other two packing materials, indicating within a fixed total bed volume, sponge offers a superior methane removal performance. First, sponge has an extremely high porosity resulting in a longer empty bed contact time (EBCT) compared to the other materials under the same total packed bed volume. Since methane removal efficiency is linearly proportional to EBCT (Fig.13), this theoretically allows for higher methane removal. Second, the sponge exhibited a higher Vmax and the highest methane affinity based on the biokinetics batch experiments with moist 50 packing materials, enabling a higher methane oxidation rate at a given methane concentration. In Fig.12, after day 50, the EC of the gravel exceeded that of the other materials, demonstrating that when EBCT is controlled, gravel can exhibit a more efficient methane removal performance. Table 8 summarizes several cases of application of biofiltration with inorganic packing materials for methane emission reduction under similar EBCT conditions. Estrada at al. (2014) reported methane elimination capacity between 24.3 g/m3/h to 63.8 g/m3/h over 100 days when their biofilter packed with polyurethane foam (PUF) was loaded methane ranging from 229 to 625 g CH₄ m⁻³ h⁻¹(46). The maximum EC achieved with PUF is comparable to the EC of sponge in this study. To our knowledge, very few studies have been performed with inlet methane concentration as high as the one considered there (3.0%v/v), as most focus on major methane emission sites where ambient air is only enriched to around 500 ppm methane(37). Higher methane concentrations, as previously mentioned, can mitigate mass transfer limitations and enhance methanotroph activity, thereby increasing EC. This explains why the ECs observed in this study exceed those reported in the limited literature on single-phase inorganic packing media for methane treatment. 51 Packing Material Inlet CH4 concentration (%v/v) Volumetric inlet load (g/m3/h) EBRT (min) CH4 EC (g/m3/h) CH4 RE (%) Reference Glass tubes 1.0 77 5 15.4 20 (68) Inorganic material 0.70–0.75 71 4.3 29 41 (26) stone 0.08-1 75 15 44.7 50 (71) polyurethane foam (PUF) 0.25 625 4 63.8 10.3 (46) Sponge 3.0 156 7.3 65.9 20 This study Gravel 3.0 156 2.9 111.8 14 This study Biofilm carrier 3.0 156 5.1 74.22 16 This study Table 8. Comparison of biofilter performance in the present work with methane biofiltration results from other studies at similar empty bed contact time (EBCT), assuming the highest gas flow rate in this study (211.9 ml/min) 3.7 Modeling the real-world conditions Fig. shows the setup of a pilot-scale methanotrophic biofilter system. CH₄ emitted from the well casing flows upward into the biofilter. To ensure sufficient oxygen for methane oxidation, air is pumped into the system using a fan. Therefore, the total gas flow entering the biofilter consists of methane-rich gas emitted from the abandoned well and supplemental air introduced through the fan system. The average methane emission factors(𝛼) of abandoned oil and gas wells is estimated 52 to be 9.6 g/h per well (7). Based on the stoichiometry of methane oxidation, The complete oxidation of CH4 requires O2 at a molar ratio of 1:2 (CH4:O2). Therefore, to fully oxidize 9.6 g/h CH4, O2 should be supplied at the rate of 1.2 mol/h. To meet this oxygen demand, the required air flow rate is estimated to be 0.138 m³/h at 22 °C and 0.130 m³/h at 5 °C, assuming standard atmospheric pressure and ideal gas behavior. For a well with a constant methane emission rate, the volumetric gas flow rate and inlet concentration are inversely proportional. Given methane concentrations ranging from 10% to 90% v/v (Cin ranging from 65.4 to 588.6 g/m³), the average exhaust gas flow rate is approximately 0.016 to 0.15 m³/h at 22 °C, using equation 23. Similarly, at 5 °C, when methane concentrations ranging from 10% to 90% v/v (Cin ranging from 69.7 to 627.5 g/m³), the average exhaust gas flow rate is approximately 0.015 to 0.13 m³/h at 5 °C. Therefore, the total volumetric inlet gas flow (Q) is 0.154 to 0.288 m³/h at 22 °C and 0.145 to 0.26 m³/h at 5 °C. Assume the biomass concentration and the porosity of these packing are not changing from lab to the field. Based on equation 6,15,19,20,24 and the laboratory-scale biofilter system, full-scale biofilter systems are designed to treat the exhaust gas from an unplugged AOG well with a consistent methane emission rate (𝛼) of 9.6 g/h with methane concentration ranging from 10% to 90% v/v at 22 °C and 5 °C. 53 Fig.23. The setup of a pilot-scale methanotrophic biofilter 3.7.1 The zero-order kinetics and methane removal efficiency Comparing Table 6 and Table 7, when the atmospheric methane concentration ranges from 10% to 90% (corresponding to methane equilibrium concentration in the liquid phase ranging from 130.04 to 1170.35 μM which are much higher than the half saturation constant of methanotrophs on the packing materials), the biofiltration process is more likely to follow the zero order kinetics under these condition. Therefore, the system performance under field conditions was first estimated using a zero-order kinetic model If the biofiltration system follows the zero-order kinetics in the field condtion, as shown in Fig. 24, using equation 24, when a sponge is used as the packing medium at 22 °C , biofilter total volumes of 0.26, 0.32, and 0.37 m³ are needed to achieve methane removal efficiencies of 50%, 75%, and 100%, respectively. When use the maximum volumetric gas flow rate (0.288 m³/h), the 54 corresponding minimum EBCT are 0.88, 1.09, and 1.26 hours, respectively. Under the low flow rate condition (0.154 m³/h), the corresponding EBCT are 1.65, 2.04, and 2.35 hours. When gravel is used as the packing medium, achieving the same methane removal efficiencies (50%, 75%, and 100%) at 22 °C using equation 24, we estimate that larger biofilter volumes of 0.52, 0.64, and 0.74 m³ are needed. However, the EBCT are shorter, at 0.70, 0.87, and 1.00hours, respectively, under a high flow rate of 0.288 m³/h. When operated at a lower flow rate (0.154 m³/h), the corresponding EBCT values increase to 1.32, 1.62, and 1.87 hours, respectively. Because the sponge has a 1higher porosity, its pore volume is greater within the same total volume, which leads to a longer EBCT. When using the biofilm carrier as the medium requires biofilter volumes of 0.82, 1.00, and 1.16 m³ to achieve methane removal efficiencies of 50%, 75%, and 100% at 22 °C using equation 24. The corresponding EBCT are notably longer, at 1.91, 2.33, and 2.70 hours, respectively, under a high flow rate of 0.288 m³/h. When operated at a lower flow rate of 0.154 m³/h, the EBCT values further increase to 3.56, 4.35, and 5.05 hours, respectively. 55 Fig.24. Biofilter design for treatment of exhaust gas from an unplugged AOG well at a constant methane emission rate of 9.6 g/h at 22 °C, determined using equation 24. . The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. The Q10 model predicts the maximum methane oxidation rate (qmax) at 5 °C: qmax =28.16 mg CH4 g biomass-1 h-1 for gravel; qmax =14.91 mg CH4 g biomass-1 h-1 for biofilm carrier; qmax =45.56 mg CH4 g biomass-1 h-1 for sponge. As shown in Fig.25, when the sponge is used as the packing medium at 5 °C , biofilter total volumes of 0.62, 0.75, and 0.87 m³ are needed to achieve methane removal efficiencies of 50%, 75%, and 100%, respectively. When use the maximum volumetric gas flow rate (0.26 m³/h), the corresponding minimum EBCT are 2.34, 2.83, and 3.28 hours, respectively. Under the low flow rate condition (0.145 m³/h), the corresponding EBCT are 4.19, 5.07, and 5.88 hours. 56 When gravel is used as the packing medium, achieving the same methane removal efficiencies (50%, 75%, and 100%) at 5 °C using equation 24, we estimate that biofilter volumes of 1.23, 1.51 and 1.74 m³ are needed. The EBCT are at 1.84, 2.26, and 2.61 hours, respectively, under a high flow rate of 0.26 m³/h. When operated at a lower flow rate (0.145 m³/h), the corresponding EBCT values increase to 3.31 4.06, and 4.68 hours, respectively. When the biofilm carrier is used as the packing media at 5 °C, biofilter total volumes of 1.94, 2.37, and 2.74 m³ are needed to achieve methane removal efficiencies of 50%, 75%, and 100%, respectively. When use the maximum volumetric gas flow rate (0.26 m³/h), the corresponding minimum EBCT are 5.00, 6.11, and 7.06 hours, respectively. Under the low flow rate condition (0.145 m³/h), the corresponding EBCT are 8.96, 10.95, and 12.66 hours. 57 Fig.25. Biofilter design for treatment of exhaust gas from an unplugged AOG well at a constant methane emission rate of 9.6 g/h at 5°C, determined using equation 24. The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. 3.7.2 The first-order kinetics and methane removal efficiency If the biofiltration system follows the first-order kinetics in the field condition, assuming the average exhaust gas flow rate (Q) is approximately 0.154 to 0.288 m³/h (assuming methane concentration ranging from 10%-90%) at 22 °C, using equations 15 and 19, Fig.26 shows the simulated methane removal efficiencies versus the total biofilter volume. The solid line represents the condition of higher methane concentration (90%) and lower gas flow rate (Q=0.0.154 m³/h), while the dash line represents the condition of higher methane concentration (90%) and lower gas flow rate (Q=0.288 m³/h).As shown in Fig.26, the biofilter packed with sponge achieved significantly higher methane removal efficiency compared to the other two 58 packing materials. Specifically, a 1 m3 sponge- packed biofilter is estimated to achieve 19% methane removal under gas flow rate of 0.288 m³/h (CCH4 = 10%) and 34% under gas flow rate of 0.154 m³/h.( (CCH4 = 90%)). The corresponding EBCT are estimated to be 6.36 h and 3.40 h. Fig.26. Biofilter design for treatment of exhaust gas from an unplugged AOG well at a constant methane emission rate of 9.6 g/h with gas loading rate from 0.154 to 0.288 m³/h at 22 °C, determined using equation 19. The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. When the biofilter is operated at 5 °C, as previously mentioned, the biokinetics parameters at 5 °C are estimated using the Q10 model: qmax =28.16 mg CH4 g biomass-1 h-1 for gravel; qmax =14.91 mg CH4 g biomass-1 h-1 for biofilm carrier; qmax =45.56 mg CH4 g biomass-1 h-1 for sponge. Using the equation 15 and 19, the Fig.27 is plotted to illustrate the variations in methane removal efficiency with respect to the total volume of biofilter at 5 °C, with gas loading rate ranging from 0.145 to 0.260 m³/h (methane concentration ranging from 10%-90%). As shown in 59 Fig.27 , a 1 m3 sponge- packed biofilter is estimated to achieve 4% under gas flow rate of 0.145 m³/h (CCH4 = 10%) and 7% under gas flow rate of 0.26 m³/h.( (CCH4 = 90%)). The corresponding EBCT are estimated to be 6.76 h and 3.77 h. Fig.27. Biofilter design for treatment of exhaust gas from an unplugged AOG well at a constant methane emission rate of 9.6 g/h with gas loading rate from 0.145 to 0.260 m³/h at 5 °C, determined using equation 19. The y-axis represents the methane removal efficiency (𝜂), %, while the x-axis represents the biofilter volume (V), m3. Combining the results from the first order kinetics and zero-order kinetics model, and assuming the average methane emission rate is 9.6g/h with methane concentrations ranging from 10% to 90% v/v, the sponge is identified as the most suitable packing material. A 1 m3 sponge- packed biofilter is estimated to achieve methane removal efficiencies ranging from 19% to 34% at 22°C, and from 60 4% to 7% at 5°C assuming the first order kinetics govern the system performance.. The methane removal efficiency estimated by the zero-order model is higher than that predicted by the first- order model. This result is reasonable because, in methane oxidation process, the reaction rate increases with the methane concentration. At low methane concentrations, the process tends to follow first-order kinetics, while at high concentrations, it approaches zero-order kinetics. Therefore, under real-world conditions, when methane emissions from abandoned oil and gas (AOG) wells are relatively high, the biofiltration system is expected to perform more efficiently. 3.8 Field investigation at Allegany State Park We investigated 28 AOG wells recorded in the DECinfo database, including 7 plugged and 21 unknown status wells, but we only found 3 unplugged and 5 plugged wells based on the geographic coordinates recorded in the database. The methane concentrations observed near the wellheads shows variation. For the 3 unplugged wells, the methane concentration in the air ranged from 11- 39 ppm. For the 5 plugged wells, out of which only 2 showed detectable methane, had methane concentrations ranging from 7-17 ppm. These findings suggest that unplugged wells tend to have higher methane concentrations compared to plugged wells, which is consistent with general trends observed in studies of methane emissions from abandoned oil and gas wells (7). Unplugged wells are more likely to leak methane into the atmosphere. Some plugged wells still exhibit relatively high methane concentrations, suggesting the possibility of potential leakage in certain plugged wells. However, the natural conditions within Allegany State Park present significant obstacles to identifying and cataloging abandoned oil wells. Dense vegetation and steep terrain obscure 61 wellheads, making visual detection difficult. Database inaccuracies in the DECinfo geographic coordinates may have contributed to the small number of wells located (only 8 out of 28). The limited sample size might constrain the representativeness of the results. 62 Chapter 4 Conclusion This study developed methanotrophic biofilters for the sustainable and cost-effective mitigation of CH4 emissions from abandoned wells. In contrast to traditional applications of methanotrophic biofilters, such as landfill biocovers and livestock barn exhaust treatment, this study focused on an operational environment with significantly higher CH₄ concentrations. Additionally, ensuring effective CH₄ removal under cold weather conditions was a key focus to enhance seasonal applicability. The inoculum used for the laboratory-scale biofilters was selected based on the highest oxidation rate and temperature sensitivity. Notably, lower temperatures (5°C) facilitated the rapid formation of biofilms. After 145 days of operation, the results showed that the biofiltration system with sponge had significantly higher methane removal efficiency than gravel and biofilm carrier at the same biofilter volume. At higher gas flow rates, although the removal efficiency was reduced, the elimination capacity (EC) was still significantly improved, indicating that the gas flow promoted the mass transfer of methane to the biofilm. The methane emission rate from abandoned wells is estimated about 9.6 g/h. Therefore, combining laboratory data with mathematical modeling, a 1 m3 biofilter packed with sponge can achieve from 19% to 34% methane removal at 22°C, and from 4% to 7% at 5°C in the Appalachian region assuming first order kinetics. 63 Chapter 5 Future Work Pilot-scale methanotrophic biofilters will be tested at abandoned wells in Allegany State Park to demonstrate feasibility as a cost-effective solution for mitigating CH₄ emissions. Biofilters optimized for high CH₄ concentrations will undergo field testing to assess their efficiency, durability, and performance across different seasons. By integrating laboratory findings with field data, the study will evaluate the effectiveness of model-based biofilter design, establish their long- term viability under varying environmental conditions, and compare their cost-effectiveness to traditional well-plugging methods. 64 REFERENCES 1. Saunois M, Stavert AR, Poulter B, Bousquet P, Canadell JG, Jackson RB, et al. The global methane budget 2000–2017. Earth System Science Data Discussions. 2019;2019:1- 136. 2. USEPA. Understanding Global Warming Potentials. United States Environmental Protection Agency. 2025. 3. Kirschke S, Bousquet P, Ciais P, Saunois M, Canadell JG, Dlugokencky EJ, et al. Three decades of global methane sources and sinks. Nature Geoscience. 2013;6(10):813- 23. 4. Kang M, Kanno CM, Reid MC, Zhang X, Mauzerall DL, Celia MA, et al. Direct measurements of methane emissions from abandoned oil and gas wells in Pennsylvania. Proceedings of the National Academy of Sciences. 2014;111(51):18173-7. 5. Interior USDot. Assessing Methane Emissions from Orphaned Wells to Meet Reporting Requirements of the 2021 Infrastructure Investment and Jobs Act (BIL): Federal Program Guidelines. 2022. 6. Kang M, Baik E, Miller AR, Bandilla KW, Celia MA. Effective Permeabilities of Abandoned Oil and Gas Wells: Analysis of Data from Pennsylvania. Environmental Science & Technology. 2015;49(7):4757-64. 7. Williams JP, Regehr A, Kang M. Methane emissions from abandoned oil and gas wells in Canada and the United States. Environmental science & technology. 2020;55(1):563-70. 8. Dilmore RM, Sams JI, III, Glosser D, Carter KM, Bain DJ. Spatial and Temporal Characteristics of Historical Oil and Gas Wells in Pennsylvania: Implications for New Shale Gas Resources. Environmental Science & Technology. 2015;49(20):12015-23. 9. Williams JP, Regehr A, Kang M. Methane Emissions from Abandoned Oil and Gas Wells in Canada and the United States. Environmental Science & Technology. 2021;55(1):563-70. 10. Boutot J, Peltz AS, McVay R, Kang M. Documented Orphaned Oil and Gas Wells Across the United States. Environmental Science & Technology. 2022;56(20):14228-36. 11. Raimi D, Krupnick AJ, Shah J-S, Thompson A. Decommissioning Orphaned and Abandoned Oil and Gas Wells: New Estimates and Cost Drivers. Environmental Science & Technology. 2021;55(15):10224-30. 65 12. Kang M, Christian S, Celia MA, Mauzerall DL, Bill M, Miller AR, et al. Identification and characterization of high methane-emitting abandoned oil and gas wells. Proceedings of the National Academy of Sciences. 2016;113(48):13636-41. 13. Miller SA. Supplementary cementitious materials to mitigate greenhouse gas emissions from concrete: can there be too much of a good thing? Journal of Cleaner Production. 2018;178:587-98. 14. La H, Hettiaratchi JPA, Achari G, Dunfield PF. Biofiltration of methane. Bioresource Technology. 2018;268:759-72. 15. Hanson RS, Hanson TE. Methanotrophic bacteria. Microbiol Rev. 1996;60(2):439- 71. 16. Gebert J, Gröngröft A. Performance of a passively vented field-scale biofilter for the microbial oxidation of landfill methane. Waste Management. 2006;26(4):399-407. 17. Melse RW, van der Werf AW. Biofiltration for Mitigation of Methane Emission from Animal Husbandry. Environmental Science & Technology. 2005;39(14):5460-8. 18. Brandt EMF, Duarte FV, Vieira JPR, Melo VM, Souza CL, Araújo JC, et al. The use of novel packing material for improving methane oxidation in biofilters. Journal of environmental management. 2016;182:412-20. 19. Fedrizzi F, Cabana H, Ndanga ÉM, Cabral AR. Biofiltration of methane from cow barns: Effects of climatic conditions and packing bed media acclimatization. Waste Management. 2018;78:669-76. 20. Farrokhzadeh H, Hettiaratchi JPA, Jayasinghe P, Kumar S. Aerated biofilters with multiple-level air injection configurations to enhance biological treatment of methane emissions. Bioresource technology. 2017;239:219-25. 21. Park S, Lee C-H, Ryu C-R, Sung K. Biofiltration for reducing methane emissions from modern sanitary landfills at the low methane generation stage. Water, air, and soil pollution. 2009;196:19-27. 22. Estrada JM, Kraakman NB, Lebrero R, Muñoz R. A sensitivity analysis of process design parameters, commodity prices and robustness on the economics of odour abatement technologies. Biotechnology advances. 2012;30(6):1354-63. 23. Girard M, Ramirez AA, Buelna G, Heitz M. Biofiltration of methane at low concentrations representative of the piggery industry—influence of the methane and nitrogen concentrations. Chemical Engineering Journal. 2011;168(1):151-8. 66 24. Ramirez AA, García-Aguilar BP, Jones JP, Heitz M. Improvement of methane biofiltration by the addition of non-ionic surfactants to biofilters packed with inert materials. Process Biochemistry. 2012;47(1):76-82. 25. Cáceres M, Dorado AD, Gentina JC, Aroca G. Oxidation of methane in biotrickling filters inoculated with methanotrophic bacteria. Environmental Science and Pollution Research. 2017;24:25702-12. 26. Nikiema J, Bibeau L, Lavoie J, Brzezinski R, Vigneux J, Heitz M. Biofiltration of methane: An experimental study. Chemical Engineering Journal. 2005;113(2):111-7. 27. Gilman A, Laurens LM, Puri AW, Chu F, Pienkos PT, Lidstrom ME. Bioreactor performance parameters for an industrially-promising methanotroph Methylomicrobium buryatense 5GB1. Microbial cell factories. 2015;14:1-8. 28. Scott D, Brannan J, Higgins I. The effect of growth conditions on intracytoplasmic membranes and methane mono-oxygenase activities in Methylosinus trichosporium OB3b. Microbiology. 1981;125(1):63-72. 29. Han JI, Semrau JD. Quantification of gene expression in methanotrophs by competitive reverse transcription-polymerase chain reaction. Environmental microbiology. 2004;6(4):388-99. 30. Lontoh S, Semrau Jeremy D. Methane and Trichloroethylene Degradation byMethylosinus trichosporium OB3b Expressing Particulate Methane Monooxygenase. Applied and Environmental Microbiology. 1998;64(3):1106-14. 31. Humer M, Lechner PP. Alternative approach to the elimination of greenhouse gases from old landfills. Waste Management and research. 1999;17(6):443-52. 32. Einola J-KM, Kettunen RH, Rintala JA. Responses of methane oxidation to temperature and water content in cover soil of a boreal landfill. Soil Biology and Biochemistry. 2007;39(5):1156-64. 33. Börjesson G, Sundh I, Svensson B. Microbial oxidation of CH4 at different temperatures in landfill cover soils. FEMS Microbiology Ecology. 2004;48(3):305-12. 34. Dunfield PF. Methanotrophy in Extreme Environments. eLS, John Wiley & Sons Ltd, Chichester. 2009. 35. R. Syed SS, K. Tate, B.H.A. Rehm. Does acidification of a soil biofilter compromise its methane-oxidising capacity? Biol Fertil Soils. 2016;pp. 1-11. 67 36. Yoon S, Carey JN, Semrau JD. Feasibility of atmospheric methane removal using methanotrophic biotrickling filters. Appl Microbiol Biotechnol. 2009;83(5):949-56. 37. He L, Groom JD, Wilson EH, Fernandez J, Konopka MC, Beck DA, et al. A methanotrophic bacterium to enable methane removal for climate mitigation. Proceedings of the National Academy of Sciences. 2023;120(35):e2310046120. 38. Y. Cai YZ, P. L. E. Bodelier, R. Conrad, Z. Jia. Conventional methanotrophs are responsible for atmospheric methane oxidation in paddy soils. Nat Commun. 2016;7, 1– 10. 39. Tentori EF. Membrane Bioreactors for the Study of Aerobic Methanotrophs: Exploring Links Between Biomarkers and Methanotrophic Activity, and Treatment of Anaerobic Digestion Wastewaters [Ph.D.]. United States -- New York: Cornell University; 2021. 40. Whittenbury R, Phillips K, Wilkinson J. Enrichment, isolation and some properties of methane-utilizing bacteria. Microbiology. 1970;61(2):205-18. 41. Eppley RW, Rogers JN, McCarthy JJ. Half-saturation constants for uptake of nitrate and ammonium by marine phytoplankton 1. Limnology and oceanography. 1969;14(6):912- 20. 42. Sander R. Compilation of Henry’s law constants (version 4.0) for water as solvent. Atmos Chem Phys. 2023;23,10901–12440. 43. Sander R. Compilation of Henry's law constants (version 5.0.0) for water as solvent. Atmos Chem Phys. 2023;23:10901-2440. 44. Wiesenburg DA, Guinasso NL, Jr. Equilibrium solubilities of methane, carbon monoxide, and hydrogen in water and sea water. Journal of Chemical & Engineering Data. 1979;24(4):356-60. 45. Duc NT, Crill P, Bastviken D. Implications of temperature and sediment characteristics on methane formation and oxidation in lake sediments. Biogeochemistry. 2010;100:185-96. 46. Estrada JM, Lebrero R, Quijano G, Pérez R, Figueroa-González I, García-Encina PA, et al. Methane abatement in a gas-recycling biotrickling filter: Evaluating innovative operational strategies to overcome mass transfer limitations. Chemical Engineering Journal. 2014;253:385-93. 68 47. Urmann K, Gonzalez-Gil G, Schroth MH, Hofer M, Zeyer J. New field method: gas push-pull test for the in-situ quantification of microbial activities in the vadose zone. Environ Sci Technol. 2005;39(1):304-10. 48. Streese J, Stegmann R. Microbial oxidation of methane from old landfills in biofilters. Waste Management. 2003;23(7):573-80. 49. Streese Kleeberg J, Rachor I, Stegmann R, editors. In situ measurement of methane oxidation activity in landfill covers by gas push pull tests. Proceedings Sardinia; 2009. 50. Urmann K, Schroth MH, Noll M, Gonzalez-Gil G, Zeyer J. Assessment of microbial methane oxidation above a petroleum-contaminated aquifer using a combination of in situ techniques. Journal of Geophysical Research: Biogeosciences. 2008;113(G2). 51. Szafranek-Nakonieczna A, Wolińska A, Zielenkiewicz U, Kowalczyk A, Stępniewska Z, Błaszczyk M. Activity and Identification of Methanotrophic Bacteria in Arable and No- Tillage Soils from Lublin Region (Poland). Microbial Ecology. 2019;77(3):701-12. 52. Gebert J, Groengroeft A, Miehlich G. Kinetics of microbial landfill methane oxidation in biofilters. Waste Management. 2003;23(7):609-19. 53. Bédard C, Knowles R. Physiology, biochemistry, and specific inhibitors of CH4, NH4+, and CO oxidation by methanotrophs and nitrifiers. Microbiological Reviews. 1989;53(1):68-84. 54. Sun H-W, Chen C-Z, Wu C-F, Zhao H-N, Yu X, Fang X-H. [Effect of Temperature on Nitrogen Removal Performance and the Extracellular Polymeric Substance (EPS) in a Sequencing Batch Reactor (SBR)]. Huan Jing Ke Xue. 2017;38(11):4648-55. 55. Wang C, Cui Y, Qu X. Mechanisms and improvement of acid resistance in lactic acid bacteria. Archives of Microbiology. 2018;200(2):195-201. 56. Rishell S, Casey E, Glennon B, Hamer G. Characteristics of a methanotrophic culture in a membrane-aerated biofilm reactor. Biotechnology progress. 2004;20(4):1082- 90. 57. Picioreanu C, Van Loosdrecht MC, Heijnen JJ. Effect of diffusive and convective substrate transport on biofilm structure formation: A two-dimensional modeling study. Biotechnology and bioengineering. 2000;69(5):504-15. 58. Zhou J, Wu Y, Liu D, Lv R. The effect of carbon source and temperature on the formation and growth of Bacillus licheniformis and Bacillus cereus biofilms. LWT. 2023;186:115239. 69 59. Morimatsu K, Eguchi K, Hamanaka D, Tanaka F, Uchino T. Effects of Temperature and Nutrient Conditions on Biofilm Formation of Pseudomonas putida. Food Science and Technology Research. 2012;18(6):879-83. 60. Josiane N, Heitz M. The influence of the gas flow rate during methane biofiltration on an inorganic packing material. The Canadian Journal of Chemical Engineering. 2009;87:136-42. 61. Kim S, Deshusses M. Determination of mass transfer coefficients for packing materials used in biofilters and biotrickling filters for air pollution control. 1. Experimental results. Chemical Engineering Science. 2008;63:841-55. 62. Sander R. Compilation of Henry's law constants (version 4.0) for water as solvent. Atmospheric Chemistry and Physics. 2015;15(8):4399-981. 63. Oliver JP, Schilling JS. Harnessing fungi to mitigate CH 4 in natural and engineered systems. Applied microbiology and biotechnology. 2018;102:7365-75. 64. Sun MT, Zhao YZ, Yang ZM, Shi XS, Wang L, Dai M, et al. Methane Elimination Using Biofiltration Packed With Fly Ash Ceramsite as Support Material. Front Bioeng Biotechnol. 2020;8:351. 65. Limbri H, Gunawan C, Thomas T, Smith A, Scott J, Rosche B. Coal-packed methane biofilter for mitigation of green house gas emissions from coal mine ventilation air. PLoS One. 2014;9(4):e94641. 66. Pachaiappan R, Cornejo-Ponce L, Rajendran R, Manavalan K, Femilaa Rajan V, Awad F. A review on biofiltration techniques: recent advancements in the removal of volatile organic compounds and heavy metals in the treatment of polluted water. Bioengineered. 2022;13(4):8432-77. 67. López JC, Quijano G, Souza TS, Estrada JM, Lebrero R, Muñoz R. Biotechnologies for greenhouse gases (CH 4, N 2 O, and CO 2) abatement: state of the art and challenges. Applied microbiology and biotechnology. 2013;97:2277-303. 68. Sly L, Bryant L, Cox J, Anderson J. Development of a biofilter for the removal of methane from coal mine ventilation atmospheres. Applied Microbiology and Biotechnology. 1993;39:400-4. 69. du Plessis CA, Strauss JM, Sebapalo EM, Riedel K-HJ. Empirical model for methane oxidation using a composted pine bark biofilter☆. Fuel. 2003;82(11):1359-65. 70 70. Cantera S, Banda DK, Ángeles-Torres R, Palmay PG, Lebrero R. Assessing macroscopic and microscopic performance of fungal-based biofilters for methane abatement. Environmental Technology & Innovation. 2024;33:103519. 71. Nikiema J, Brzezinski R, Heitz M. Influence of phosphorus, potassium, and copper on methane biofiltration performance. Canadian Journal of Civil Engineering. 2010;37(2):335-45. 72. El Hachem K, Kang M. Methane and hydrogen sulfide emissions from abandoned, active, and marginally producing oil and gas wells in Ontario, Canada. Science of The Total Environment. 2022;823:153491. 73. Berger J, Fornés L, Ott C, Jager J, Wawra B, Zanke U. Methane oxidation in a landfill cover with capillary barrier. Waste Management. 2005;25(4):369-73. 74. Einola J-K, Sormunen K, Rintala J. Methane oxidation in a boreal climate in an experimental landfill cover composed from mechanically–biologically treated waste. Science of the total environment. 2008;407(1):67-83. 75. Pratt C, Walcroft AS, Tate KR, Ross DJ, Roy R, Reid MH, et al. Biofiltration of methane emissions from a dairy farm effluent pond. Agriculture, ecosystems & environment. 2012;152:33-9. 76. Zhang Q, Taleghani AD, Elsworth D. Underground energy storage using abandoned oil & gas wells assisted by geothermal. Journal of Energy Storage. 2025;110:115317. 77. Kunz K, Tixier M. Temperature surveys in gas producing wells. Transactions of the AIME. 1955;204(01):111-9.