Satellite Structural Overhead Factor

What is the structural overhead mass multiplier -- the ratio of total satellite mass to the sum of primary payload subsystems (solar arrays, thermal radiators, compute hardware)?

What is the structural overhead mass multiplier for orbital compute satellites?

Answer

The structural overhead multiplier ranges from 1.15x (optimistic) to 1.40x (conservative), with a central estimate of 1.25x. This covers the satellite bus structure, wiring harness, ADCS, propulsion (station-keeping and deorbit), communications, and command & data handling -- everything not in the three primary subsystems.

The values are supported by three convergent lines of evidence: (1) standard spacecraft mass allocation tables adapted for compute-satellite architecture, (2) specific compute-satellite analyses (Mach33/Space Intelligence, Dwarkesh Patel) allocating 15-25% of total mass to bus/chassis, and (3) bottom-up subsystem estimation summing to 20-44% of the pre-overhead payload total, corresponding to multipliers of 1.20x-1.44x.

Analysis

Defining the overhead for a compute satellite

For a conventional satellite, standard mass allocation tables (as reflected in NANOSTAR nanostar-methodology.1) assign 46% of dry mass to non-payload/non-power/non-thermal subsystems: structure 27%, ADCS 6%, propulsion 3%, communications 2%, C&DH 5%, and other 3%. But for a compute satellite, the "payload" is the compute hardware, "power" is the solar arrays, and "thermal" is the radiators — all three are already counted as primary subsystems. The overhead covers only the residual bus, wiring, ADCS, propulsion, comms, and C&DH.

Bottom-up overhead estimation

For a 100 kW compute satellite with pre-overhead mass of ~20 kg/kW_IT (central case: ~2,000 kg for a 100 kW system):

This translates to multipliers of 1.20x-1.44x on the pre-overhead mass, bracketing the model's 1.25x central value.

Cross-check against published estimates

• Mach33/Space Intelligence mach33-cooling.8: 100 kW satellite at ~5,400 kg, with ~3,600 kg in primary subsystems (solar + radiators) and ~1,800 kg in overhead (includes compute). Adjusting for antenna removal from Starlink V3 baseline, overhead is ~1,300 kg on ~3,600 kg, or ~1.36x — between central and conservative. (Note: this denominator omits compute mass because the source bundles compute into the +500 kg net bus increase rather than breaking it out separately. Including compute in the denominator — as this page's definition requires — would lower the ratio to ~1.33-1.35x; the difference is small.)

• Dwarkesh Patel dwarkesh-space-gpus.1: 25% chassis allocation = 1.33x multiplier, between central and conservative.

• Per Aspera peraspera-realities.6: 3-5 metric tons total, excluding batteries (~2.5-4.5 MT), with ~1,700 kg identifiable primary subsystems (solar ~700 kg + radiators ~1,000 kg; compute mass not separately itemized). Overhead spans 1.47-2.65x, though the wide range reflects uncertainty in the source's "structure, cooling loops, computers, and more" residual rather than a precise structural overhead estimate.

Why the range is asymmetric

The optimistic case (1.15x) assumes tight structural integration — compute modules mounted directly on solar array structures (per Handmer's concept), minimal separate bus frame, electric propulsion with low propellant needs in dawn-dusk SSO, and streamlined wiring through distributed architecture. Aggressive but achievable for a purpose-built, mass-optimized design.

The conservative case (1.40x) accounts for a less integrated bus design, conventional wiring harness at 5-8% of mass, more propellant for contingencies and debris avoidance, and redundant ADCS. Battery mass for eclipse ride-through is now modeled separately (see eclipse-duration-sso) and included in the pre-overhead mass sum alongside solar, thermal, and compute subsystems.

The central case (1.25x) represents a well-engineered but not revolutionary bus leveraging Starlink heritage for ADCS, propulsion, and communications, with moderate structural integration. Note that this sits at the lower end of the bottom-up range (1.20-1.44x) and below both cited cross-checks (Mach33 at ~1.36x, Dwarkesh at 1.33x). The 1.25x central is therefore optimistic-leaning rather than a true midpoint; a more conservative central of 1.30-1.33x would be equally defensible. This choice has a modest impact on TCO — increasing structural overhead from 1.25x to 1.33x adds ~2-3% to satellite mass and thus to launch cost, which at central 2035 launch prices changes TCO by <1%.

Evidence

1. NANOSTAR spacecraft design methodology provides mass allocation tables for LEO satellites with propulsion: structure/mechanisms 27%, thermal 2%, ADCS 6%, propulsion 3%, power (including harness) 21%, communications 2%, C&DH 5%, payload 31%, other 3%. — nanostar-methodology

2. Mach33/Space Intelligence analysis of a Starlink V3-derived 100 kW compute satellite models the bus mass as a +500 kg net increase from the 20 kW baseline, capturing antenna removal offset by added compute, power conditioning, and thermal transport. Total 100 kW satellite mass: ~5,400 kg, of which solar arrays ~2,600 kg, radiators ~1,000 kg, with bus/structure/propulsion accounting for the remaining ~1,800 kg. — mach33-cooling

3. Dwarkesh Patel's mass budget assumes "a fourth of the mass of the satellite has to be the chassis," arriving at a whole-system specific power of 85 W/kg. This 25% chassis allocation corresponds to a multiplier of 1.33x on non-chassis components. — dwarkesh-space-gpus

4. Patel explicitly flags the 25% chassis assumption as a rough estimate: "feel free to plug in your own numbers." It is not derived from a detailed subsystem analysis. — dwarkesh-space-gpus

5. McCalip's orbital data center model uses a Starlink V2 Mini heritage bus class with station-keeping propellant mass assumed rolled into Starlink-like specific power (W/kg). No separate structural overhead is broken out; the model scales a single bus class linearly to target power. — mccalip-space-dc

6. Per Aspera estimates a 100 kW orbital data center at 3-5 metric tons total, itemizing solar panels (~700 kg for ~100 kW generation), batteries (a few hundred kg), radiators (~1,000 kg), plus "structure, cooling loops, computers, and more." In a separate power-sizing discussion, the source estimates ~930 kg of solar panels for 140 kW generation (oversized to maintain 100 kW delivery after eclipse and efficiency losses) and ~500 kg of batteries for 50 kWh storage. — peraspera-realities

7. NASA GSFC Reliability Practice PD-ED-1238 (via LLIS lesson 722) documents spacecraft electrical harness design standards: minimum wire gauge AWG 24 (AWG 22 for power), harness diameter limited to 1 inch, EMI mitigation via signal-type grouping and shielding, and rigorous contamination control. The practice focuses on fabrication quality and environmental survivability rather than mass budgets. — nasa-harness-llis