BankableReal Bet? · The Electrolyser Teardown · one-page verdictWhich electrolyzer is the real bet?
There is no single winner. The real bet is the match — machine to power to site. Companion to the video episode; every number below traces to the cited research file.
| Machine | CAPEX $/kW | Flexibility | Stack life | Key material | Verdict |
|---|---|---|---|---|---|
| Alkaline (AWE) OH⁻ · liquid KOH · 60–90 °C |
~$2,000 West $750–1,300 China |
modern 5–100%, ~10%/s; classic floor ~20% (crossover) |
60–90k h | nickel — no PGM | REAL BET today's workhorse — firm, cost-first sites |
| PEM H⁺ · solid acid membrane · 50–80 °C |
~$2,400 West | ~5% min load, sub-second; most compact (1–6 A/cm²) |
60–80k h | iridium + platinum | REAL BET the flexible future — variable power, tight sites |
| Solid oxide (SOEC) O²⁻ · hot ceramic · 600–850 °C |
$2,000–3,000 (pilot) | hates cycling (thermal inertia) | ~20k h demo | ceramics + nickel — no PGM | WATCH 80–90% efficient — needs cheap heat + durability |
| AEM OH⁻ · solid alkaline membrane · 40–60 °C |
early stage | aims PEM-like (0.2 A/cm² today) | ~1,000 h proven | Ni/Co — no PGM | NOT YET highest upside, least proven — track it |
Sources: IEA Global Hydrogen Review 2025 · World Bank electrolyser report · Kiemel et al. IJHE (iridium: ~8 t/yr, ~90% South Africa; loadings 2 g/kW → 0.3 → 0.05) · IJHE 2025 SOEC · Ammonia Energy Assn AEM status · RSER dynamic-operation review · Applied Energy 2024 · S&P Global / Hydrogen Insight (Sinopec Kuqa ~20% utilization). Full [S#] mapping in the workbook's Sources sheet.
Read the site, then choose the configuration
- Power firm (hydro / baseload)? The cheap alkaline machine runs all day and never meets its weak spot — hydro country (Quebec → Washington, Norway → Sichuan) is its home.
- Power bouncy (solar / wind)? Pay for PEM's ~5% floor and sub-second response — and budget the iridium supply question.
- Land or purity tight? PEM: far more power per square metre, near-pure H₂ off the stack.
- Cheap heat next door (nuclear · steel · steam)? Watch SOEC — 80–90% efficiency once stack life clears ~40k hours.
- Water: budget 20–30 L/kg real (9 L stoichiometric); PEM wants ultrapure, chloride-free feed. Seawater = a desalination plant, not just a pipe. Groundwater = water rights + permits.
Bankable — which hard-tech bets actually pencil out. The editable comparison workbook (matrix + site matcher) came with this download; the video is at bankable.show.
Prepared from public data by an engineer with over a decade commercializing first-of-a-kind energy technology. Illustrative, not project-grade engineering. We judge technologies, never stocks; nothing here is investment advice. Advisory: hello@bankable.show