Synthesis and Assessment of Process Systems for Production of Ammonia Using Nitric Oxide in Combustion Exhaust Gas

Presented on November 14, 2019 during the Ammonia Energy Conference 2019

Authors Hideyuki Matsumoto Shotaro Ogawa Shiro Yoshikawa Tokyo Institute of Technology

Download this Presentation Download a PDF version of this presentation.

Abstract

Recently, ammonia is regarded as an alternative fuel without carbon dioxide (CO2). Numerous studies have been performed using ammonia as a fuel. Iki and Kurata confirmed the working of a prototype for the ammonia gas turbine, where ammonia burned in an environmentally benign way to generate electricity, exhausting only water and nitrogen [1]. From the view of cycle of ammonia for the development of a society with low carbon, it is required to synthesize carbon-free ammonia (green ammonia) in small plants. This green ammonia can be synthesized using renewable energy, with hydrogen from electrolysis of water and nitrogen from pressure swing adsorption of air.

As turning an eye to the cycle of nitrogen that is related to ammonia, nitrogen oxides (NOx) generated in combustion engines are returned to harmless nitrogen molecules by using a large amount of energy. So far, various selective catalytic reduction (SCR) methods have been reported to remove NOx in combustion exhaust gas. For example, in the urea SCR system, ammonia is utilized as a reducing substance. Previous study for the SCR system using hydrocarbon (HC-SCR systems) has reported that ammonia formed in the middle of SCR could accelerate the SCR process.

Since NOx is considered to be the cause of air pollution, it is considered that NOx is much more reactive compared to nitrogen in air and contains reactive nitrogen. Therefore, our research group proposes to utilize the reactive nitrogen in the combustion exhaust gas to produce ammonia as an alternative fuel or a useful chemical. Reuse of the reactive nitrogen without converting it back to nitrogen enables to save a large amount of energy for treatment of environmental pollutants. Moreover, it is expected that the proposed process of converting the reactive nitrogen to ammonia could suppress the enormous energy consumption spent on the activation of nitrogen in the air by combining with the reuse of produced ammonia as a fuel. It is also expected that the recovery and utilization of carbon-free reactive nitrogen will bring about a significant effect for reduction of CO2 emissions.

In development of the proposed process for conversion of reactive nitrogen to ammonia (NTA reaction process), a high selectivity is necessary. NOx reduction with hydrogen is one of the reactions to obtain high productivity of ammonia. However, it is difficult to obtain almost complete conversion to ammonia in NO-H2 reaction under the condition for feed of reactant gas with stoichiometric ratio. Nanba et al. have demonstrated that the Pt/TiO2 catalyst showed high selectivity to ammonia under NO-CO-H2O reaction at ambient pressure [2].

2NO + 5CO + 3H2O → 2NH3 + 5CO2 (1)

Recently, Kobayashi and Nanba have investigated ammonia formation in NO-CO-H2O reaction on TiO2-supported platinum group metals [3]. It was shown that Pt/TiO2 was higher activity than M/TiO2 (M: Rh, Ir, Ru, Pd). Effect of the crystal type of TiO2 on catalytic activity was also investigated. In our experimental analysis, the anatase TiO2 showed higher activity than the rutile TiO2 for Pt/TiO2 catalysts. It was observed by the diffuse reflectance infrared spectroscopy that carbon monoxide exhibited strong interaction with platinum on the rutile TiO2. Therefore, it was estimated that the strong interaction of carbon monoxide with platinum reduced the reaction activity by comparing with Pt/anatase-TiO2 catalyst.

Since the Pt/TiO2 developed by our research group could perform high activity under the temperature condition of 300 ° C or less, the use of waste heat is expected to save energy in the proposed NO-CO-H2O reaction process. Then choice of resource of carbon monoxide and its supply amount are considered to influence the reduction effect of the greenhouse gas (GHG) emissions. However, design and assessment of the process system that is composed of the NO-CO-H2O reaction process and other unit operation processes (e.g. the reduction of hydrocarbon, the adsorption of nitric oxide in combustion exhaust gas) has not been investigated sufficiently.

In the present study, we investigate synthesis of two process functions that are NO-CO-H2O reaction and reduction of hydrocarbon, as the first step of the conceptual design of process system for production of ammonia using nitric oxide in combustion exhaust gas. The framework for procedure of the conceptual design is composed of four activities; “investigation and data collection”, “generation of alternatives”, “simulation” and “evaluation”. In the process synthesis for this case, information about the type of combustor and its scale is considered to be a control factor for the activity of generation of alternatives. When the reciprocating engine generator was set as a type of combustor in this study, reduction of hydrocarbon could be caused by combustion under conditions of low excess air ratio. Therefore the lean-rich cycling operation [4] in combustor was considered to be an alternative to the reduction reactor with addition of a small amount of hydrocarbon.

Effects of adaptation of the lean-rich cycling operation were investigated by using numerical simulation in the present study. In order to acquire composition data for the exhaust gas of combustion engine that was fed to the NO-CO-H2O reaction process, a tool of zero-dimensional simulation based on the extended Zel’dovich mechanism for NO formation and six equilibrium reactions for combustion was developed. Numerical data of composition of the exhaust gas were collected by changing the excess air ratio (λ) for a case when decane was used as a fuel. Temperature and pressure that were input to the combustor simulation tool were set at 300 K and 1.5 atm, respectively. Then behavior of reactor for conversion of NO to NH3 (NTA reactor) was estimated by using Gibbs reactor model in Aspen HYSYS V9. In simulation of the equilibrium reactor, NO-CO-H2O reaction (Eq. 1) and NO-H2 reaction (Eq.2) were applied.

2NO + 5H2 → 2NH3 + 2H2O (2)

Influence of the lean-rich cycling operation to the reduction effect of the GHG emissions was investigated by coupling the combustor simulation on to the NTA reactor simulation. In the simulation of rich burn combustion, value of the λ was changed to 0.797, 0.661 and 0.526. For lean burn condition, value of the λ was assumed to be 1.41, and the NO concentration was changed from 500 to 2000 ppm by considering actual behavior of diesel engine generator. Ratio of the cycle time (rich-burn operation time/ lean-burn operation time) was calculated so that component ratio in feed to the NTA reactor was the stoichiometric ratio for NO and (CO + H2), i.e. 2 : 5. It was assumed that the NTA reactor was operated under adiabatic condition. For a case when NO concentration in the feed gas was 2000 ppm and the temperature of the feed was 250 ° C, it was seen that the ratio of the cycle time for rich-burn / lean-burn operation changed in the range between 1.4% and 4.9%.

Then, the above-mentioned simulation results were compared with the urea SCR system under lean burn conditions (λ = 1.41), by using an evaluation index of ratio of CO2 emissions [mol] to the total heat of combustion for the consumed decane and the produced ammonia [kJ]. As a result for the evaluation, the reduction rate for the value of evaluation index was estimated to be approximately 1% by comparing with the urea SCR system. It was considered that reason why the reduction rate was not large depended on the low heat of combustion for ammonia. The decrease in the reduction rate was seen, as value of the λ decreased. On the other hand, such the change in the reduction rate was not clearly seen, when the NTA reactor was assumed to be operated under isothermal condition.

Hence, it was estimated that positive emission of NOx could be an innovative process system for reduction of GHG emissions by combing with NTA reaction. Since the rich-burn operation time is much shorter than the lean-burn operation time, it necessary to asses effects of the lean-rich cycling operation from the view of controllability, by utilizing dynamic model for the NO adsorption process and the NTA reaction process in the activity of “simulation” in future work. Then it is expected that heat integration among the NTA reaction process and other unit operation processes could bring about more reduction of GHG emissions.