High-voltage PMAD: reducing cable mass on high-power spacecraft

The claim

A NASA Glenn Research Center presentation at the 2012 Space Power Workshop lays out design considerations for high-power spacecraft electrical systems, arguing that high-voltage power distribution can reduce cable mass and ohmic losses significantly compared to low-voltage architectures.¹ The presentation, delivered by Baez, frames the choice of distribution voltage as the central PMAD (Power Management and Distribution) architecture decision for high-power missions. At high power levels, the conductor mass needed to carry current at low voltage becomes a major fraction of total power system mass, making voltage step-up an attractive lever. The trade is not free: higher voltage introduces arc risk in vacuum and surface flashover conditions, adds converter mass and complexity, and changes fault protection requirements across the harness.

It is important to be honest about the source. The NTRS record for this presentation (document 20150010178) contains metadata and keywords but no abstract, and the full presentation slides are not available as a downloadable technical report in the standard sense.¹ What is documented in the primary record is the title, the author affiliation (NASA GRC), the venue (2012 Space Power Workshop), and the keyword set including "high voltage," "cable mass," "ohmic losses," and "power conversion." The broader engineering reasoning below draws on standard PMAD architecture knowledge that is consistent with the presentation's stated scope, but the specific numerical claims should be understood as general PMAD engineering principles rather than verbatim findings from a full paper.

How it works

The mechanism is straightforward electrical engineering. For a given power level P, the current I flowing through a distribution cable is I = P / V, where V is the distribution voltage. Ohmic losses in the cable scale as I squared times R, so doubling the voltage quarters the resistive loss for the same conductor gauge. Equivalently, to deliver the same power at the same acceptable loss, a higher voltage allows a smaller conductor cross-section, which means less cable mass. Since cable and harness mass scales roughly linearly with conductor cross-sectional area and total run length, and since high-power spacecraft can have tens to hundreds of meters of harness routing, the mass savings from voltage step-up can be substantial at the system level.

The Baez presentation emphasizes minimizing the number of power conversion stages to maximize end-to-end efficiency.¹ Each DC-to-DC conversion step incurs efficiency losses (typically 90 to 98 percent per stage at the bus level, measured at the converter input-to-output boundary), so an architecture that distributes at a voltage close to the source output and the load input avoids stacking conversion penalties. The design tension is that raising distribution voltage to cut cable mass may require additional conversion at the source or load if the native voltages of solar arrays, batteries, and payloads do not match the chosen bus voltage. The presentation frames this as an architecture-level trade, not a component-level optimization.

The counterforces are arc risk and insulation. In vacuum and partial-discharge conditions (such as during ascent venting or on surfaces exposed to plasma), higher voltage increases the probability of sustained arcing across insulation gaps and along cable surfaces. Spacecraft charging in auroral or polar orbits can drive differential potentials that interact with high-voltage buses. Arc mitigation adds mass (thicker insulation, potting, corona shields) and design complexity, partially offsetting the cable mass savings. The net mass benefit is therefore a function of voltage level, orbit environment, and arc protection strategy, not a monotonic function of voltage alone.

Steelman: the best case for high-voltage distribution

The strongest argument for pushing distribution voltage up is that cable mass is one of the few mass line items in a power system that scales with power level in a way that architecture choice can directly attack. Solar array mass scales with area (and thus power), battery mass scales with stored energy, and converter mass scales with processed power. But cable mass is a function of the current you choose to push, and that is a free parameter the architect controls. At power levels of tens to hundreds of kilowatts, which are the regime the Baez presentation targets for "high power spacecraft,"¹ low-voltage distribution (28 V or 50 V heritage buses) would require conductor cross-sections that are impractically large for long harness runs. The mass penalty is not marginal; it is structural.

Furthermore, the efficiency argument compounds. Ohmic loss in cables is dissipated as heat that the thermal subsystem must reject. Reducing distribution losses by a factor of four by doubling voltage not only saves cable mass but also reduces thermal control mass and radiator area, which is itself a major mass driver on high-power spacecraft. The presentation's emphasis on minimizing conversion stages¹ means that a well-chosen high-voltage bus can simultaneously cut cable mass, cut ohmic loss, and cut thermal load, a rare triple win in spacecraft design. The arc risk is real but is a known engineering problem with established mitigation approaches (insulation design, voltage derating in critical phases, arc detection and fast disconnect), not a fundamental physics barrier.

Skeptic: where the argument weakens

The honest skepticism begins with the source itself. The Baez presentation is a workshop talk from 2012 whose full content is not available as a citable technical report; only metadata and keywords are in the NTRS record.¹ This means the specific mass-saving claims, voltage levels, and architecture recommendations that the presentation presumably contains cannot be independently verified from the primary record. We are relying on the title and keyword set to infer the presentation's scope, which is a weaker evidentiary basis than a full paper with quantified trades. Any specific numerical claim about how much mass is saved at what voltage should be treated as general PMAD knowledge, not as a finding documented in this source.

Substantively, the arc risk is not always as manageable as the steelman suggests. The interaction between high-voltage buses and spacecraft charging in plasma environments is a coupled problem that depends on orbit, materials, and geometry, and flight heritage for very high voltage distribution (above 100 V, say) on spacecraft is limited. The International Space Station uses a 160 V primary distribution bus and has experienced sustained arcing and plasma interaction issues that required operational workarounds. This is a bench-to-flight gap: arc behavior in vacuum chamber tests does not fully predict arc behavior in the orbital plasma environment. The converter mass and complexity penalty is also understated in the steelman. High-voltage DC-to-DC converters for space are not off-the-shelf components; they require qualified switchgear, insulation systems, and fault protection that add mass and development cost. The net system mass savings after arc mitigation and converter mass may be smaller than the cable-only calculation suggests.

What it means for power in space

Implication: The non-obvious implication is that PMAD architecture, not just generation technology, sets the mass ceiling for high-power spacecraft. A mission that buys advanced high-efficiency solar arrays or a high-power nuclear reactor but inherits a low-voltage distribution architecture may give back a large fraction of the generation mass savings in cable and thermal mass. The Baez presentation frames this as a system-level trade, meaning the power generation and power distribution design decisions are coupled and should not be made independently.¹

Threat: The genuine threat is arc-driven unreliability. If high-voltage distribution is adopted to save cable mass but the arc mitigation strategy is inadequate for the actual orbital plasma environment, the result is not just mass penalty but potential mission-ending single-event faults. A sustained arc on a primary distribution bus can cascade into power loss across multiple subsystems. The flight heritage gap for high-voltage PMAD above heritage levels means the risk models are partly analytical, not fully empirical.

Opportunity: The opportunity is for standardized high-voltage PMAD bus architectures that are qualified once and reused across missions. If the arc mitigation, converter design, and fault protection for a 120 V or 270 V distribution bus are developed and flight-qualified as a reusable module, the cable mass savings become available to all high-power missions without each one re-deriving the trade. This is a platform-level investment, not a per-mission optimization.

Bottom line

The Baez 2012 Space Power Workshop presentation frames the correct central question for high-power spacecraft PMAD: at what voltage do you distribute, and what mass and risk does that choice buy?¹ The physics is unambiguous that higher voltage reduces cable mass and ohmic losses, and that these savings compound into thermal mass reductions. The engineering is unambiguous that higher voltage introduces arc risk and converter complexity that partially offset those savings. What the primary source does not provide, because only its metadata is available, is the quantified trade: at what voltage level, for what power level, in what orbit, does the net system mass reach its minimum. That answer requires the full presentation or a follow-on technical paper. Calibrated reading: the presentation identifies the right trade and the right framing, but the specific architecture recommendations should be sourced from the full slides or subsequent NASA GRC technical publications before being cited as findings. TRL framing: the voltage-mass-loss physics is at TRL 9 (basic principle, flight-demonstrated on ISS at 160 V), but specific high-voltage PMAD architectures for high-power missions beyond ISS heritage are at TRL 4 to 6 depending on the component, with arc mitigation in relevant environments at the lower end of that range.