(173k) Recovery of NH3 from Livestock Manure for Renewable, Net-Zero Fuel

The nitrogen (N) fertilizer consumption in the U.S. has quadrupled since 1960. This trend has resulted in two consequences. One is the excess N in the soil, which is prone to run off to waterways, causing eutrophication and groundwater contamination. Another is greenhouse gas (GHG) emissions associated with nitrogen. Agriculture is responsible for 75% of the total U.S. N₂O emissions in 2021. 95% of agricultural N₂O emissions are due to soil management, including fertilizer applications. N₂O emissions are 265 times as potent as CO₂ emissions as GHG emissions, accounting for 58% of the total GHG emissions from agricultural activities in the U.S. Every MT of NH₃ generates 2.6 MT of CO_2eq emissions by the Haber-Bosch process. Currently, the U.S. consumes 14 MMT of NH₃, most of which is used as fertilizers. This corresponds to 36.4 MMT of CO_2eq emissions. In addition, applications of N fertilizers to the soil generate N₂O emissions, which is 5.6 % of the total CO_2eq emissions in the U.S., equivalent to 353 MMT of CO_2eq emissions. Together, the consumption of N fertilizers generates 389 MMT of the life-cycle assessment (LCA)-based CO_2eq emissions in the U.S. Studies show that 30 ~ 50 % of those N₂O emissions from soil management originate from applications of animal manure.

Recently, Green NH₃ or H₂ have been receiving strong attention as net-zero emission energy sources. However, both technologies use water electrolysis, which consumes a large amount of renewable electricity. Green NH₃ will require dedicated electricity with a capacity that would be 4.6 times the current global renewable energy production if the goal of the net-zero emission society is to be achieved by 2050. Further, the recent Oxford Report estimates the price of Green NH₃ to be 3.7 times more expensive than Gray NH₃ based on natural gas due to the high cost of electricity for water electrolysis. However, there is a vast, untapped resource for N in the livestock industry. According to one estimate, there were 6 MMT of N as a nutrient in manure discharged from livestock animals in the U.S. This number is more than half the total N fertilizer consumption in the U.S. and almost 90% of the N fertilizers the U.S. imports every year. This paper presents a novel N recovery process with very low LCA-based CO_2eq emissions and LCA-based energy consumptions. The recovered NH₃ can be used as renewable N-fertilizers, renewable net zero-carbon fuels, or H₂ carriers. H₂ can be produced by reforming NH₃.

The stripping/scrubbing process is a common process for NH₃ removal and recovery at industrial facilities. However, its CAPEX/OPEX still need to be lower for livestock farmers to adopt the technology widely. Also, the N concentration in the recovered solution is low, ~ 1 N% which is not high enough for efficient NH₃ condensation to produce liquid NH₃. It also requires high water consumption. More importantly, the current stripping/scrubbing design has limited mass-transfer interface areas, resulting in slow mass-transfer kinetics. Mass transfer is the key to designing the NH₃ recovery process in a commercially viable way and scale. Mass transfer by stripping/scrubbing has been extensively studied experimentally and theoretically. Still, the mass transfer-based NH₃ recovery is an unexplored field. For example, there has yet to be a university-accepted valid correlation between the liquid-to-gas (L-G) or the gas-to-liquid (G-L) mass-transfer rates and the design/operating parameters of a wastewater treatment plant. We believe that a thorough understanding of the relationships between the mass-transfer rate constants and the controllable design/operating parameters in both NH₃ stripping and absorption will help advance a fundamental scientific and engineering understanding of an L-G and G-L multiphase and multi-component mass transfer coupled with chemical reactions, a common phenomenon in many chemical and biochemical processes. Our system has a multi-component mass transfer; when NH₃ is stripped by air, it is mixed with N₂ and O₂ gases. Hence, the G-L transfer involves a mass transfer of the NH₃ gas mixed with N₂ and O₂ gases. Our mass transfer is coupled with chemical reactions because the process uses chemical reactions to facilitate the NH₃ stripping and absorption, especially when stoichiometric chemicals are used.

Accordingly, we first set out to develop a new mass transfer kinetic model (Model I): a consecutive L-GG-L mass transfer kinetic model which has rarely been reported in the literature: the L-G mass transfer for the NH₃ stripping and the G-L for the NH₃ absorption.

Then, we applied the model to our NH₃ recovery process. The model guided us to adopt column systems for their larger interfacial areas and their lower maintenance cost than those of conventional tower systems. According to Model I, the concentration in the absorption tank is expressed as follows:

(1)

where , , , , refer to the concentration in the absorption tank at the time, t, the absorption recovery yield in %, the stripping removal yield in %, the concentration in the stripping column in the beginning, and the mass transfer rate per unit volume for the absorption, respectively. The subscripts “a” and “s” refer to the absorption and the stripping, respectively. The mass transfer rate per unit volume for the stripping, , is embedded in . Hence, eq 1 combines the L-G and G-L mass transfers for the first time. Based on the model, we devised the ranges of the operation parameters to maximize both L-G and G-L mass transfers. Hence, our technology is a model-based recovery process. The operation parameters include the air flow rate for aeration, the orifice diameter of the air diffusers, the column design, the acid/base concentrations, and others.

In addition, we have developed a theoretical kinetic model (Model II) to evaluate the mass-transfer coefficient, , by applying the transition state theory to the mass transfer resistance at the interface, which is equivalent to . The L-G mass transfer coefficient, , has been widely investigated for a variety of gases and chemicals. Yet, the rationale for the dependence on has been scarce in the literature. As to the G-L coefficient, , very few data are available for NH₃ in the literature. With Model II, we were able to calculate the energy barrier for using the density functional theory. Then, we used Model II to investigate the dependence of on the concentration of either base (the NH₃ stripping) or acid (the NH₃ absorption) and gained a better understanding of how the operating parameters affect at the atomistic level.

Through repeated experiments under different operation parameters, we were able to establish the relationship between the mass transfer rate and the operation parameters. Using the relationship, we found the optimized condition to maximize the L-G or G-L NH₃ mass transfers. We employed chemical reactions to facilitate the NH₃ stripping and absorption. One example is the following reactions:

Ca(OH)₂ + 2 → 2NH₃ + Ca²⁺ + H₂O (2)

2NH₃ + H₂SO₄ → (NH₄)₂SO₄ (3)

The stoichiometric amounts of the chemicals produce a 20 N% solution as the final product, which is high enough for the N condensation. Under the optimum conditions, we achieved the L-G mass transfer rate of 1,274.5 mg L^-1 hr^-1, quadrupling some of the highest reported for the NH₃ stripping in the literature. As to the G-L mass transfer rate of 300 mg L^-1 hr^-1, it is still higher than 250 mg L^-1 hr^-1 in a previous study. Both and were more than 90%. Our estimate shows that the LCA-based energy consumption for the recovery of 1 kg of NH₃ by our process is 6 kWh kg-NH₃^-1 which compares with 11.3 and 13.7 kWh kg-NH₃^-1 for Gray and Green NH₃, respectively. As to the LCA-based CO_2eq emissions, our proposed process using eqs 2 and 3 is estimated to generate 0.74 kg CO_2eq kg-NH₃^-1, which compares with 11 and 1.35 ~3.8 kg CO_2eq kg-NH₃^-1 for Gray and Green NH₃, respectively. Our new process consumes about half the energy and generates less CO_2eq emissions than Gary or Green NH₃. Our cost analysis suggests that the cost of producing 1 MT of NH₃ is $1,258, $2,427, and €2,400 by our process, based on the current price of (NH₄)₂SO₄ and the estimate for the stripping/scrubbing process from literature, respectively. Once the N fertilizer is produced by eq 3, NH₃ can be recovered by conventional evaporation and liquefaction, both of which are routine processes. For the transportation of NH₃, infrastructure, such as storage and pipelines, already exist. As a demonstration, we will show the liquid NH₃ obtained by boiling a (NH₄)₂SO₄ solution and condensing the NH₃ gas under -30 °C.

The long-distance cargo shipping industry is currently pilot-testing NH₃ for bunker fuel, which is expected to be the next generation of fuel. However, the currently high cost of Green NH₃ prevents the industry from widely spreading its application. Our process, if successful, would promote a faster adaptation of renewable NH₃ fuel in this industry and beyond.