Converting Neutron Heat into Power

Abstract

Fusion becomes an energy source only after heat crosses several controlled boundaries and emerges as a useful product. The blanket coolant collects neutron energy, an intermediate boundary can limit tritium migration and chemical interaction, a power cycle converts thermal exergy, and heat rejection closes the balance. I compare water, helium, and liquid-metal primary coolants with steam, gas-Brayton, and supercritical-carbon-dioxide cycles. The reference architecture keeps a replaceable primary nuclear heat-transport system separate from the turbine island, reserves thermal storage for pulse smoothing and grid services, and measures every internal electrical load. I treat gross efficiency as secondary to net output, material lifetime, confinement, and maintainability.

Heat is abundant; useful temperature is scarce

The deuterium-tritium reaction releases energy in alpha particles and neutrons. Alpha energy eventually joins plasma exhaust and surface heat. Neutron energy becomes volumetric heat in the blanket and shield. The plant must collect these sources without exceeding material limits.

Thermal efficiency is bounded by source and sink temperatures. The Carnot limit is

$$\eta_C=1-\frac{T_c}{T_h}.$$

For a hot reservoir at $773\ \mathrm{K}$ and sink at $303\ \mathrm{K}$, the ideal limit is about $61\%$. Real cycles achieve less because heat transfer, compression, expansion, pressure drop, moisture, component efficiency, and finite temperature differences create irreversibility.

Raising outlet temperature can improve conversion and industrial usefulness. It can also accelerate creep, corrosion, permeation, and irradiation damage. The blanket material window selects the practical cycle.

Primary coolant is a nuclear design choice

Water has high volumetric heat capacity and mature pumps, exchangers, and steam cycles. High pressure creates stored energy, and water ingress into a vacuum vessel or reactive breeder can produce challenging events. Its operating temperature may limit efficiency.

Helium is chemically inert and transparent to many nuclear processes. It needs high pressure and large compression power because of low density. Leakage is difficult, and large ducts consume blanket space. Its high-temperature potential is attractive if structure and joints permit it.

Liquid metals offer high thermal conductivity and low vapor pressure. Conducting flow through strong magnetic fields experiences magnetohydrodynamic drag. Chemical reactivity, corrosion, activation products, freeze protection, and tritium behavior depend on the specific fluid.

Molten salts could combine useful temperature with lower pressure but need fusion-specific radiation, corrosion, chemistry, and tritium evidence. I retain them as a development branch rather than a default.

Coolant	Heat-transfer advantage	Plant penalty
Pressurized water	High heat capacity and industrial maturity	High pressure, lower temperature, interaction hazards
Helium	Inert and potentially high temperature	Pumping power, large ducts, leakage
Lead-lithium	Breeder and coolant functions can combine	MHD pressure drop, mass, corrosion, tritium extraction
Molten salt	Low pressure and high-temperature potential	Chemistry, freeze, radiation, and qualification gaps

The heat balance has no invisible terms

I write the plant boundary as

$$P_{\mathrm{fusion}}+P_{\mathrm{aux,plasma}} =P_{\mathrm{blanket}}+P_{\mathrm{divertor}}+P_{\mathrm{shield}}+P_{\mathrm{loss}}+\frac{dU}{dt}.$$

Some auxiliary power reaches the coolant as heat and contributes to gross generation, but returning a fraction of purchased electricity does not erase the original internal demand. Electrical accounting remains separate:

$$P_{\mathrm{net}}=P_{\mathrm{turbine}}-P_{\mathrm{heating}}-P_{\mathrm{magnets}}-P_{\mathrm{cryo}}-P_{\mathrm{pumps}}-P_{\mathrm{fuel}}-P_{\mathrm{site}}.$$

Every reported net value states whether transformer, cooling-tower, storage, outage, and startup loads are inside the boundary.

Intermediate loops can protect the turbine island

Blanket coolant can carry tritium, activated corrosion products, and contamination. Sending it directly through a large turbine may spread inventory across difficult equipment. An intermediate loop introduces another heat exchanger and efficiency loss but reduces nuclear and tritium scope.

The exchanger becomes a permeation boundary. Tritium can diffuse through hot metal even when no conventional leak exists. Barriers, purge spaces, pressure arrangement, cleanup, and monitoring limit transfer. Double-wall tubes can reveal a breach before primary and secondary fluids mix, at the cost of heat-transfer resistance and complexity.

The pressure hierarchy should drive leakage toward the region with the stronger cleanup and confinement system. That hierarchy may change during shutdown, so check valves and isolation alone are insufficient. Every state needs analysis.

Power cycles follow temperature and fluid

A water-cooled blanket naturally supports a Rankine steam cycle. High-temperature helium can support a gas turbine directly or transfer heat to steam or supercritical carbon dioxide. Supercritical carbon dioxide offers compact turbomachinery and promising efficiency over selected temperatures, with demanding seals, materials, transient control, and heat exchangers.

The first plant should favor a power block whose uncertainties do not coincide with every nuclear uncertainty. A mature steam cycle may produce lower efficiency but clearer commissioning and supply chains. Advanced cycles can be tested on non-nuclear heat before becoming critical to fusion demonstration.

I therefore make the primary-to-secondary interface modular. The nuclear island can evolve while the energy product remains a controlled thermal stream.

Pulses need thermal mediation

If the reference tokamak operates in pulses, the turbine should not necessarily cycle with the plasma. Repeated thermal ramps consume fatigue life and complicate grid delivery. Thermal storage or auxiliary heat can smooth the source.

For a dwell interval $\Delta t$ and desired constant thermal output $P_o$, required usable storage is approximately

$$E_s=P_o\Delta t.$$

Supplying $1.5\ \mathrm{GW_{th}}$ for a ten-minute dwell requires $0.25\ \mathrm{GWh_{th}}$. Real capacity must include efficiency, temperature range, reserve, and degradation.

Storage can also decouple plasma operation from electricity price and provide reserve or ramping. Its vessels, fluid, pumps, heat exchangers, and losses remain within plant cost and safety boundaries.

Decay heat persists after fusion stops

The plasma terminates quickly when fueling, heating, or confinement ends. Activated structures continue to produce decay heat. Stored heat remains in armor, blanket, vessel, and coolant. Chemical reactions may add heat during faults.

The decay-heat removal system uses passive conduction, natural circulation, independent loops, or combinations chosen from inventory and temperature limits. The safety case does not credit an operating turbine as the only heat sink. Ultimate heat rejection survives loss of offsite power and isolates damaged loops.

Heat rejection defines the site

Every thermal plant rejects more heat than it exports as electricity. For thermal input $P_{th}$ and efficiency $\eta$,

$$P_{\mathrm{reject}}=(1-\eta)P_{th}.$$

A $2\ \mathrm{GW_{th}}$ plant at $40\%$ efficiency rejects about $1.2\ \mathrm{GW}$. Cooling water availability, ambient temperature, drought, plume, ecology, and land therefore influence siting. Dry cooling reduces water use while increasing capital and hot-weather performance penalties.

CollectRemove blanket and exhaust heat

Protect material limits and preserve quantified decay-heat paths.

ConfineSeparate tritium and activated material

Use barriers, pressure hierarchy, cleanup, and leak monitoring.

ConvertMatch the cycle to temperature

Select proven or advanced machinery with honest parasitic loads.

DeliverShape output and reject residual heat

Integrate storage, grid services, water, and environmental constraints.

Commissioning starts without fusion heat

Electric heaters, temporary boilers, and non-nuclear test loops can commission pumps, valves, heat exchangers, chemistry, storage, turbines, controls, and heat rejection before deuterium-tritium operation. This separates ordinary power-plant defects from nuclear commissioning.

ColdPressure, leak, flow, and logic tests HotArtificial heat and cycle commissioning NuclearLow-power neutron heat balance RatedNet-output and transient demonstration

The thermal plant can be built from known industrial ideas, but its inner components become activated and life-limited. Achieving useful output repeatedly depends on replacing those components through a remote maintenance system designed as part of the machine from the first layout drawing.