Installed electrolyser CAPEX in 2026 spans $700–$1,300/kW for Chinese alkaline, $1,800–$2,300/kW for Western alkaline, $900–$1,500/kW for Chinese PEM, and $2,000–$2,550/kW for Western PEM, with the spread driven by manufacturer geography, PGM catalyst loading, and stack current density [S2][S3].
The U.S. Department of Energy 2026 PEM technical target remains $250/kW uninstalled (system) and $100/kW (stack), against a 2022 status of $1,000/kW and $450/kW respectively, paired with a $2/kg H2 production cost goal [S1]. Real-world installed costs still run 2–3x above that target, with Western PEM averaging $1,500–$2,000/kW for commercial systems producing 99.999% purity H2 [S8].
CAPEX Bands by Technology and Manufacturing Origin
Chinese alkaline systems from LONGi, Peric, and Sungrow are quoted at $700–$1,300/kW installed in 2026, with several sources reporting sub-$1,000/kW at multi-hundred-MW scale [S2]. The 40–60% cost gap versus Western alkaline (Nel, ITM, Plug Power, Siemens Energy) is attributed to lower labour, diaphragm (Zirfon-type) sourcing, and nickel-electrode supply chain depth [S3]. Western alkaline sits at $1,800–$2,300/kW installed, down roughly 30% from 2022 baseline [S2].
PEM pricing remains 1.5–2x alkaline in every geography. Chinese PEM stacks are reported at $900–$1,500/kW and falling fast, while Western PEM holds $2,000–$2,550/kW with only a 25% reduction since 2022 [S2]. A separate 2026 dataset pegs alkaline at $600–$1,000/kW and PEM at $1,500–$2,000/kW for commercial units, reflecting a narrower gap in some Chinese-dominated segments [S7][S8]. The DOE 2026 stack target of 0.5 mg/cm² PGM (down from 3.0 mg/cm² in 2022) and iridium-specific loading of 0.1 g/kW is the structural lever for further PEM cost compression [S1].
Stack Architecture and the Cost-Per-Spec Breakdown
Alkaline stacks run paired nickel-coated steel electrodes in 25–30% KOH electrolyte at 70–90°C, 1.8–2.4 V per cell, with current density limited to 0.2–0.4 A/cm² and cold-start times of 20–60 minutes [S5][S6]. PEM stacks operate at 50–80°C, 1.0–2.0 A/cm² (with 2–3 A/cm² cited in OEM literature), reaching 3.0 A/cm² at 1.8 V per cell in the 2026 DOE target, with sub-5-minute cold start and millisecond response [S1][S5][S6]. That current density gap (roughly 5–10x higher in PEM) is the physical reason PEM commands a price premium per kW installed: fewer cells, smaller footprint, and higher differential pressure output.
System efficiency on an LHV basis is reported at 63–70% for alkaline and 56–68% for PEM, though DOE 2026 targets 65% (51 kWh/kg) for PEM systems and 69% (48 kWh/kg) at the stack level [S1][S5]. Operating cost in the field tracks at $2.00–$3.50/kg H2 for alkaline versus $3.50–$5.00/kg for PEM according to comparative production cost data [S4]. Stack lifetime diverges sharply: 60,000–90,000 hours for alkaline versus 40,000–60,000 hours for PEM, with the DOE 2026 PEM target set at 80,000 hours and 2.3 mV/kh average degradation [S1][S5].
Decision Matrix: When to Pick Alkaline, When to Pay for PEM
For baseload hydrogen production above ~4,000 hours/year with stable renewable input, Chinese alkaline at $700–$1,000/kW delivers the lowest LCOH, with capital amortisation dominating total cost [S2][S3]. Halving CAPEX from $2,000/kW to $1,000/kW reduces green H2 cost by roughly $1/kg and PtL-SAF cost by $500–$800/tonne in Power-to-Liquid applications, where the electrolyser is 30–40% of total plant CAPEX [S2].
For dynamic-load profiles (wind penetration above ~50% of grid mix, sub-minute ramp requirements, or 0–100% load range duty), PEM's millisecond response and ability to operate at zero load justify the $1,000–$1,500/kW premium over alkaline [S5][S6]. Denmark is the textbook case for PEM flexibility: PEM electrolysers offer response times under one second and a wider operating range that captures surplus variable wind power, while alkaline electrolysers remain the cheaper option for steady, baseload production with lower capital costs and longer operational lifespans [S5]. A related reference on how plant-level demand response translates into industrial cost savings in 2026 is covered in 2026 electricity price volatility and industrial demand response, which frames the same dynamic-load economics from the electricity-buyer's side.
High-purity H2 for fuel-cell mobility or electronics (99.999% from PEM versus lower downstream purification cost on alkaline) further tips the scale for PEM in mobility and small-scale on-site generation, despite the 5–10 year stack lifetime penalty [S4][S8]. For utility-scale ammonia, methanol, or refinery hydrogen, alkaline remains the default.
Project Economics: LCOH and the Path to $2/kg
Reaching the $2/kg H2 threshold requires three conditions firing simultaneously: electrolyser CAPEX below $500/kW, electricity below $20/MWh, and utilisation above 4,000 hours/year [S2]. Chinese alkaline is approaching the CAPEX line; Middle East solar PPAs at $13/MWh and parts of Chile/Australia have hit the electricity line; the utilisation gap is the binding constraint, with most intermittent-renewables plants running 2,000–3,000 hours/year without battery buffering or hybrid solar+wind smoothing [S2].
Regional 2026 H2 cost bands cluster at $2.50–$4.00/kg in the Middle East, $3.00–$5.00/kg in Chile/Australia and the US (with the 45V production tax credit at $3/kg), $3.50–$5.50/kg in China, and $5.00–$7.50/kg in Europe, where both electricity and CAPEX are higher [S2]. The $2/kg DOE target for PEM is contingent on hitting 0.5 mg/cm² PGM loading, 3.0 A/cm² at 1.8 V per cell, 80,000-hour stack life, and $250/kW uninstalled system cost, all simultaneously on the same stack [S1]. None of those four parameters is independently sufficient.
What the 2026 Numbers Don't Tell You
All CAPEX figures above are stack-and-BOP; balance-of-plant civil work, grid connection, water treatment, and hydrogen compression are not included and typically add 30–60% to the installed project cost depending on site [S2]. The "Chinese alkaline at $250–$400/kW" figure cited in some 2026 surveys reflects ex-works stack pricing for bulk shipments, not delivered-and-installed cost in Europe or North America, where shipping, certification (CE/ATEX/IEC 60079 series for hazardous-area integration near pressure transmitter and flow meter skids), and EPC markup push effective CAPEX 2–3x higher [S3].
For engineers building process plant integrations that already specify industrial valve trains, PLC interlocks, and servo-motor-driven pressure sensor manifolds, the relevant question is not the $700/kW or $2,500/kW headline, but the delivered-and-commissioned figure on a specific site with a specific load profile. The 2026 cost gap between alkaline and PEM is structural and will not close below roughly 1.3–1.5x in the next 24 months; bankability, dynamic response, and lifetime remain the three real selection criteria, not headline CAPEX.