That conclusion is as premature as a toddler modeling thermohydraulics in crayon.
While your math easily penitrates the target you chose, its scope is narrower than a collimated neutron beam. You only modeled a single scenario. BSF’s parameter space is vast, uncharted, and riddled with knobs no one’s dared to twist. To draw a general conclusion at such an early and unexplored stage is unjustified.
The way you portrayed BSF, as a necropolis—where capital goes to die, was humorous. However, your analysis was based on several assumptions detrimental to plant performance. Obviously, if the assumptions are bad enough, the payback timeline can be stretched into geological epochs. Instead of focusing on examples with poor efficiency, let’s focus on how to improve the assumptions.
Red’s Assumptions
1. Laser diode modules: $10,000
2. Number of modules: 100,000
3. Target rate: 1 every 15 seconds
4. Energy per target: 330 MJ
5. Thermal-to-electric conversion efficiency: 25%
#1. Future Cost Reduction: Moore’s Law & Learning Curve
The good news is that laser diode costs are expected to plummet with large-scale production and technological improvements – much like semiconductors and solar panels have in the past. Two conceptual models can be used to project future prices under high-volume manufacturing:
- Moore’s Law Analogy: In the microchip industry, Moore’s Law observed that the number of components (transistors) on a chip doubled roughly every 2 years with minimal cost increase. In practice, this meant cost per function halved with each doubling of technology – an exponential improvement. If a similar trend held for high-power laser diodes, the cost per watt could halve with each doubling of cumulative production. This is an extremely rapid learning rate.
- 20% Reduction per Doubling (Learning Curve): Many energy technologies follow a more modest learning curve. Swanson’s Law for solar PV is a classic example – solar module prices tend to fall ~20% with each doubling of cumulative output. This corresponds to an “80% learning rate.” Every doubling of production cuts cost to 80% of its previous value. For instance, 20 doublings (~1048576×volume) would reduce cost to about (0.8)^{20}≈0.012 (approximately 1.2% of the original cost). In practical terms, if a module costs $100k in early low-volume production, the price would fall to about $1,200.
This analysis spanned a range from very aggressive improvement (Moore-like) to conservative improvement (20% learning rate). Real-world outcomes could fall in between. Notably, semiconductor lasers are semiconductors, so with large demand, they could start behaving more like mass-produced microchips or LEDs in terms of cost curve. Historical data shows laser diode prices have fallen dramatically as power and efficiency improved. In 2015, state-of-the-art pump diodes for high-energy lasers cost on the order of $5/W; a few years ago this had dropped below $1/W, and roadmaps aim for a few cents per watt in the coming decade. This trend results from improvements in diode efficiency (reducing waste heat hardware), higher output per chip, and automated manufacturing.
#2. Laser Power (+) Shared Amongst Multi-chambers
The drivers are the most expensive components in inertial confinement fusion (ICF) power plants. In a multi-chamber ICF plant, several blast chambers can share one driver system through time-multiplexing. If a laser is used, the driver only needs to pump for a few microseconds, while clearing the chamber between shots takes several seconds. Thus, the laser modules spend most of their potential duty cycle idle in a single-chamber setup. By rotating pulses between many chambers, the same diode array can be kept busy nearly continuously, increasing utilization without exceeding thermal or duty-cycle limits. This means the high capital cost of the driver is spread over many more fusion events per second, lowering the effective cost of the laser per unit of power produced.
The same principle applies to turbines and power conversion systems. When multiple chambers fire in sequence, the combined energy flow into the heat exchangers and turbines becomes smoother and more continuous, allowing the use of larger, shared infrastructure instead of duplicating equipment for each chamber. Large turbines are cheaper per watt and more efficient than many smaller ones, and the balance-of-plant (cooling, piping, generators, grid connection) can also be consolidated. Together, these shared components distribute fixed costs across greater output, so the gigawatts of electricity delivered per dollar invested rise substantially, reducing the plant’s overall cost-per-watt of electricity.
#3. Faster Target Delivery
In Red’s senario, one 330 MJ target gets ignited every 15 seconds. When these targets are fused, their heat energy gets deposited near the ignition site, in coolant at the center of the blast chamber. This heat needs to be exchanged (removed from the blast chamber) by circulating coolant through it. The per shot volume of coolant that needs to be exchanged (inlet → outlet) can be calculated:
where:
- energy supplied, E = 3.3E8 J
- temperature swing, \Delta T = 1100-750 => 350 K
- density, \rho = 2000 kg/m^3
- specific heat, C_p = 2414 J/kg\cdot K
So, under Red’s assumptions, 0.2 m^3 of coolant would be flowing through the inlet/outlet every 15 seconds (0.8 m^3/\text{minute}). This is two orders of magnitute slower than typical fission reactors, which have volumetric flow rates ranging from 57 to 95 cubic meters per minute.
#4. Higher Yield Targets
#5. Increased Carnot Efficiency
The theoretical maximum efficiency of any heat engine is given by:
Where, based on molten FLiBe coolant:
- T_{\text{hot}} = (750-1100) temperature of the working fluid (in Kelvin)
- T_{\text{cold}} = 300 temperature of the heat sink (in Kelvin)
- η_{\text{Carnot}} = 60-73%
The following ICF expenses where absent in Red’s analysis, but their inclusion would lower the cost of a power plant based on BSF compared to ICF:
Target Fabrication & Injection: 5-10%
Reaction Chamber & Blanket: 15-25%
But sharing the laser system, and turbines between a large number of multi-chambers would produce the most bang for a given buck.