Aluminum Combustion Thermophotovoltaic for High Energy Density, Long Duration Power Applications

Long range and long duration portable power require some kind of chemical fuel for practical use. Combustion of metals is a neglected method of producing power in a way that can emit no carbon dioxide pollutants. Here we propose a system using a thermophotoelectric (TPV) device to convert the emitted IR and visible light from the oxidation of aluminum directly into electrical power. A qualitative overview of the advantages of this scheme is presented, along with a simple techno-economic analysis of aluminum combustion. The advantages of this device warrant additional research into this area.

the use of internal combustion piston engines or turbines infeasible. In conventional mechanical engines, the combustion of solid fuels requires the use of a boiler and steam turbines, which would dramatically increase the size of the overall system, and also decrease efficiency compared to other engines.
Photothermalvoltaics: With the desirable mechanical engines likely infeasible for metal combustion, a new approach is needed. Interestingly, the combustion of aluminum occurs at extremely high temperatures (>2500 K based on fitting of IR spectral data [7]). This makes the fuel an attractive candidate for thermophotovoltaic (TPV) systems which convert heat/light into electricity directly by using photovoltaic (PV) cells. TPV systems are unique compared to solar PV because the heat source can be completely contained, surrounded by PV cells and reflectors, which enable a high number of low energy photons to be recycled, and increased in energy to above the band gap of the PV cell, significantly increasing efficiency. TPV systems have been shown to reach 30% spectral efficiency at 1500K in lab scale tests, and can likely reach 50% with reasonable improvements in reflector, chamber, and PV cell design [10]. This is comparable to some of the best mechanical combustion engines. Other groups have achieved 34% experimental efficiency at 1700K as well [11]. With a flame temperature of 2500K, the practical efficiency limit may be even higher for an aluminum combustion system. In addition to efficiency, TPV systems have no moving parts which the solid combustion products will interfere with, enabling the heat engine to function without mechanical failure. However, it is unclear how the resulting oxide particles will interfere with a PV cell operation. For example, the oxide particles may stick to the PV cells, blocking incoming light, and reducing power output. Particle generation and capture: In order to avoid transporting and storing powdered aluminum metal, it is useful to take advantage of the relatively low melting point of aluminum (~950K) to enable an atomizer to disperse particles to the proper size, where they can then rapidly combust. Atomizers, of which many types exist, are commonly used in the production of metal powders in spray forming processes. An additional heater can be added at the end of the atomizer to kick-start the reaction. It has been shown in previous research that (in the proper concentrations) aluminum particles will continuously burn once ignited [7]. Once the particles combust, the resulting oxide can be collected in a second container, and transported back to the aluminum smelting facility, where electrical power can be used to convert the oxide back to metal. Particle collection can be done via filtration or by other separation techniques. Finally, the NOx pollutants from aluminum combustion have yet to be quantified, but NOx reduction systems are common and can be used to reduce the NOx levels to acceptable values. A schematic of the system is shown in Figure 1.
Technoeconomic Analysis, fuel: A rough estimate of the cost of aluminum as a fuel is given in table 1, compared to other kinds of zero carbon fuel sources. This calculation makes a few key assumptions: 1. Cost of aluminum is $1.8/kg, a rough estimate based on market prices. 2. The cost of alumina, and carbon anodes constitute 52% of the market price and are neglected, as these will be re-usable [12]. 3. Hydrogen and ammonia are assumed to be produced from steam methane reforming and do not include the costs of carbon capture and storage. Ammonia costs include estimates of liquification costs at $0.13/kg.  Table 1 shows that under these assumptions, only compressed hydrogen is competitive with diesel in terms of cost. However, as mentioned before, the low energy density of the compressed hydrogen fuel is low, and therefore unsuitable for many long-range transportation applications. Furthermore, if truly CO2 free hydrogen and ammonia would be included in this list, which would be produced via electrolysis or using carbon capture, the cost would likely be significantly higher for the hydrogen and ammonia options (Also interesting is the retail price of hydrogen for passenger vehicles in California is currently well over 10 $/kg [13]). While these calculations are filled with speculations, they show that an aluminum-based combustion system may be cost competitive with other zero-carbon fuel sources. Many thanks to Richard Randall and Eddie Sun for their contributions to this analysis.
Technoeconomic Analysis, TVP system: Previous work has already been done to estimate the cost of a TVP system for grid scale storage applications [14]. In this work, the PV cells themselves constitute the largest portion of the cost of the power costs, and the production costs are assumed to be 10 k$/m 2 of PV cells, an estimate which comes from the cost of GaAs production (the authors note that silicon PV production costs are significantly lower, at ~50 $/m 2 ). A key technical metric for a TPV system is the power density of the PV cells. In the work cited, the assumed power density of the PV cells is 100 kW/m 2 . This value is consistent with many concentrated solar power systems [15]. The higher this number, the lower the overall cost of power of the system will be, as less total area will be required for the system. As an interesting side note, if these power densities can be achieved, they will be roughly a factor of 20 times greater than the areal power densities of fuel cells. The PV cells therefore contribute 0.1 $/W of cost to the system. Other additional costs of the system are found in ref [14].
The final cost of power of the system is computed to be 0.34 $/W, which is comparable to gas turbine power costs at 0.60 $/W (2015 data) [16]. However, academic cost modeling tends to underestimate the cost of new technologies, so significantly more work will need to be done in order to truly estimate the cost of a full TPV system.