Lau to function. After launch, the satellite would

Lau Jing Yi Pearlyn1, Ng Yi Da2, Chua Jia Yun3, Tan Chng Kiat31Dunman High School, 2River Valley High School, 3Defence Science & Technology Agency_____________________________________________________________________________________________AbstractThe Electric Power System (EPS) of a satellite is responsible for managing power input from solar-cells, charging of on-board batteries, and providing electrical power, at the required voltages, to all other satellite components. This paper attempts to find an efficient power supply system to fit mission and requirement of an optical sensing CubeSat, with mission of image capturing of space debris at low earth orbit (LEO), through comparison of traditional and unconventional power sources of a satellite. Traditional methods of powering a CubeSat includes the use of photovoltaic cells or solar panels, paired with rechargeable batteries, such as Ni-Cd or Li-ion batteries. Other available methods to power a satellite such as thermoelectric generators are also adopted by certain satellites to meet energy and other demands of the mission. They may also be adapted as a more efficient power supply system for an optical sensing CubeSat, provided that they are able to fit into the space restriction of a CubeSat. Through sizing up of solar panels and batteries through experimentation and a power budget, an analysis of various types of solar cells and rechargeable batteries is done to find the most suitable type of solar cells and battery for the CubeSat in this project.  IntroductionSimilar to all other machines, a satellite would require a source of energy to function. After launch, the satellite would be relying on the Electric Power System (EPS) to supply the energy required. The EPS generates, stores, conditions, control, and distributes power within specified voltage band to all bus and payload equipment on board the satellite, through the 4 subsystems in the EPS: Power source, Energy storage, Power distribution, Power regulation and control. Designing a satellite’s power system requires consideration of multiple factors such as sizing the solar cells and batteries required, effects of radiation from the space environment and the power requirement of the other satellite subsystems.This project aims to find an efficient power source that is suited for the use of an optical sensing CubeSat that is designed with mission of image capturing of space debris, using the conventional method of solar arrays and batteries, and also to obtain the appropriate sizing of solar arrays and batteries through reference of the appropriate case study of an existing CubeSat. The project aim also includes a comparison between conventional and unconventional power sources to investigate the suitability of each power source for use on an optical sensing CubeSat.  Conventional methods of powering a satellite usually includes use of photovoltaic cells and batteries, both primary and secondary depending on the duration of the mission. Nuclear power in the form of Radioisotope Thermoelectric Generator (RTG) and other sources may also be adopted, provided that they are able to fit into the space restriction of a CubeSat. Solar energy is currently the main power source in a CubeSat, but other unconventional sources of power may also be fitted to power an optical sensing CubeSat effectively, such sources of power includes: Thermoelectric generator Magnetohydrodynamic generator Fuel Cells Electrodynamic TethersCase Study – Aalto-1In order to determine the power budget and facilitate sizing up of the solar arrays and batteries for the CubeSat, assumptions of the requirements of the CubeSat has to be made by taking reference of other similar Nanosatellites. This project takes reference from the data of a research nanosatellite, Aalto-1, from the Aalto University School of Electrical Engineering.Aalto-1 is a 3U CubeSat, with a size of34cm×10cm×10cmand a mass of 4 kg. Being an Earth Observation satellite tasked with the mission of orbital science observations, Aalto-1 has similar mission properties of optical imaging with the CubeSat in our project, thus making it a suitable reference.Power RequirementThe satellite requirements, which consist of the mission information as shown in Table 1, and the power requirement throughout the satellite’s lifetime must be defined in order to size up the solar arrays and batteries for Aalto-1.Mission Lifetime 2 yearsAltitude 505 KmOrbit Inclination 97.45Initial Orbit Period 6042.4015 seconds 100.71 minutesTable 1: Basic Mission InformationWith the mission information shown in Table 1, the following results in Table 2 is also obtained:Maximum Eclipse 2279.850763 seconds 38.00 minutesMinimum Orbit in Sunlight 62.71 minutesPercentage of Orbit in Sunlight 62.27%Table 2: Orbital InformationWith the mission and orbital information, the power budget is obtained as follows in Table 3:Subsystem Peak Power required (W) Standby Power required (W) Day Duty Cycle (%) Day Avg Power (W) Night Duty Cycle (%) Night Avg Power (W) Orbit Avg Power (W)Payload 8.5 2.3 0.45 3.185 0.25 2.885 3.07181Communication Subsystem 5.05 0.2 0.25 0.7525 0.07 0.3375 0.5959205On-board Computer 0.55 0.25 0.5 0.4 0.5 0.4 0.4GPS 0.13 0.045 0.45 0.08325 0.45 0.08325 0.08325ADS Subsystem 7 0 9.51294E-07 6.6591E-06 0 0 4.15E-06ADCS Subsystem 1.8 0.5 0.1 0.63 0.1 0.63 0.63Total Power 23.03 3.295 5.05075666 4.33575 4.78E+00Power Design Margin (5%) 1.1515 0.16475 0.25253783 0.2167875 2.39E-01Subtotal Power 24.1815 3.45975 5.30329449 4.5525375 5.02E+00Table 3: Power BudgetSolar Panel & Battery sizingAll available outer surfaces of Aalto-1 is covered with solar cells, manufactured using GaInP/GaAs/Ge material. Aalto-1 uses Lithium Polymer Battery cells. Table 4 and Table 5 shows the configurations and performance of the solar panel and battery:Cell size 32cm^2Power output 375W?m^2 Solar flux density (SF) 1350 W?m^2 Nominal solar cell efficiency (?n?_SC?_?) 28%Initial losses factor (IRF) 0.90Worst case sunlight incident angle (?) 63.5Efficiency of power conversion at peak power point ( n?_peak?) 95%Efficiency drawback due to back-off from peak power point (n?_bo) 95%Degradation factor (DF) 0.95Table 4: Solar PanelThe values of n?_peakand n?_boare determined using the Peak Power Tracking approach, which sets it at 95%. The value of IRF is estimated to be 0.9 to account for the cell mismatch, wiring & diode losses, temperature correction and etc. Likewise for the value of DF after end of mission life accounts for UV degradation, micrometeorite damage, random failure, sun intensity variation, radiation effects and etc.Based on the data in the above tables, the area of the solar array and battery capacity can be deduced as shown in Table 6.Solar Array Sizing P_SA=P_sunlight/(n?_bo×n?_peak )+(P_eclipse×T_eclipse×V_BC×BRF)/(n?_bo×n?_peak×T_sunlight×V_BD ) 10.94W. Array P_SA/(SF×n_SC×IRF×DF×cos?) 0.0759m^2Battery Sizing E=P_eclipse×T_eclipse. 2.913Wh Capacity=E/DOD 14.566WhTable 6: Battery and Solar Array SizingExperiment: Deduce the duty cycle of the subsystems, the Raspberry Pi and the Camera, based on the COTS solar array and battery on hand.The orbit for this experiment is set at 400 km orbit height with 15 degrees inclination, due to the higher concentration of space debris in the set altitude.Mission Lifetime 1 yearAltitude 400 KmOrbit Inclination 15 degreesInitial Orbit Period 5540.5 sec 92.34 minMaximum Eclipse 36.01 minTable 7: Mission Information (2)The solar panel at hand mimics the solar array that is on a CubeSat, while the power bank represents the battery and power regulator on board, the former that supply power to the various subsystems while the latter regulate power distribution in the CubeSat itself.As shown in the picture below, the solar panel is connected in series with the battery, Raspberry Pi and the Camera. The setup is then placed in direct sunlight (solar irradiance of 698 W/m2) for 56.33 min, then disconnected from the solar panel for 36 minutes to investigate if the battery is sufficient to sustain the 2 subsystem during eclipse.     Figure 1: Experiment SetupFrom the experiment, it was known that the solar panel and battery are in excess of the power required by the Raspberry Pi and the camera. From there, we attempt to size up the actual size of the solar array and battery capacity needed for the CubeSat in this projectBatteryA 2200mAh, 3.7V Lithium Ion power bank was used as the battery in the setup, which gives a capacity of 8.14Wh is able to power the Raspberry Pi and the Camera for 6.78 hour inclusive of the time in eclipse. This shows that the battery used is in excess, and hence using proportion, we can deduce the smaller size of battery for the experiment.Time of Eclipse: 0.6 hour, 8.85% of total time of battery capacityActual Sizing of Battery required=8.85×2200=194.7 mAhIn comparison to other rechargeable batteries available such as the Ni technology Space batteries, Lithium Ion batteries are more suited for use on a CubeSat, due to the latter’s lighter weight and high specific energy of 100Wh/kg. Table 8 shows the comparison between the 2 main forms of secondary batteries commonly used in satellites.Battery Type Specific Energy (Wh/kg) Energy Density (Wh/L) Operating Temp. Range (?) Cycle life Mission Life (years)Ni-Cd,Ni-H2 24-35 10-80 -5 to 30 >50,000, @25% DOD >10Li-Ion 100 250 -20 to 30 > [email protected]% DOD >2Table 8: Comparison of Ni and Li BatteriesAs shown in Table 8, although Li-Ion batteries have a lower cycle life then Ni type batteries, it has higher specific energy and could last for a mission life of 2 years and below, making it the ideal battery type to be used in a CubeSat, which has short mission lifetime with strict space and weight restriction.Solar PanelOEM TopMall-BW337 solar panel made of polycrystalline silicon was used in the experiment.Cell size 187.2cm^2Power output 4W 213.7W?m^2 Nominal solar cell efficiency (?n?_SC?_?) 14.8%End of charge voltage 4.5VEnd of discharge voltage 3.5VTable 10: Information on solar panel usedThe power output of the solar panel (4W) sets the constraints on the power demanded by the CubeSat subsystems. The duty cycle is therefore set to meet the power constraints and allow for a 10% minimum overall design margin, as shown in table 11. Using the same parameters from the Aalto-1 case study and formula in table 6, the power requirement is found to be 3.54WSubsystem Peak Power required (W) Standby Power required (W) Day Duty Cycle (%) Day Avg Power (W) Night Duty Cycle (%) Night Avg Power (W) Orbit Avg Power (W)Payload 1.22 0.708 1.02Camera 1.6 0 0.25 0.4 0.03 0.048 Raspberry PI 3.7 0.5 0.1 0.82 0.05 0.66 Communication 0.66 0.58 0.629Under Raspberry PI module 2.1 0.5 0.1 0.66 0.05 0.58 Total Power 7.4 1.0 1.88 1.29 1.65Power Design Margin (5%) 0.34 0.05 0.094 0.0644 0.0825Subtotal Power 7.77 1.05 1.97 1.35 1.73Table 11: Power Budget of CubeSatFrom the experiment, it was shown that the solar panel used was also in excess of what is needed, also, with the high solar irradiance in space of 1358 W/m2, the size of the solar panel can also be reduced to produce enough power. Since solar irradiance of 698 W/m2 is able to power the components, by proportion, only 92.6cm2 or 51.4% of the original size of the solar panelCommon solar cells used in space are silicon solar cells, like those on the International Space Station, gallium arsenide solar cells and the current most commonly used multifunction gallium arsenide (GaAs) solar cells. These solar cells vary in efficiencies, with multijunction GaAs as the most efficient at 22% on average, depending on the number of light absorbing layers in the cell.According to a research by Emcore Photovoltaics regarding solar array trades between very high-efficiency multi-junction and Si space solar cells, under end-of-life (EOL) orbital conditions at LEO, multijunction solar cells significantly outperforms high efficiency silicon solar cells as shown in Table 9.Solar Cell Technology EOL Efficiency on Orbit (%) W/m2 Normalized Cost ($/W)3-mil High Efficiency Si 10.6 143 1.00Dual-Junction Solar Cells 18.1 245 1.29Triple-Junction Solar Cells 20.3 275 1.15Table 9: Comparison between Si and multijunction solar cellsAs shown in Table 9, not only are multijunction solar cells much more efficient than high efficiency solar cells, they are also able to provide more power per square metre than a high efficiency Si solar cell, with only a 15-29% difference in cost. Thus, multijunction solar cells, namely the triple junction (3J) solar cell are more suitable for use particularly for CubeSats, as they are more cost efficient, and able to provide higher power than Si solar cells given the same area of solar cells used.For example, given that 0.01m2 of Si solar cells and triple junction solar cells are pasted on each side of a CubeSat, the 3J side is able to generate more power with only a slight cost difference, while taking up the same amount of space as the Si solar cells thus making 3J solar cells more suited for use for a CubeSat, given the limited space available.LimitationsThe 2 main limitations of this experiment is one, it was conducted under ambient daylight with varying light intensity throughout the duration of the experiment, hence resulting in varying output of power by the solar panel. Two, in an actual optical imaging CubeSat, it includes other subsystems such as the attitude determination and control subsystem, resulting in higher power required as determined by the power budget. However, the experiment was done with exclusion of the other subsystems is to increase the accuracy of the tabulated values in the power budget as shown in Table 11. To improve the accuracy of the results, the oriel xenon arc lamp solar simulator or other sun simulator can be used to stimulate the sun intensity in space, so as to more accurately size up the solar array and batteries needed to power the CubeSat.Unconventional Power SourcesThermoelectric GeneratorAdvantage: Slow degradationThermoelectric generator has a relatively longer operational life with increased reliability.Plutonium is used by most thermoelectric generator. This material has a half-life of 87.7 years, therefore the power output diminish by 0.787% per year.The generator has no mechanical or chemical processes involved, therefore maintenance of generator is not required.Disadvantage: Poor efficiencyEfficiency of thermoelectric generator used in space mission lies between 3-7%. Although a thermionic converter—an energy conversion device which relies on the principle of thermionic emission—can achieve efficiencies between 10–20%, it requires higher temperatures than those at which standard RTGs run.Hydrogen Fuel CellsAdvantage: SustainabilityUnlike alkaline fuel cells or batteries, the electrolyte is not consumed and can produce electricity as long as more fuel and oxygen are provided. Moreover, the reaction in the fuel cell is reversible, thus making regenerative fuel cells possible.  Advantage: High EfficiencyHigh specific power of 275 W/kg of fuel used, and shuttle fuel cells produce 16 kW of peak powerDisadvantage: Heavy with high risk of failureFuel cells require significant support equipment and take a fair amount of propulsion fuel to place one into space, which adds significant weight and introduces potential failure modes. The two fuels must be stored in tanks cryogenically, and are delivered to the fuel cells by plumbing and valves, which could fail mechanically and cause mission failure.Fuel cells that are large enough to provide energy for very long space missions would require a lot of fuel or an external power source to provide energy to resplit the water back into hydrogen and oxygen. A large amount of fuel required would add on to the cost, while an external power source such as solar panels would also add on to the weight and deplete space in the satellite.Magnetohydrodynamic generatorAdvantage: High EfficiencyThe efficiency of MHD generators can reach up to 50%-60% through recycling the energy from the hot plasman gas. High reliabilityAs there is no mechanical process involved, MHD components are less likely to be damaged and cause malfunction.Disadvantage: High working temperatureMHD requires a working temperature of 2000-4000K. This high temperature is not suited for built-in satellite generator, otherwise a powerful and high power demand thermal control system will be required. CorrosionDue to the high working temperature, components of the generator will encounter high corrosion. This is unsuited for satellite use that requires generator to function for the whole of mission life and requires zero/minimal maintenance.Electrodynamic Tethers (EDT)Advantage: Stable Power ProductionA tether moving through the Earth’s magnetic field will experience a current flow, and the anode end of the tether collects electrons from the ionosphere and ejects them from negatively charged cathode; the electrically conductive ionosphere then completes the circuit which produce steady current for on board power results. A 20 kilometre tether in LEO could produce up to 40 kW of powerDisadvantage: Produces DragAn EDT can generate electric current flow towards Earth which can provide enough electricity to run a satellite; this also causes the tether to experience a force from Earth’s magnetic field that is opposite the tether’s direction of motion, produces drag and hence lowering the EDT’s orbit.However, EDT is a possible solution to resolve the problem of space debris as it is able to produce drag to lower a satellite’s or a spacecraft’s orbit. With that, the satellite can release a long wire antenna at the end of its mission lifetime, thus decreasing speed and lowering its orbit, allowing it to burn up in the atmosphere.ConclusionIn conclusion, the conventional way of generating power using solar arrays and batteries is still the preferred method of generating power, due to its stability in power generation and also the flexibility of sizing up of solar array and batteries to fit the power requirement, space and weight restriction of the satellite. Unconventional power sources is not as suited for use on the optical sensing CubeSat in this project, mainly due to the space and weight restriction in the design of the CubeSat. Given that the power required in a typical CubeSat is usually low, usage of solar cells and batteries is sufficient to power the CubeSat through its mission lifetime.AcknowledgmentWe would like to give thanks to the Defence Science & Technology Agency for giving us this research opportunity at YDSP, thereby allowing us to gain more knowledge and insights into space exploration and on satellites. We would also like to thank Ms Chua Jia Yun and Mr Tan Chng Kiat for their invaluable guidance throughout, which allowed for smooth completion of this project.References Chessab Mahdi, Mohammed & Sadiq, Jaafer & Abd- AL-Razak, Shehab. (2014). Design and Implementation of an Effective Electrical Power System for Nano-Satellite. International Journal of Scientific and Engineering Research. 5. 29-35. Rao, Divya & Venugopalan, S & B. M., Dayanand & C. N., Shanmugham & V, Sambasiva Rao & K. Agrawal, V. (2012). Power system design of Student Imaging Satellite. Miller, D. W., & Keesee, J. (n.d.). Spacecraft Power System. Retrieved December 26, 2017, from Graaf, H. D. (2010, May 19). Electric power subsystems in satellites. Retrieved December 26, 2017, from Maloney, S., Waddle, H., McCullar, T., & Stogsdill, J. (2007, September 7). Satellite Electrical Power System. Retrieved December 26, 2017, from “Aalto-1: the first Finnish nanosatellite,” Feb, 5, 2012, URL: N. S. Fatemi, H. E. Pollard, H. Q. Hou and P. R. Sharps, “Solar array trades between very high-efficiency multi-junction and Si space solar cells,” Conference Record of the Twenty-Eighth IEEE Photovoltaic Specialists Conference – 2000 URL: Green, M. D., Emery, K., Hishikawa, Y., Warta, W., & Dunlop, E. D. (2014, December 20). Solar cell efficiency tables (Version 45). Retrieved December 26, 2017, from Christensen, B. (2004, November 9). Electrodynamic Tethers: Getting into the Swing. Retrieved December 26, 2017, from