GlobalNuclearNewsGLOBAL NUCLEAR INTELLIGENCE
← Nuclear Theory

Society, Environment & the Future · 284

Stairs or Elevator? — Why Temperature Matters in Nuclear Hydrogen Production

The total energy to split water stays the same; higher temperature shifts the burden from expensive electricity to reactor heat.

A white marble mountain model with a pool of water at its base, a glowing golden glass elevator rising halfway up the rock face, and a staircase continuing to a glass platform at the summit
Splitting water is one summit with two ways up: stairs (electricity) and an elevator (heat). Conceptual illustration.
Key idea

Splitting water takes about 39.4 kWh per kilogram of hydrogen, and that total barely changes with temperature. What changes is the split: the share that must be electricity falls from 32.7 kWh at 25°C to 26.0 kWh at 800°C, and heat covers the rest. Because a reactor makes heat first and electricity only at about 33% efficiency, hotter reactors can make hydrogen more efficiently. [1]

One summit, two ways up

Split nine one-liter bottles of water and you get about 1 kilogram of hydrogen, enough for a hydrogen car to drive a little over 100 kilometers. Water is everywhere; the problem is the effort. Water is very stable, and pulling it apart takes about 39.4 kWh per kilogram of hydrogen.

Think of it as climbing a mountain. The summit is 39.4 kWh high, and there are two ways up: stairs you climb on foot, and an elevator that carries you partway. The stairs are electricity; the elevator is heat.

ΔH = ΔG + TΔS

ΔH is the total energy (the height of the mountain), ΔG is the minimum share that must be supplied as work such as electricity even with no losses (the stairs), and TΔS is the share that can be supplied as heat (the elevator). Because temperature T multiplies the last term, the hotter the mountain, the higher the elevator goes. [1]

Graph of energy per kilogram of hydrogen against water or steam temperature from 0 to 1000 degrees: the blue minimum-electricity band falls from 32.7 at 25 degrees to 26.0 at 800 degrees, the orange heat band sits on top, a dashed line marks the total of 39.4, and a side card lists real electricity use of about 50 to 55 for low-temperature electrolysis and 37.7 for a vendor hot-steam test
Calculated from standard thermochemical tables at 1 atm under reversible conditions, in kWh per kg of hydrogen. The 5.6 kWh heat of evaporation at 100°C is supplied separately as heat.
ConditionMinimum electricity (ΔG)Heat share (TΔS)Reaction total (ΔH)
Liquid water, 25°C32.7 kWh/kg6.7 kWh/kg39.4 kWh/kg
Steam, 800°C26.0 kWh/kg8.2 kWh/kg34.2 kWh/kg (+ about 5.6 of evaporation heat)

Starting with room-temperature water and ending with room-temperature hydrogen and oxygen, the total is 39.4 whichever route you take once all heat in and out is counted. The summit is just as high; only the climb on foot gets shorter. That matters because a reactor's direct product is heat: a typical light water reactor converts heat to electricity at about 33%, so 1 kWh of electricity costs about 3 kWh of heat.

Route one: stairs all the way

Low-temperature electrolysis runs electricity through water at about 60–80°C and works with any reactor. At the Nine Mile Point nuclear plant in New York State, a 1.25 MW electrolyzer began producing about 560 kg of hydrogen a day in March 2023, about 54 kWh per kilogram, in line with the 50–55 kWh commonly quoted in industry. [2]

That is far above the minimum of 32.7 because electrode and electrolyte losses leak away as heat. And since two-thirds of the reactor's heat is already gone by the time the electricity is made, reactor heat to hydrogen comes to roughly 25%. [5]

Route two: switch to the elevator with hot steam

High-temperature steam electrolysis splits steam at 700–850°C with a ceramic electrolyte. The heat shrinks the stairs, and hot electrode reactions lose less. A solid oxide electrolyzer tested at Idaho National Laboratory with steam simulating nuclear plant conditions was reported by its manufacturer at about 37.7 kWh per kilogram. The measurement basis differs from the 50–55 of low-temperature electrolysis, so compare cautiously, but it clearly saves electricity. [3]

A hot reactor can also supply the elevator section, boiling and heating the steam, directly with its own heat. That is why reactors with outlet temperatures of about 750–950°C, such as high-temperature gas-cooled reactors, are often named as partners; linking light water reactor steam with extra heating is also being studied.

Route three: a factory with a different temperature on each floor

Splitting water with heat alone in one step takes temperatures well above 2,000°C. A thermochemical cycle links several reactions in a loop so each runs at a manageable temperature, driven by heat apart from auxiliary power. The leading example is the sulfur–iodine (S-I) cycle.

On the first floor (about 120°C), water, iodine and sulfur dioxide form hydrogen iodide and sulfuric acid. On the second floor (about 400–450°C), hydrogen iodide splits and hydrogen comes out. On the top floor (about 850–900°C), sulfuric acid decomposes and oxygen comes out. Sulfur and iodine return to be reused, so only water is consumed.

A vertical temperature axis from 0 to 1000 degrees marking a light water reactor outlet at about 320 degrees and high-temperature reactor outlets of about 750 to 950 degrees, with bands for low-temperature electrolysis at 60 to 80, high-temperature steam electrolysis at 700 to 850 and the sulfur-iodine cycle from about 120 to 850-900 degrees, next to cards for the three cycle reactions by level
Temperatures are typical ranges; actual process temperatures run lower than reactor outlet temperatures by the heat-exchanger temperature difference. The S-I cycle has run only in test facilities.
LevelTemperatureReaction
1 (Bunsen, exothermic)about 120°CI₂ + SO₂ + 2H₂O → 2HI + H₂SO₄
2about 400–450°C2HI → H₂ + I₂
3about 850–900°CH₂SO₄ → SO₂ + H₂O + ½O₂

A light water reactor at about 320°C can't reach the top floor. In 2019 the Japan Atomic Energy Agency produced about 30 liters of hydrogen per hour for 150 hours straight in an electrically heated test facility, not connected to a reactor; materials that withstand hot sulfuric acid and hydrogen iodide remain a challenge. [4]

Two common mix-ups

"Raising the temperature lowers the energy needed" is only half right: the height of the mountain stays about the same, and what shrinks is the expensive electricity share, taken over by heat. And nuclear hydrogen isn't only for high-temperature reactors: low-temperature electrolysis already runs on today's light water reactors. Either way, reactor coolant never flows into the hydrogen feedstock; only heat crosses through heat exchangers.

How far up are we now?

Low-temperature electrolysis is in demonstration beside nuclear plants, high-temperature steam electrolysis is at test scale of a few megawatts or less under simulated plant conditions, and thermochemical cycles are at laboratory scale. Japan is working toward connecting a hydrogen plant to its High Temperature Engineering Test Reactor (HTTR), with the first goal of establishing technology to couple a reactor and a hydrogen plant safely. [4]

Key takeaways

Whether you take the stairs all the way or ride the elevator partway, the summit is the same. A hotter reactor lets heat carry more of the climb, trimming the electricity share from 32.7 to 26.0 kWh per kilogram between 25°C and 800°C, which is why temperature matters so much in nuclear hydrogen.