Dynamic modelling and flexibility analysis of a methanol production process for renewable energy storage

Abstract

The energy transition from fossil fuel (predominantly coal and natural gas) resources to renewable alternatives is driven primarily by the need for cleaner energy to alleviate detrimental global warming effect and ensure sustainable socio-economic development. Candidate solutions features unsteady renewable energy sources such as wind and solar. To make use of the traditional steady state grid infrastructure and serve the energy demand timely, the integration of intermittent energy sources requires the development of energy storage solutions. In this caveat, technologies that can store energy from the time-scales of seconds to seasons to reduce supply and demand mismatches are required. Chemical energy storage is by far the most promising solution due to its unmatched energy density and ability to couple various sectors. Chemical energy storage solution that have ease of handling and will optimise the utilisation of CO2 without putting a burden on the planetary boundaries' constraints are sought after. Methanol is among the most promising solutions as a chemical energy storage and a green-hydrogen carrier with various market application, growing demand, good economic value, and handling and transportation ease in contrast to pure hydrogen. However, to be accepted as an effective solution, several questions regarding the dynamic performance of the methanol synthesis infrastructure under intermittent resource supply need to be sufficiently answered to alleviate technical feasibility concerns such as its ability to handle variable load. From a process system engineering perspective, this includes developing improved kinetic models, advance control, reactor configurations and heat management strategies to optimise and ascertain the potential.

To address these concerns, the work presented in this thesis, has squarely and majorly focused on process system engineering front with dynamic modelling of the power-to-methanol process. To demonstrate this, first, a detailed review study is conducted on the power-to-methanol processes from system perspective with details on the status of the encompassing technologies such as the electrolysis technologies, catalysts, and methanol reactors and process integration, mathematical models and optimization thereof, and the overall techno-economics status of the process. Secondly, an improved kinetic model has been developed for methanol synthesis from CO2 hydrogenation. Detailed comparative steady state search of the candidates and best performing process configurations for power-to-methanol and dynamic modelling thereof are performed. Dynamic load flexibility and load dependent energy efficiency are investigated for steam- and co-electrolysis-based processes. A comparison with electrified reverse water gas shift reactor-based (eRWGS) process is also performed. From this, techeconomics evaluation of the retrofitted power-to-methanol process with consideration of gridconnected energy arbitrage vs standalone renewable energy is conducted. Generalisable effect trends of the key economic feasibility bearing parameters on the levelized cost of methanol and hydrogen are discussed. The work further extends to investigate the dynamic flexibility andstart-up and shut-down physical-chemical characteristic of the water-cooled, gas-cooled, and adiabatic fixed bed tubular methanol reactor under warm and cold start mode. This features, steady state simulation, dynamic simulation with controller tuning optimizations for the different systems.

The results show that the developed kinetic model performed better than other available different literature kinetic models with different assumptions on active sites, rate-determining steps, and hence, model formulations. The modified model (MOD) proposed in this thesis has reduced number of parameters-it excludes CO hydrogenation, but it takes into consideration the morphological changes of active sites and CO adsorption. Furthermore, the integration of methanol synthesis with co-electrolysis yielded higher energy efficiency than the system with steam electrolysis and eRWGS. Steam electrolysis has higher energy efficiency than the eRWGS. From comparing the different flexibility performance of the modular reactors designed in parallel-series and series-series architectures for a 100kton/annum methanol synthesis plant, the parallel-series and series-series have a comparable load flexibility range and limited by the auxiliaries such as compressors and pumps. The parallel-series can adapt to disturbances faster and reach the new steady state in a shorter time. While the series-series configuration performed better in most parameters including the higher load dependent energy efficiency (up 87%) and it is preferred when the maximum load flowrate is less than double its rated flow.

Furthermore, the thesis findings indicate that it is possible with effective controller design and tuning optimization to implement fast ramp rates for load change (from maximum to minimum or vice versa, i.e., ramp rates of 2.22 %/min), start-up and shutdown. The ramp rates are limited by the system pressure fluctuations, available pressure drop, and allowable catalyst heating rates. Fast load ramp rates, particularly during start-up, reduces the CO2 emissions and valuable material losses. The load flexibility of the methanol synthesis using a single adiabatic and water-cooled reactors are comparable (20% to 110%) and limited by the auxiliaries such as compressors and pumps for their physical protection. While the load change flexibility of the gas-cooled reactor is limited to a range of 40 -110%. The maximum load change is limited by the increase in the system pressure drop. The trend of energy efficiency follows the pressure drop very closely, such that the energy efficiency changes in the trend of adiabatic reactor (77.4%) > gas-cooled reactor (75.65%) > water-cooled reactor (73%). The full load reactor CO2 conversions are 18%, 25.4% and 28.3 % for the adiabatic, water- and gas-cooled reactors, respectively. Reactor conversion increases at part load due to reduction in the residence time inside the reactor. Product recirculation is effective to manage the heat generation in the adiabatic reactor, while the use of recirculation and steam drum pressure variation for watercooled reactor, and recirculation and portioning some reactants to the shell side of the gascooled reactor are effective to manage the mildly exothermic methanol synthesis reaction heat generation. No wrong way behavior and thermal run away, and violation of the path constraints were observed when effective control was implemented. It was demonstrated in the thesis that the methanol synthesis can handle dynamic load in the range comparable to other Power-to-X energy storage technologies such as power-to-methane and different ramp rates/degree of intermittency, and can have fast start up (in minutes scales, down to 15 min) and shutdown time (in minutes scales, down to 6min) depending on the reactor design and operation.

Moreover, from a techno-economic perspective of power-to-hydrogen-to-methanol, the results show that economic parity of H2 can be reached with an electricity price of 30 €/MWh and 70% of the CAPEX. While the levelized cost of methanol will still be above 2 €/kg at 80% of the CAPEX and electricity price of 20 €/MWh. This indicates that even if the CAPEX reduces to 20% of its original in this study, and the electricity price reduces to about 20 €/MWh, the green methanol will still not reach economic parity. Plants that still have 20 years and beyond of lifetime left are recommended for retrofitting, below the 20 years, the investment will be difficult to justify, and demand significant capital and electricity price reduction. The results show that to make the retrofitted plant, with minimum of 20 years of lifespan, profitable, a feasible reduction in the electricity price to below 10 €/MWh along with favourable incentives such as CO2 credit and reduction in CAPEX, particularly the electrolyser, and treatment of the PtMeOH as a multiproduct plant with capabilities to co-sell excess products such as electricity/H2/O2 will be required.

The primary practical use of the findings from this thesis relates to the design of the system to be integrated with variable energy for flexible operation and offers guide to ensure technoeconomic feasibility. The thesis contain five core chapters with interconnected findings; three of the chapters have already been published as original full-length peer-reviewed articles and the last two chapters are framed as manuscripts to be peer-reviewed before publication.