Elsevier

Renewable Energy

Volume 187, March 2022, Pages 987-994
Renewable Energy

Uninterrupted photovoltaic power for lunar colonization without the need for storage

https://doi.org/10.1016/j.renene.2022.02.016Get rights and content

Highlights

  • Continuous PV electricity production without storage for lunar colonies

  • High-latitude deployment permits viable transmission to near-polar consumers

  • Specific power (kW/kg) is far superior to nuclear reactors or adding battery storage

  • No atmosphere, a near-zero axis tilt and small size enable a unique lunar solution

Abstract

Can uninterrupted photovoltaic power feasibly be realized without energy storage? Although on planet Earth the answer appears to be negative, we depict and evaluate how it can be achieved on the Moon with a strategy that exploits the combination of the absence of a lunar atmosphere and the near-zero tilt of the Moon's polar axis with respect to the ecliptic plane. This strategy may be of considerable value in the current revival of lunar colonization initiatives, where the envisioned level of uninterrupted electricity for the oxygen production plants that will constitute the principal energy consumer is of order ∼10 MW, with the flexibility of large distances between power production and power consumption. This approach is also shown to offer noticeably superior specific power (kW/kg) than the current proposed designs of either nuclear reactors or solar with battery storage.

Introduction

We start with the central notion and then motivate its uniqueness for electricity generation in future lunar colonies, along with estimating the magnitudes involved and the feasibility relative to alternatives.

A chief limitation of photovoltaic (PV) systems is accommodating the temporal mismatch between energy delivery and consumption, in particular when 100% of year-round power demand must be satisfied. Namely, PV power does not offer a totally renewable solution if all electricity needs must be satisfied at all times, but rather must be enhanced with energy storage that invariably engenders its own non-renewable characteristics, or with supplemental renewable sources such as wind or hydroelectric that are also limited by their ephermeral nature. The limited degree to which renewables can supply close to 100% grid penetration has recently been investigated and quantified over a broad range of climates. Not unexpectedly, there is an acute diminishing-returns relation between grid penetration and the amount of both PV area and storage, well before the 100% level is approached [1,2].

Although battery storage can solve the problem, it requires a massive capacity of the order of weeks, rather than just hours, of storage. The associated exorbitant price has precluded its widespread use in large-scale PV systems on Earth, the power generation of which currently is reaching close to 1 peak TW worldwide and expanding annually at close to 0.2 TW [3]. The individual utility-scale systems comprising and accounting for this global grand total have peak power generation from tens to hundreds of MW. Some have incorporated buffer battery storage, mainly to maintain short-term system power delivery stability, and typically not exceeding a few hours. Their grid penetration, however, remains well below approaching 100% [[4], [5], [6]]. Trying to mitigate the problem by connecting PV systems across many thousands of km - with the potential to remain totally renewable - similarly engenders unacceptably immense cost and infrastructure today and in the forseeable future.

But the situation on the Moon differs in three ways that permit a completely solar solution. (Other renewable resources such as wind and hydroelectric do not exist on the Moon.) First, the absence of an atmosphere means that normal solar irradiance (i.e., at zero incidence angle to the solar beam) is independent of latitude and time of day. Hence different locations on the Moon are distinguished by their latitude (geometry), and not by atmospheric effects that are so dominant on Earth. Second, the near-zero tilt of the Moon's rotational axis with respect to the ecliptic plane (Fig. 1a) permits lower cosine losses than on Earth for judiciously chosen collector orientations. And third, the radius of the Moon is small enough that the distances between power-generation and remote power-demand locations need not exceed the order of 102 km (elaborated in Section 2.1).

The proposal here is to install PV arrays across a 360° latitudinal ring at high latitude, illustrated in Fig. 1b for 87° (but not exceeding ∼88° for reasons quantified in Section 2.1). Conventional transmission lines that can be far lower-mass than those used on Earth (elaborated in Section 3.2) would deliver electricity to remote oxygen (O2) production plants. The collection and transmission of PV power can ensure the uninterrupted electricity demand of each O2 factory 100% of the time. Namely, even when the O2 plants are in the dark, there will always be sufficient PV array area on the opposite irradiated side of the Moon to satisfy the power demand. This point is illustrated schematicaly in Fig. 1b.

The O2 factories are located at a slightly higher latitude than the PV arrays (rather than at lower latitudes) so that the length of transmission lines can be minimal. The number of such factories remains to be determined from future optimization analyses, and is arbitrarily chosen to be four in Fig. 1b for illustrative purposes. These factories are also envisioned as being anywhere from several km to tens of km from the PV systems in order to avoid the PV panels becoming covered with the highly adhesive dust from the lunar soil which will be thrust into the immediate environment of the O2 production plants during processing, harvesting, O2 storage and subsequent transfer to consumers.

This approach will be shown to offer a specific power (kW/kg) that is substantially superior to the current proposed alternatives of either nuclear reactors or solar with battery storage.

At the outset, we stress that our aim is not a precise, detailed blueprint, but rather highlighting the feasibility of a novel, viable concept, and establishing the approximate magnitudes for the key system components. These estimates are based on commercially-available, laboratory-tested technologies only, and hence available for immediate implementation. Without diminishing the potential of future prospective materials and devices, we do not consider them due to their speculative nature. Indeed, any proposal adopting existing technologies, and targeted for projects to be realized within a few years (as opposed to decades into the future), could be delayed based on the hope that alternative breakthroughs are en route. But the strategy portrayed here is geared toward current multi-national plans for Moon colonization within the coming few years. And while economics will not be an insignificant factor, it is premature to relate to the prices of individual devices and their lunar installation.

Moon colonization programs formulated decades ago have recently been revived in earnest by the space agencies of the US, the EU, Russia, China, Japan, and India [8,9]. Every aspect of suitable lunar locations, power systems, energy storage, construction, maintenance, cost, and power management and system control strategies have returned to the agenda [[10], [11], [12]]. Oxygen production from the lunar soil (regolith, comprising metal oxides that are ∼45% O2 [9]) in reactors will constitute the overwhelming majority of the total energy requirement [13,14] (Fig. 2). Essentially all that energy will be electricity that either powers high-temperature (∼1800 K) electrolyzers or is converted to the high-temperature heat (∼1500–2500 K) used to drive regolith decomposition, typically via vacuum pyrolysis [15]. (The proposal of installing a nuclear fission reactor-cum-generator or turbine in lieu of solar power [16] will be addressed at the end of Section 3.) The overwhelming majority of the O2 will serve as propellant for assorted space missions subsuming refueling Earth-orbiting satellites, return flights to Earth, and journeys to Mars and beyond. Part of the attraction is the dramatically smaller launch energy and cost due to the gravitational acceleration on the Moon being only approximately 1/6th of that on Earth.

Annual O2 generation levels from 1000 tons up to several thousand tons are envisioned [17]. Given that the most efficient and feasible reactors developed to date have demonstrated a specific energy consumption of ∼30 kWh/kg of O2 [9,13,18], the necessary annual average electrical energy delivery (including distribution losses) would approach the order of ∼100 GWh. For plants in continuous operation, the delivered solar power would then need to be of order ∼10 MW - a magnitude used to set the scale for the solar installations in the remainder of this paper.

With launch costs from Earth currently exceeding $100,000/kg, maximum specific power (kW/kg) for lunar energy systems is paramount. PV systems constitute a proven space technology, offering a viable, practical, reliable, efficient and long-lifetime option for lunar colonies. One must distinguish, however, between specific power and conversion efficiency. The former determines the mass required for a given PV power demand, whereas the latter determines the necessary PV array area independent of mass.

Commercial PV arrays for space (Fig. 3) have achieved module conversion efficiencies of 33% [19], based on a normal-incidence yearly-average extraterrestrial solar irradiance Isun of 1.37 kW/m2, the AM0 solar spectrum, and the standard PV testing temperature of 298 K. A couple of adjustments must be made for estimating array efficiency. First, the operating PV temperature on the Moon (where the sole means of heat rejection is thermal radiation) will be ∼60 K higher. Based on measured temperature coefficients for the III-V cells, conversion efficiency would then be lower by ∼10% relative [20]. Second, assembling modules into arrays leaves a small area devoid of solar cells, representing another penalty of ∼10% relative. Accordingly, we conservatively use an array-level conversion efficiency of 25%.

The PV systems and electronics on the Moon will be exposed to energetic electrons and protons. Typically, adequate protection is provided by a 0.2 mm thick glazing of high transmittance to solar radiation, and high absorptance to the high-energy cosmic radiation. The track record over years of space missions attests to a satisfactory robustness, to wit, an efficiency degradation not exceeding ∼1% per annum and lifetimes of ∼25–30 years. Moreover, this problem is less severe on the Moon which is spared geomagnetically-trapped high-energy particles to which satellites in Earth orbit are exposed.

Specific power for III-V PV modules has reached 0.7 kW/kg [21]. But the complete system will also include (a) frames that do not have to contend with wind loading and need to support only about 1/6th of what the weight would be on Earth, and (b) power conditioners and electrical connections that incur losses of no more than a few percent. In addition, no accommodation for rain or corrosion based on atmospheric gases is necessary. The combined effect is taken to roughly double the mass of the PV system. Hence an approximate specific power of 0.35 kW/kg for the PV installation at the test conditions cited above is used in the estimates that follow.

But with a lunar synodic rotation period of 708.7 Earth hours, it would seem that maintaining full-time operation of the O2 production plants would require either:

  • (1)

    Enormous battery storage. The highest energy densities for today's mass-produced rechargeable lithium-ion batteries are ∼0.25 kWh/kg. But operation on the Moon must account for the temperature-control systems essential to functioning in the extreme lunar environment. This roughly halves the actual energy density [19,22]. Hence the battery storage that can satisfy an instantaneous load of 10 MW for at least 354 h (half the Moon's synodic rotation period) would have to be close to 3 × 107 kg of batteries, which is two orders of magnitude larger than the entire requisite PV + transmission system mass (see Sections 3 Results and discussion, 2.3 Estimates of PV area and mass). Moreover, battery system lifetimes are only several years, exacerbated by the harsh environment on the lunar surface. Hence the batteries would have to be replenished several times over the lifetime of the solar power and distribution components.

  • (2)

    Accepting that, on average, each O2 factory would function only 50% of the time, requiring launching and installing twice as many factories, thereby markedly compromising project feasibility.

Some reactors developed for O2 production from regolith have been purely thermal with almost no electrical input. Sunlight is concentrated into the reactor to decompose the metal oxides into O2 and pure metals (primarily Si, Fe, Al, Ca, Mg and Ti) at temperatures of ∼1500–2500 K [9,18,23,24]. Direct use of sunlight could in principle be noticeably more efficient than first converting solar to electricity and then converting the electricity to heat inside the reactors (inductively or resistively). However, the concentrators tailored to this lunar application have demonstrated unduly low optical efficiency and specific power [18,[24], [25], [26]]. Additionally, thermal stress as well as condensation on the reactor window from the ablated regolith resulted in unacceptably short reactor lifetimes [25,27]. Moreover, direct use of solar mandates that concentrators of tight optical tolerance (acceptance half-angles less than 1°) must be sited alongside the reactors. In contrast, using electrical input permits the complete decoupling of solar power generation from remote reactors, and the adoption of established high-temperature, opaque furnace technologies.

Taking maximal advantage of the O2 reactors, and hence of the massive investment delivering them to, and installing them on, the Moon, corresponds to operating the facilities continuously. Although part-time production may serve the initial stages and the transition to large-scale production, we treat the likely eventual scenario as one where O2 generation is required to be nominally constant. Account also needs to be taken of the possibility of malfunctions in power generation that could cease O2 production for short periods, a contingency with which electrical energy storage in batteries could contend. For example, based on the estimate cited above, 2 h of buffer storage for a complete shut-down of the 10 MW feeding all the plants would necessitate ∼1.6 × 105 kg of batteries. As derived in Section 2.2, this is comparable to the mass of the entire PV system. Accommodating a partial shut-down of a lesser percentage of the PV systems would clearly require a commensurately smaller storage capacity.

Section snippets

Basic parameters and their relevance

The mean radius of the Moon is 1737.4 km (varying from 1738.1 km at the equator to 1736.0 at the poles). For comparison, the Earth's mean radius is 6371.0 km which, due to its small asphericity, varies from 6378.137 km at the equator to 6.356.752 km at the poles. Fig. 4a plots the length of latitudinal rings on the Moon, along with the longitudinal arc length from a given point to its diametrically opposed point at the same latitude via the pole (drawn in Fig. 1b). The latter variable becomes

Siting of the power consumers

Maintaining reasonable transmission distances for the PV-generated power while still permitting sizable separation of power generation and power consumption would place the O2 production factories within no more than the order of 102 km from the PV installation. This is based on achieving low transmission losses and low specific mass for all transmission lines. Accordingly, we consider O2 factories in the region between the PV arrays and the pole (Fig. 1b).

Power centralization and distribution

Conclusions

Lunar colonization is being planned to commence in earnest within the coming few years. For the lunar environment, solar is the sole renewable resource, principally PV electricity production. An overriding challenge is to renewably drive the main energy consumer - plants that synthesize oxygen from the lunar soil for rocket propellant, orbiting satellite refueling and human sustenance - 100% of the time, lest large essential factories either lie dormant on average for half the Moon's rotational

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Jeffrey M. Gordon: Writing – original draft, The sole author performed all the research, formulated the manuscript, wrote the paper and is fully responsible for all its contents.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

I am grateful to the scientific team at Helios Ltd. (Israel) for having inspired consideration of this problem.

References (44)

  • W. Hemetsberger et al.

    Global Market Outlook for Solar Power 2021-2025

    (July 2021)
  • I.M. Peters et al.

    The role of batteries in meeting the PV terawatt challenge

    Joule

    (2021)
  • W.J. Cole et al.

    Quantifying the challenge of reaching a 100% renewable energy power system for the United States

    Joule

    (2021)
  • R. Pappa et al.

    Solar Power for Lunar Pole Missions

    (April 3, 2019)
  • M. Kaczmarzyk et al.

    Parametric study of a lunar base power systems

    Energies

    (2021)
  • A.J. Colozza et al.

    Solar energy systems for lunar oxygen generation

  • S.S. Schreiner et al.

    Integrated modeling and optimization of lunar in-situ resource utilization systems

  • R. Alexander et al.

    Moon-based advanced reusable transportation architecture: the MARTA project

  • A.C. Muscatello et al.

    Comparison of direct solar energy to resistance heating for carbothermal reduction of regolith

  • A. Dono et al.

    State of the art of small spacecraft technology

  • Cited by (0)

    View full text