Program · readiness

Readiness, risks & what would close them.

Where the program actually stands: technology and manufacturing readiness level by level, the open risks stated plainly, and what evidence would close each one.

Readiness · Breeder · Hyperion

Readiness · Technology & Manufacturing

An honest maturity assessment — Technology Readiness (TRL) and Manufacturing Readiness (MRL) by subsystem, the long-lead / critical-path items, and the supply-chain constraints. Fusion is early-stage: most subsystems sit at TRL 3–6, not 9. Saying so is the credible position.

⚠ Assessment, not a committed plan. TRL/MRL are self-assessed against the standard scales (TRL 1–9 / MRL 1–10). The two binding constraints are named plainly: REBCO tape scale-up (breeder & burner magnets) and He-3 supply (burner fuel).

Avg TRL

4.4/9

early-stage, honestly

Lowest-TRL subsystem

Breeding blanket

TRL 3 · Tritium breeding at scale unproven

Critical-path items

REBCO tape

Binding supply

REBCO

HTS capacity

Readiness by subsystem

Subsystem	TRL	MRL	Key risk
HTS magnets (REBCO)	TRL 5	MRL 4	Tape production scale-up	critical path
Plasma / ST confinement	TRL 4	MRL 3	Integrated burning plasma unproven
Breeding blanket	TRL 3	MRL 2	Tritium breeding at scale unproven
Divertor (90% radiated)	TRL 4	MRL 3	Detachment control margin	binding
Tritium plant	TRL 4	MRL 3	Throughput + inventory
Vacuum vessel / structure	TRL 6	MRL 5	Large low-activation forgings
Remote handling	TRL 5	MRL 4	ITER-heritage, extend to swaps

Long-lead & critical-path items

Item	Quantity	Lead time	Supply note	Tier
REBCO tape	7.99 t (peak stage, H72)	24–36 mo	global HTS capacity constraint	CRITICAL
SC magnet fabrication	16 TF + CS + PF	~24 mo	winding + cryo test	HIGH
Tritium startup inventory	6.47 kg	—	scarce (fission by-product)	HIGH
Vacuum-vessel forgings	1 set	18–24 mo	large-scale forging	MED
Beryllium / lithium	blanket	—	multiplier + breeder supply	MED

Supply-chain constraints

REBCO / HTS tape

constrained

global capacity ≪ fleet need

Tritium

watch

fission by-product · scarce

Tungsten

watch

China-dominated

Li / Be

manageable

Readiness roadmap now → prototype → product

The binding constraints

Binding constraint #1REBCO / HTS tape scale-up

Binding constraint #2Tritium startup inventory

Overall system TRL4.4 — component demo before ~2030

Manufacturing advantageone isolatable breeder

These maturity levels gate the Roadmap ↗ — the ~2030 prototype needs TRL 6–7, the ~2036 product needs TRL 8–9. The gap is honest and is the work.

Self-assessed TRL (1–9) / MRL (1–10) against the standard scales · REBCO scale-up and tritium inventory are the binding constraints. MATURITY ASSESSMENT · EARLY-STAGE

Radical honesty

Open risks & what would close them

The binding uncertainties in both machines — stated plainly, with severity and the evidence that would retire each one. Most fusion pitches hide this list. Ours leads with it: if you are vetting Kronos, start here.

For DOE / national labsEach item names the experiment or high-fidelity calculation that closes it — several map cleanly onto existing lab capabilities and user-facility beam time.

For investorsThese are the technical gates between here and a fundable prototype. None is a known show-stopper; each has a defined path and is tracked to a readiness gate.

HIGH RISKBREEDERConfinement quality (H₉₈)

Breeder performance leans on the H₉₈ confinement factor — the single largest driver of Q in the uncertainty study. At H₉₈ = 1.0 the design closes; below it, margins erode.

WHAT CLOSES IT → Dedicated confinement scaling data at spherical-tokamak aspect ratio + higher-fidelity transport modelling (declared HPC port).

HIGH RISKBREEDERDivertor radiated-power fraction

The breeder divertor must radiate ~90% of the exhaust power to survive a ~14.3 GW/m² unmitigated load. Negative triangularity helps (no ELMs), but the detachment requirement is demanding.

WHAT CLOSES IT → Detached / radiative-divertor demonstration at prototypical parameters.

HIGH RISKBURNERPlug-density ratio (net-power gate)

Burner net electricity requires a plug-to-central density ratio of ~16 (≈347× the GDT demonstration). This is the binding requirement behind every net-power figure; at a ratio of 10 it does not close.

WHAT CLOSES IT → A tandem-mirror plug-density experiment reaching the required ratio.

HIGH RISKBURNERHigh-field plug magnet — fatigue

Static, bore-resolved stress at the 26.49–39.74 T plug pair sits under the reinforced allowable (≈87–350 MPa vs ~1091 MPa) — which refutes the naïve field-only scaling. But cyclic fatigue at operating stress is currently infeasible (margin 0): this is a gate, not a solved result.

WHAT CLOSES IT → Bore-resolved winding-pack FEA + a fatigue-qualified reinforcement/preload scheme.

MEDIUM RISKBURNERMagnet neutron lifetime

Even low-neutron, the plug coils need shielding: ~15 cm of tungsten gives only ~0.56 full-power-years before REBCO degradation dominates.

WHAT CLOSES IT → Shield optimisation + advanced-pinning (APC) conductor; confirm with transport (MCNP) runs.

MEDIUM RISKBOTHREBCO under fast-neutron fluence

Long-term REBCO critical-current degradation under fusion-spectrum neutrons is an open lifetime question for both machines.

WHAT CLOSES IT → Irradiation campaign on candidate tape at fusion fluence.

STRATEGIC RISKBURNERHelium-3 supply

The burner runs on ³He. The breeder co-produces it, but grid-scale burner fleets need a larger supply (terrestrial + longer-term lunar).

WHAT CLOSES IT → Staged supply: breeder co-production → tritium-decay ³He → lunar sourcing.

Severity is our own assessment, not a third-party rating. "What closes it" is the evidence we would accept as decisive — see Readiness ↗.

Context

How our approach compares

Honest, side-by-side positioning against the alternatives — at the energy-source level and the fusion-architecture level. Advantages and costs both shown; nothing cherry-picked.

Firm clean power — the options

Source	Fuel / risk	Waste	Siting
Kronos burner (D–³He)	Low-neutron; ³He supply is the dependency	No HLW; low activation	Point-of-use (low-neutron)
Fusion D–T (tokamak)	14 MeV neutrons; tritium self-sufficiency	Activated structure; no HLW	Isolated / shielded
Fission (SMR)	Chain reaction; meltdown path exists	Long-lived actinides / HLW	Regulated exclusion zone
Solar + storage	Intermittent; land + battery scale	Panel / battery end-of-life	Land-limited, non-firm
Gas (CCGT)	Fuel price + carbon	CO₂ (or capture cost)	Flexible, emitting

Breeder architecture — ST vs conventional tokamak

Spherical tokamak (Hyperion)

Compact, high bootstrap fraction potential, strong shaping; here run at negative triangularity to suppress ELMs. Trade: tight central column, neutron load on a small machine.

Conventional tokamak

Mature scaling & databases (ITER/IPB98), larger bore eases the centre-column problem. Trade: large, expensive, ELM control on positive-δ H-mode.

Burner architecture — D–³He tandem mirror vs D–T tokamak

D–³He tandem mirror (burner)

Low neutron fraction (~5.4%), direct energy conversion, linear/simple vacuum, no disruptions. Trade: needs a very-high-field plug and a large plug-density ratio; ³He fuel.

D–T tokamak

Highest reactivity, most-studied path. Trade: 80% of energy in 14 MeV neutrons, disruptions, tritium handling, thermal (steam) conversion.

Comparisons use published characteristics of each option and our frozen design point. "Advantage" and "cost" are inherent trade-offs, not marketing — the burner's low-neutron siting advantage, for example, comes with a helium-3 supply dependency.

From 3-D model → digital twin

Future Capability

What this platform is today, and what each surface becomes as the machine is built and instrumented. We are honest about the line: most of this is a twin-ready model; the live twin activates only when there is hardware feeding it.

The three stages — model → shadow → twin

A "digital twin" is a specific thing: a live, two-way link between a real machine and its model. You can't have one without the machine. Here is the honest path, and exactly what separates each stage — the direction the data flows.

Stage 1YOU ARE HERE

Digital Model

data: no live data · —

A physics-grounded 3-D representation of a machine that does not exist yet. It runs the real design calculations and every number reproduces from open data — but nothing is measured, because there is no hardware.

Live now: geometry, the in-browser solver, analysis, V&V, safety, and the whole assurance stack.

Stage 2NEEDS A PROTOTYPE

Digital Shadow

data: one-way · machine → model

Once a prototype is built and instrumented, real sensor data flows ONE-WAY into the model. It now mirrors the actual machine — you compare as-built vs as-designed and watch it run — but the model does not yet act on the machine.

Unlocks with: a built, instrumented prototype + a data pipeline. The Live-Ops, metrology and anomaly surfaces are the scaffold, running on simulated data today.

Stage 3NEEDS THE OPERATING PLANT

Digital Twin

data: two-way (closed loop) · machine ⇄ model

A live, two-way sync. The model ingests plant data AND feeds predictions, optimisation and control back to operations in real time. It forecasts, catches anomalies before they matter, and informs decisions continuously.

Unlocks with: an operating machine + models validated against real data + control integration + safety qualification.

◈ Model— no data —→◈ Shadowmachine → model→◈ Twinmachine ⇄ modelThe only thing that changes across the three stages is whether — and which way — real data flows.

What it takes to advance

Hardware

A built prototype, then an operating plant. Everything downstream waits on this — see the Roadmap gates.

Instrumentation

Sensors and diagnostics feeding a reliable data pipeline (for the one-way shadow, a data-diode keeps it safe).

High-fidelity models

Full FEA / CFD / neutron-transport (MCNP) validated against real measurements — the declared HPC/Quantum ports come online here.

Integration & safety

Binding the model to plant control (PCS/SIS), plus the qualification and cyber-security a live loop demands.

The maturity ladder

3-D model — Geometry, correct proportions, negative-triangularity plasma

✅ built

Physics-grounded model — Every number frozen from the deposit; the breeder solver genuinely computes

✅ built

Design & assurance environment — Analysis · V&V · metrology · safety · control · RAM · thread — representative where no data

✅ built

Digital shadow — Real sensor data flows IN, one-way — the model mirrors the running machine

⛔ needs the machine

Digital twin — Two-way — model and machine stay in sync; the model predicts and informs control

⛔ needs machine + integration

We are at rungs 1–3. The live-telemetry, anomaly and As-Operating features are scaffold for rungs 4–5 — they show the shape of the twin but run on simulated / frozen data today.

What activates, and when

Surface	Today (model)	With a prototype (shadow)	With the live machine (twin)
3-D reactor	Interactive geometry + As-Simulated overlay	Overlay fed by real FEA/CFD/MCNP fields	Live plant state rendered on the model
Analysis	Frozen + screening multiphysics	Full solver runs replace representative fields	Continuously re-run against as-operating data
Metrology	Receiving structure (synthetic Δ)	As-built CMM / scan data ingested	Live drift vs as-designed
Control	Architecture + SIL targets	Hardware-in-the-loop test	Live PCS/SIS bound to the plant
Live Operations	Simulated telemetry	Prototype telemetry (shadow)	Real-time operations
RAM	Allocated MTBF/MTTR	Early field reliability	Live availability + prognostics
Quantum	Simulator PoC + resource estimate (no crossover yet)	Chemistry / sensing on a maturing QPU	Quantum-accelerated design where it genuinely wins

The distinction is deliberate: an operational digital twin requires a physical asset and live sensor data. Until then this is the model, physics and assurance environment — built, verified, and ready to receive the machine. Tied to the Roadmap ↗ and Compute ports ↗.

KRONOS FUSION ENERGY · Kronos Model Conceptual design and simulation study; no machine has been built