We have developed an innovative hybrid-power system for the Plateau Observatory (PLATO). This Antarctic astronomical site-testing facility was successfully deployed by an expedition of the Chinese Polar Research Institute in January 2008. PLATO is located at Dome A, the coldest, highest, and one of the most remote sites in Antarctica. The plateau is characterized by high altitudes of up to a few kilometers above sea level, and the atmosphere is thinner compared to the same physical altitude at nonpolar locations. Of all the high Antarctic research sites shown in Figure 1, Dome A is the only one that does not have the infrastructure to provide power and heat for research instrumentation. Antarctica's unusual day–night cycle dictates that continuous solar-power generation is impossible for much of the year. The local wind speed is very slow, thus preventing the effective use of wind turbines. This, combined with extremely cold temperatures (ranging from −30°C in summer to as low as −80°C in winter), poses interesting challenges and boundary conditions for in situ power and heat sources for small remote facilities.
Figure 1. Map of Antarctica with contour levels in meters. The locations of various research sites are indicated. (Map courtesy of the Australian Antarctic Division.)
Small power systems for remote facilities have operated successfully in Antarctica.1 The British Antarctic Survey used solar and wind energy and rechargeable batteries to achieve up to 2.5W of electrical power.2 Lockheed Missiles and Space Company Inc. developed the Automated Geophysical Observatory with a power system operating on the basis of a thermoelectric generator fueled by liquid propane, providing up to 100W.3 The University of New South Wales (Australia) designed the Automated Astrophysical Site-Testing International Observatory (AASTINO),4 which used a hybrid power system consisting of solar panels and two Stirling engines to provide up to 400W of electrical power. PLATO is designed to continuously provide at least 1kW of electrical power throughout the year, with peaks of 4kW.
Figure 2. (top) Inside the PLATO engine module, showing three of the Hatz 1B30 diesel engines. (bottom) PLATO's solar array. (Image by Zhenxi Zhu.)
The PLATO power system has incorporated many of AASTINO's features. However, during the winter months we decided to use diesel instead of Stirling engines because the former can produce greater power output with a better efficiency. There is no preferred wind direction at Dome A, making it impossible to protect the instruments from the diesel engine's exhaust stream. Therefore, PLATO consists of an instrument module fitted with onboard heaters, connected by an electric umbilical cable to a separate engine module located at a distance of 50m. The waste heat from the engines cannot be used to warm the instruments. Instead, it heats the fuel tank and oil systems.
PLATO's engine module includes six Hatz 1B30 diesel engines arranged in two banks of three, mounted on a 4000l fuel tank: see Figure 2 (top). They run on Jet A-1 fuel mixed with a small amount of two-stroke motorcycle oil. To extend the engines' operating life they share a 60l external tank of Delvac 5w-40 lubricating oil, which is continuously filtered. Typically, only one engine runs at any one time. They are started electrically using a stack of 3000 Farad ultracapacitors. These were chosen over batteries for simplicity of charging and improved low-temperature performance. Four of the engines are coupled to eCycle brushless alternators, and the other two run Mavilor disk motors as direct-current generators. During the summer months, two 500W solar-panel arrays can be used as the primary power source: see Figure 2 (bottom).
A low-pressure test chamber was constructed to study the effect of high altitude on performance. The engine was found to start and run reliably at simulated altitudes of up to 5000m. Several test runs of ~10min each, at a particular speed and load, were conducted at both sea-level atmospheric pressure (1000hPa) and a pressure of about 540hPa, corresponding to a slightly higher altitude than that of Dome A. The fuel consumed, engine runtime, load voltage and current, and the exhaust-gas temperature were recorded. We present two examples of the specific fuel-consumption behavior as a function of engine power (from a much larger data set) in Figures 3 and 4.
Figure 3. Comparison of the diesel-engine brake specific-fuel consumption at 2200 rotations per minute (rpm) between sea level and an altitude slightly higher than that of Dome A.
Figure 4. Comparison of the diesel-engine exhaust-gas temperature at 2200rpm between sea level and an altitude corresponding to an atmospheric pressure of ~540hPa.
Here, the power recorded is the electrical output from the alternator and thus includes alternator and rectifier losses. The brake specific-fuel consumption figure is therefore an overall fuel-to-electricity efficiency. The increased exhaust-gas temperature at high altitude observed in Figure 4 agrees with theoretical expectations. As similar amounts of heat must be generated to produce a given power, but only 54% of the mass of air is present in the cylinder at Dome A altitudes, combustion will raise the temperature of the gas there by almost twice as much.
Only minor modifications to the diesel engines are required for satisfactory operation at altitudes of up to 5000m. Little loss of efficiency, if any, is encountered when the engine is run at high altitude. PLATO has successfully been in autonomous operation for 204 days. Its innovative power system has wide potential applicability to future small astronomical facilities on the Antarctic plateau, requires little human intervention, and poses minimum environmental impact.
This research was supported by the Australian Research Council. We thank the team at Hatz Australia for generously sharing their knowledge and experience. We also thank Apollo Solar, Switchgear Commissioning and Maintenance, and Eaton Electric Systems. Special thanks are due to Jason Allen for his development of the oil-circulation system.
Shane Hengst, Michael C. B. Ashley, Jon R. Everett, Jon S. Lawrence, Daniel M. Luong-Van, John W. V. Storey
School of Physics
University of New South Wales
Graham R. Allen
Solar Mobility Pty. Ltd.