What this research found
India's National Green Hydrogen Mission favours high-irradiance desert sites, which raises the question of whether the heat that comes with the sunshine cancels out the advantage. An hourly year-long simulation of a 1 MW solar array feeding matched alkaline and proton exchange membrane (PEM) electrolysers at Jodhpur, Rajasthan, found that the answer depends on the technology: thermal penalties consume 113.6% of the solar gain for alkaline, erasing it, but only 48.1% for PEM.
- For alkaline electrolysis the thermal offset ratio reaches 113.6%, so heat-driven losses slightly exceed the benefit of Rajasthan's stronger sunlight. For PEM the ratio is 48.1%, leaving roughly half the solar advantage intact.
- Over 20 years PEM returns 2.97 times the net benefit of alkaline — 74,397 kg of additional hydrogen against 25,027 kg, measured against an average-India reference — despite beginning with lower efficiency, 59.6% versus 62.0% on a lower-heating-value basis.
- The divergence traces to operating temperature. Alkaline runs at 70.3°C for an Arrhenius acceleration factor of 1.83, or 83% faster ageing, while PEM runs at 60.3°C for a factor of 1.03, a 15-fold difference in thermal impact.
- The photovoltaic array itself gives up 7.42% of its annual energy to heat relative to standard test conditions, about 170 MWh forgone on a 1 MW system.
- First-year output is nearly identical between the two technologies, 40,050 kg of hydrogen for alkaline against 39,255 kg for PEM, but the ranking inverts over the full horizon: 703,664 kg cumulative for alkaline against 708,189 kg for PEM.
How it was done
A typical meteorological year for Jodhpur, Rajasthan (26.24°N, 73.02°E) was synthesized with the pvlib library using the Ineichen clear-sky model, producing 8,760 hourly values of irradiance, temperature and wind: annual global horizontal irradiance of 2,091 kWh/m², a mean temperature of 26.1°C and a 43.6°C peak, with cloud conditions applied to 15% of hours. That series drove an hourly model of a 1 MW photovoltaic array coupled to matched 1 MW alkaline and PEM electrolysers, with efficiency and stack degradation made temperature-dependent through Arrhenius kinetics using activation energies of 55 kJ/mol for alkaline and 65 kJ/mol for PEM. Output was projected across 20 years and compared against an average-India reference case at 85% irradiance, 3% photovoltaic loss and 60°C operation. The work was published as a 19-page two-column paper with 19 figures and more than 60 references, plus a 9-slide deck.
Data sources
- Synthesized typical meteorological year for Jodhpur, Rajasthan — 8,760 hourly records generated with pvlib's Ineichen clear-sky model
- Garcia-Navarro et al. (2024) on PEM stack degradation and Bonanno et al. (2024) on elevated-temperature PEM operation
- IEA levelised cost of hydrogen maps (2024) and the European Hydrogen Observatory
- Sharma & Sagar (2024) policy analysis of India's National Green Hydrogen Mission
Limitations
The climate input is synthesized rather than measured, which the peer review flags as a source of uncertainty and suggests validating against NASA POWER or PVGIS records for Jodhpur. Degradation is represented by a single activation energy per technology at one site, so it captures neither multiple simultaneous ageing mechanisms nor variation between locations.
Figures from this analysis
How this research was produced
K-Dense Web planned and ran this energy investigation end to end — gathering the sources, carrying out the analysis, producing the figures, and drafting the report. The full session transcript, including every intermediate step, is available to view.
