(374c) Green Hydrogen from PEM Electrolysis: Uncovering Its Worldwide True Cost and Production Boundaries When Accounting for Intermittency

Even when the concept of electrolytic hydrogen has its origins in the 18th century, it was not until the beginning of the last century that hydrogen started to gain attention from an industrial point of view. From 1920 to the present, several hydrogen hype cycles have come and gone, particularly due to the possibility of using free power coming from renewable sources. The production of hydrogen from renewable and non-renewable sources has been highly regarded due to its potential as an energy carrier and its utilisation in several fields, including heat, power and mobility services. In this regard, electrolytic, or green, hydrogen is expected to play a crucial role in any net-zero transition, with polymer electrolyte membrane (PEM) electrolysis considered one of the most promising technologies in this space. However, the costs of electrolytic hydrogen are still high compared to the BAU production routes (grey hydrogen), even when carbon capture and storage is considered (blue hydrogen). This notwithstanding, there is an abundance of literature presenting bullish estimates of significantly reduced future cost. Given that much of this literature rests on so-called expert opinion, or the potential for cost reduction of individual elements of a PEM process, disregarding the entire system, there is a danger of over-estimating the real scope of cost reduction. Furthermore, the intermittent nature of renewables, which is strongly affected by location, is often disregarded in this kind of analysis, although energy storage has been proven to have a high share of the production costs.

Hence, we present a bottom-up analysis for the large-scale implementation of PEM electrolysis, decomposing the whole system into distinct technology components, and evaluate the potential for cost reduction on this basis following a learning rate approach. We conclude that a likely minimum cost for a mature technology cost is in the range of 980 ± 380 USD/MW, relative to current installed costs of 1700 ± 350 USD/MW.

As the cost of electrolytic hydrogen is greatly affected by the intermittency of renewable electricity, we followed a previous framework aimed at the optimal production of green fuels from air, water and electricity (Ganzer and Mac Dowell, 2020). This model accounts for the fluctuating nature of solar energy throughout utilising location-based capacity factors (Pfenninger and Staffell, 2016; Staffell and Pfenninger, 2016) to estimate the electricity produced in any location. For this study, we divided the Earth's surface into 1140 points based on the Universal Transverse Mercator (UTM) coordinate system. We then modified the existing model to allow the use of hybrid systems, including solar and wind power and worked with 20-year hourly-based capacity factors for each location (see Fig. 1). To calculate the production cost of hydrogen in each location, the installed costs for solar and wind systems, electrolysers and energy storage media, and the economic parameters, were chosen according to each region in the map. This MILP model was implemented and solved using GAMS 32.1.0 with CPLEX 12.10.0 as the solver.

We estimated the current (2019) and the future (2050) production cost of green hydrogen at each map point with this information. The obtained costs leased between 3 and 40 USD/kg H₂, depending on the location, proving the considerable influence intermittency has over those production costs.

These values compare relatively poorly with the costs of blue hydrogen, reported in the range of 2-3.5 USD/kg H₂. However, green hydrogen from PEM electrolysis powered with wind electricity could reduce CO₂ emissions to 1.0 kg CO_2-eq. This is not the case for hydrogen from solar power owing to the panel manufacturing processes, whose emissions reach values of around 5.0 kg CO_2-eq. These emissions surpass those of blue hydrogen from autothermal reforming, i.e., 2.6 kg CO_2-eq, and from steam methane reforming, i.e., 4.0 kg CO_2-eq. The environmental results also exhibit burden-shifting, particularly concerning human health and ecosystem diversity toward resource availability. A crucial factor identified for the large-scale deployment of PEM electrolysis is that of iridium, vital to the production of PEM electrolysis units. While the total amount of iridium available on the Earth would satisfy the scenarios envisioned for the hydrogen economy, this would appear to exhaust proven reserves of this critical raw material. Thus, identifying a viable substitute for iridium, a balance with different technologies, significant technological improvements and regulatory policies to manage this finite resource will be critical as this agenda progresses.