THE PLAN FOR SUSTAINABLE
SPACE DEVELOPMENT
- establishing an initial infrastructure on the Moon -

REFERENCES:
1. https://onlinelibrary.wiley.com/doi/10.1002/pip.3542

2. https://www.hoymiles.com/wp-content/uploads/2021/05/Datasheet_HMT-18002250_Global_EN_V202108.pdf

3. https://www.semanticscholar.org/paper/The-compressive-response-of-carbon-fiber-composite-Finnegan-Kooistra/f4ff946af3ab18605ace98bd7336988dc8cfd0f6/figure/0

DISCUSSION & CALCULATIONS:

SOLAR DRAPES
Efficiency
One option for the PV blanket is an expensive three-junction thin film made using a lift-off process (1). This gives an alpha of 2.6 kW/kg (efficiency of 26%) at end of life (after ~20 years of irradiation in space).

A cheaper two-junction thin film made with PVD or CVD techniques might achieve an alpha of 600 W/kg at 20% efficiency for perhaps a tenth the price and better handling characteristics. Let's assume this is selected and gain some confidence that if necessary there are higher-performing options coming in the near term if money is no object.

Size
Assume each blanket is 2 meters wide and separated by 2 meters; each meter of height then generates 547 W and masses 911 grams. A 30-meter tall blanket would generate 16.4 kW and mass 27.34 kg.

Voltage
Further assume that the system uses three-phase microinverters. These collect ~30V DC power from a section of each blanket and convert it on the spot to 480V three-phase (Wye) AC. Using a three-phase system greatly reduces the onboard power storage required, so the boards can use thin film capacitors instead of electrolytic caps. That offers mass savings and greatly increased longevity as well as high efficiency, typically around 98%, which in turn reduces heat loads while spreading them across a much wider area. Higher voltage AC power minimizes resistive losses in cabling (due to drastically lower current) and maximizes the performance of whatever central transformer is used to step down to the settlement's grid standard (probably 48V DC). Lastly, 480v three-phase AC is a common industrial standard and three-phase AC power is ubiquitous in industrial motors and certain high-power applications like iron and aluminum production.

Cabling
Current in a three-phase system equals power / (sqrt(3) * power factor * line voltage). Our 30-meter blanket at 480 V, pf = 0.95 and 16.4 kW generates a peak of 20.8 amps, which can in theory be handled by 14-ga (1.4 mm) wire. If we use a copper-clad steel cable then the mass per cable should be 530 grams each or 2.12 kg per blanket. The steel core provides additional strength and significantly higher max working temperatures while the copper cladding provides low resistance; these cables would support the mass of each blanket through kapton standoff tabs that prevent direct contact.

Microinverters
Industrial three-phase microinverters are available in higher power ratings than typical single-panel microinverters. Assume units produce approximately 1.8 kW each, which means we need nine per blanket. Each unit adds about six meters of insulated wire, about 130 grams. (DC power from the PV cells to the inverters is presumed to be carried through the blanket itself.) Unit masses are roughly 5-6 kg in outdoor-rated enclosures; let's assume that the mass of the bare unit and a radiator panel comes to a total of 5 kg. See reference 2 for an example microinverter, with some differences from this design.(2)

Tracking Motor
The top-of-blanket motor for sun tracking would likely be a servomotor in a sealed nitrogen-filled enclosure to prevent vacuum welding. This would be an assembly of the motor, horizontal support bar, cable attachments and controller with built in sun tracking. Suppose the wiring supports a single 360° rotation, meaning each blanket 'rewinds' a full rotation once a month. Call it a total of 4kg at top of blanket.

Drape Spacing & Design
That puts us at about 80 kg per blanket. Suppose we leave two meters between blankets to reduce self-shading and we space support poles at the 30-meter blanket height apart; that's 15 blankets or 1,200 kg per span. That's also 246 kW per span or ~311 amps, which would require 60 meters of 4/0 cable (~30 kg) per span (30m across and 30 m down, and more than strong enough to support the load of the span). Given the uncertainties involved we suggest using a 30% mass margin, which gives a working figure of 1600 kg per span or a static load of 2600 N in lunar gravity. Supports are likely to be made of carbon fiber composite sandwich core structures (3).

Material performance was observed in the range of 1 to 11 MPa (compressive strength) and density of 0.01 to 0.1 Mg/m³. A conservative assessment might use 1 MPa and 100 kg/m³. Support height would be perhaps 40 meters (including subsurface anchoring length). Load of 2535 N at a load limit of 0.5 MPa (safety factor of 2) gives a cross sectional area of 0.0052 m² and a support mass of 20.8 kg. Very fine stainless steel wire (~0.2 mm dia.) would have about twice enough strength for this; 180 meters of fine wire (two runs up and back plus margin for attachments) would be about 45 grams. Including a motor, controller, wire spool and attachment points, let's just call it 2 kg to extend the support. Total of 22.8 kg per support. Presumably the ends of a row of supports will need additional support to resist lateral loads, so let's assume an extra support's worth of mass is used at each end for a diagonal brace.

A system built from 30 spans (33 supports) weighs in at 48.8 tonnes and produces 7.38 MW. That leaves 30 tonnes for the deployment hardware and cabling from the array to the settlement, plus 20 tonnes of payload margin. (We're using additional margins elsewhere in the estimates, but these numbers overall are very achievable.) That gets us a power-system performance of about 150 W/kg and an all-in mission performance of 92 W/kg with comfortable margins. The highest one could reasonably estimate here is 12 MW, which assumes 80 t is reserved for power hardware and 20t for everything else. We'll conservatively use the low-end 7.38 MW value for productivity numbers.

Other production is quoted daily and does not include 'off' days due to shading or darkness. For high temp processing we expect to use parabolic reflectors, so most of the energy cost of heating is assumed to be thermal rather than electrical.

POWER USAGE
Life Support
Assuming 20 kW per person for a bioregenerative life support system including full food production, separate power storage, 93% sunlit time and 10% shading loss that's 308 people supported.

Electrolysis for Propellant
Assuming a 90% efficient electrolysis method, we need 4.94 kW per hourly kg of water (equates to 17.78 MJ/kg). This setup could process 1,494 kg per hour or 35.85 tonnes per day.

Metals from Regolith
Assuming molten oxide iron reduction at 15.25 MJ/kg, production is 41.8 tonnes per day. Assuming molten salt aluminum reduction at 55.33 MJ/kg, daily production is 11.5 tonnes.