MBFR — ~1 MWe Net Demonstrator (v2)

Executive summary

This demonstrator aims to be net-electric positive while staying small enough for fast licensing and rapid iteration. The most defensible point is 2 MJ/shot @ 5 Hz with G≈35, laser wall-plug ≈15%, and auxiliaries ≈ 0.5 MWe. That combination yields about +0.9 MWe net (≈1 MWe-class) while keeping window fluence, panel apertures, and pellet logistics modest.

Net-positive operating points

Net ≥ 0 if and only if:
Ef·f × ( ηth→e − 1/(G·ηlaser) ) ≥ Paux

With G=35, η_laser=0.15, η_th→e=0.33, the bracket is 0.1395. Below are concrete points that clear net.

Recommendation: start at Case C (2 MJ @ 5 Hz) for a clean net-positive claim; qualify 1 Hz physics at 4 MJ (Case A) during bring-up.

Economy of scale & cost of electricity

Using the demonstrator (Case C, ≈0.9 MWe net) as a reference, we can fit your cost points (0.5, 0.9, 2, 4, 10, 20, 50 MWe) with a simple power-law:

C(P) ≈ 352 P−0.48  [US$/MWh], where P is net electrical output in MWe. This reproduces the demo-scale values (≈370 US$/MWh at 0.9 MWe) and can be used to project larger plants. For illustration, taking 1 US$ ≈ ₹80 gives the following indicative levelized costs:

These are order-of-magnitude figures assuming MBFR capital and fixed O&M scale mainly with shielding, driver hardware, and tritium systems, while fuel and pellet-factory costs stay nearly flat per MWh. Under those assumptions, national-grid-scale MBFR stations (≳1 GWe net) drive LCOE well below 1 ₹/kWh, and multi-GW “fusion harbours” or propulsion power-plants for ships and spacecraft push it even lower.

Operating modes (1 Hz vs 5 Hz vs 10 Hz)

• 1 Hz: ideal for early optics/chamber debugging; net needs 4 MJ/shot or better efficiency. Least debris per hour; easiest diagnostics.
• 5 Hz (recommended): balances pellet throughput with clear net. Window/film life is manageable; pumps/cryo don’t dominate auxiliaries.
• 10 Hz: full-rate feasibility and availability testing. Use after debris, symmetry, and injector QA are proven.

4096-brick driver — how the “many small lasers” combine

• Architecture: 4096 DPSSL bricks → 256 final beams (≈16 bricks/beam) arranged over ≈12 panels. Incoherent addition; no phase lock required.
• 2 MJ @ 5 Hz, G=35: total UV per shot ≈ 57 kJ ⇒ ~223 J/beam. With a 3 J/cm² damage threshold (2× margin), the clear aperture per beam is ≈ 9–11 cm.
• Per-brick duty: 57 kJ / 4096 ≈ 14 J/brick/shot; at 5 Hz ⇒ 70 W optical/brick (~470 W wall-plug/brick @ 15%).
• Tripling near the panel: combine at 1064 nm, then convert to 3ω just ahead of the final window to keep UV path short.
• Window survival: large apertures + gas curtain or thin liquid film; cassette swaps restore performance; per-beam power balance ±3%.
• Graceful degradation: install 18 bricks/beam, use 16; controller rebalances on failure; hot-swap brick cassettes.

Single-seed timing & jitter budget

• One master seed at 1064 nm fan-out by stabilized fibers; each brick gated by a Pockels cell with fine delay.
• Measured/trimmed jitter < 50 ps panel-to-panel using pilot pulses and cross-correlation. Way below what pellet flight demands.
• Maser pre-heat: pulsed gyrotron, e.g., 1 MW peak for 20–100 µs (≈ 20–100 J delivered). Average draw is only a few kW at 5–10 Hz.

Dual-chamber handover & buffer sizing

Two chambers alternate so availability stays high. The thermal handover between chambers is on the order of ~60 s; a small molten-salt buffer rides this short ramp so the turbine island sees nearly constant power. Longer 5–10 minute pauses are reserved for diagnostics and maintenance at reduced load, not full-rate buffering.

• Buffer sizing heuristic (for handover): 60 s of full thermal output. For 10 MW(th) demo heat, E = 600 MJ. With FLiBe (c_p≈2.4 kJ/kg-K) and ΔT=100 K → m ≈ 2.5 t (≈1.3 m³). Scale linearly with thermal power.
• Continuous pellet factory: stays at 5–10 Hz and feeds whichever chamber is live.

Pellet tracking & targeting accuracy

• Pellet speed ≈ 300 m/s; 2 m standoff gives 6.7 ms flight. Hitting a 100 µm window needs 0.33 µs timing accuracy—easy vs 50 ps laser jitter.
• Pointing: 100 µm at 2 m → 50 µrad. Fast tip/tilt stages (kHz) can hold this.
• Shot sequence: stereo/ToF → predict → maser (−30…−3 µs) → UV foot → 5 ns spike → veto outside gate.

Meters & acceptance test (for policymakers)

• Revenue-grade meters on: (i) bus export, (ii) laser racks, (iii) all other auxiliaries.
• Public display: 15-min rolling net power = P_bus − P_laser − P_aux.
• Acceptance criterion: sustain ≥ +0.5 MWe net for ≥ 30 minutes at the selected operating point (e.g., Case C).

Figures

Figure 1 — Chamber cross-section: injector, maser, beam panels, alpha capture, blanket stack, HX.

Figure 2 — System block: factory → injector → maser → UV panels → chamber → buffer → double-wall HX → power, with closed tritium loops.

Figure 3 — UV panel arrangement (example: 12 panels totaling 256 beams; each panel carries ≈16–24 beamlets).

Figure 4 — Heat & tritium loops (helium-swept double-wall HX; DTU; pellet factory recirculation).