Comparison of Power-to-X-to-Power technologies for energy storage in 2030

Abstract

The energy transition is advancing rapidly, and the Dutch electricity grid is changing with it. Increasing shares of variable renewable energy sources create mismatches between electricity supply and demand. These mismatches create a need for large-scale energy storage. Existing large-scale energy storage technologies are pumped hydro energy storage and compressed air energy storage, but their storage potential in the Netherlands is limited. The alternative is to store energy in chemical bonds, for example by producing hydrogen, and regenerate the electricity later. This type of storage is called a Power-to-X-to-Power (PtXtP) system. Power-to-X technologies have existed for a long time. This
report evaluates their potential in an energy storage application. First, PtXtP systems are compared to compressed air energy storage (CAES) and pumped hydro energy storage (PHES). The geographic potential of CAES and PHES in the Netherlands is limited for both technologies, proving large scale energy storage is a challenge to which PtXtP. Next, all PtXtP technologies are investigated and compared based on available literature. The three technologies with the most potential (hydrogen, ammonia and methane) are further investigated. This report gives a current and comprehensive overview of data on PtXtP system components, including amongst others their OPEX, CAPEX, efficiency and energy use. This data was used as input for several models of hydrogen, methane and ammonia storage systems, to determine system cost and performance in a dynamic system. Simulations are run with these PtXtP systems as energy storage technologies for a 1GW wind park. The simulations are used to identify main system bottlenecks, investigate the impact of intermittent use on system performance, and evaluate the potential of a PtXtP storage system. The first important bottleneck is the size of the hydrogen buffer required for operation of the Haber-Bosch reactor and Sabatier reactor. It is as large as or larger than the storage capacity in a hydrogen storage system. The second bottleneck is the size of ammonia and hydrogen fuel cells. The required fuel cell power is 425 MW, which is larger than any current or expected fuel cells. Next the simulations were used to investigate the performance of a PtXtP system as energy storage medium in a VRES system. The first important finding is the tradeoff between system flexibility and system sizing. An ammonia system with 33%-100% flexibility can be 60% smaller than a 0%-100% system, while still processing the same annual amount of hydrogen. Intermittent system use increases the levelized cost of storage significantly, in these models by factor 2.2-4, due to the unchanged CAPEX which must be paid for a reduced system output. The first important finding related to the PtXtP system is that the cost and energy consumption of hydrogen transport and storage are relatively small compared to energy conversion steps. The electrolyser proved to be the system component with the highest cost and energy loss. Finally, the added value of the storage system is significant wind park size reduction. The 1GW wind park size could be reduced by 35% when connected to a hydrogen storage system, while still meeting the same demand. In addition, zero grid exchange can only be achieved when implementing a storage system. With lower shares of grid exchange, storage becomes increasingly more valuable. The overall conclusion drawn is that hydrogen or methane systems seem to have the most potential for energy storage purposes. The report also shows energy storage is necessary, and no alternatives to PtXtP are available in the Netherlands. PtXtP will therefore have to play a large role in the future Dutch electricity grid. However, use of PtXtP storage will increase the price of electricity and several technological developments, mostly scale-ups, are necessary before a PtXtP
system is feasible.