Dynamic behavior of a flexible small­scale ammonia synthesis process

Abstract

Due to climate change, the growth in renewable energy is still accelerating and reaching new records each year. Not everything can be electrified, hence, the production of green hydrogen as an energy vector is gaining momentum. However, lowest cost hydrogen will not always be produced at the demand centres, necessitating transport of hydrogen which is nascent. In this, it has been realised that due to its easy liquefaction, proven transport at scales and matured production process, ammonia can be an excellent energy carrier. Currently, ammonia production accounts for more emissions than any other chemical and to achieve net-zero targets these emissions must be reduced ∼ 25-fold. Therefore,
ammonia synthesis with renewable energy will be imperative. Operating with variable hydrogen feed rates will be part of the challenge.

In this work, the dynamics of a small-scale ammonia synthesis plant with a capacity of 60 𝑡𝑜𝑛𝑠/𝑑𝑎𝑦 was studied. The application of chemisorption as replacement for the traditional condensation step for ammonia recovery was adopted in the study. In the system, the hydrogen feed was sourced from an
electrolyser powered by renewable energy. Hence, it was imperative to study the sensitivity and the response of the synthesis loop to the fluctuations in hydrogen feed. For the dynamics model, different scenarios of ramp-up and ramp-down were tested for two control structures to evaluate the response of various system parameters to these deviations. Therefore, the feasibility of these control structures was assessed and some of their potential limitations were identified.

The objective of this work was achieved with three simulation models. A steady-state model was employed to study the equilibrium conditions of the ammonia synthesis loop. A second more extensive model included reaction kinetics, heat integration and a chemisorption operation. The widely applied Temkin-Pyzhev rate equation was modified into LHHW form to fit this application. With the selected ZA-5 iron-based catalyst the operating pressure was established at 130 𝑏𝑎𝑟 to avoid overheating the catalyst. The simulation resulted in the reactor dimensions to attain the maximum possible conversion of 22.1% under these conditions. For the third model, the inputs from the two models were translated
into the a dynamic model. The latter was used to study the behavior of the system parameters when subjected to fluctuations in the hydrogen feed flow.

Three scenarios were tested with respect to the hydrogen variations: a (step and linear) reduction in the hydrogen feed of 10%, 25% and 50% of the initial value. With the default control system, the pressure varied by 16 bar with only a 10% step reduction of the hydrogen feed. This necessitated the development of control strategies to control the pressure deviations in order to eliminate metal
fatigue in the equipment. The first control structure compensated the lack of hydrogen by adding more nitrogen into the system. The second control structure reduced the recycle flow rate to decrease the ammonia production rate in the reactor.

Both control philosophies were successfully applied to control the system pressure for the 10% and 25% H2 ramp-down/up scenarios. The greatest pressure range measured during transient state in a linear 25% H2 reduction scenario for control philosophy 1 and 2 were 3.8% and 4.5%, respectively.
The effects of the variations and the control strategies have also been studied. In this research the nitrogen supply was assumed to be infinite, therefore a nitrogen buffer must be present. In addition, the use of a hydrogen buffer is required for a gradual decrease and dampens rapid changes in the hydrogen supply in order to minimize pressure variations.