This analysis applies only to polar locations with "moderately" cold temperatures, therefore it is a reliable guide for sites on the Antarctic continental margin, West Antarctic Ice Sheet, and periphery of Arctic landmasses. It does not apply to the Antarctic Plateau or interior of the Greenland Ice Sheet where much colder temperatures prevail. For this reason, and also the scarcity of sites for field verification, no figures are presented beyond 84 degrees latitude.
TABLE 1. Minimum SLA amp-hour capacity needed for year-round operation of a 5-watt system.
LATITUDE |   60   |   62   |   64   |   66   |   68   |   70   |   72   |   74   |   76   |   78   |   80   |   82   |   84   |   86   |   88   |   90   |
2 Panels (170W)   |   900   |   900   |   900   |   1000   |   1200   |   1400   |   1600   |   1800   |   2000   |   2100   |   2200   |   2300   |   2400   | ---- | ---- | ---- |
3 Panels (255W) |   700   |   700   |   800   |   900   |   1100   |   1300   |   1500   |   1700   |   1900   |   2100   |   2200   |   2300   |   2400   | ---- | ---- | ---- |
4 Panels (340W) |   500   |   500   |   600   |   800   |   1000   |   1200   |   1400   |   1600   |   1800   |   2000   |   2200   |   2300   |   2400   | ---- | ---- | ---- |
Although wind power has been demonstrated to significantly reduce the number of batteries required, wind power input and turbine reliability are highly variable. Because this analysis is intended to be a conservative lower bound on the number of batteries required at a given latitude, wind power is not included. However for reference, a 5W system has been operated year-round at 79 S with 2 solar panels, 1000 amp-hour batteries, and two Forgen 500 wind turbines. A similar design has been operated at 85 S with a ~1 month wintertime data outage.
Power Analysis:
The power analysis relies in part on output from the PV-DesignPro-S v6.0 software package from Maui Solar Software. This program uses solar intensity, earth orbit parameters, and local clearness index observations to calculate solar irradiance at ground level on an inclined solar panel. The program generates irradiance on an hourly basis, with weather variations modeled using randomization functions. A total of 28 locations between 59 and 80 degrees north and south latitude were individually analyzed to generate the results presented here.
The annual power budget is then produced by tracking input from the solar panels, acceptance and delivery of power from the battery bank, and usage by the continuous load. The battery bank size is increased until it has enough capacity during winter to maintain operation until sunlight returns to power the load. The selected solar panel model was a BP585 85-Watt module, and the battery calculations are based on experience with the Deka 8G31 gel cell battery.
Validation:
Accuracy of this analysis was validated against independent observations from existing year-round GPS installations in the Antarctic. For both Fishtail Point (78.9 S) and Cape Roberts (77.0 S), year-round operation was predicted, including two versions of solar/battery arrays deployed at Cape Roberts. Furthermore, observed performance of the UNAVCO GPS Test Facility at McMurdo Station (77.8 S) during winter 2006 was modeled. This system was intentionally underpowered to observe power cycle behavior, and the dates of both autumn shutdown and spring startup were accurately predicted by the analysis.
Assumptions:
1. From summer solstice to winter solstice, 100% of available solar panel power (minus the load) is used to charge the battery bank. From winter solstice to summer solstice, 30% of available solar panel power (minus the load) is used to charge the battery bank. These figures are meant to account for the actual acceptance of solar charge by "warm" batteries as the sun sets in autumn versus "cold" batteries when the sun rises in spring. At no point are the batteries allowed to overcharge, i.e. nominal capacity is not exceeded.
2. For latitudes lower than 64 it is assumed that batteries will deliver 100% of their nominal capacity during wintertime discharge. For latitudes higher than 64, it is assumed that batteries will deliver 70% of their room-temperature capacity. The 70% figure has been validated by testing at McMurdo Station (78 S). There are two reasons for such a high yield of battery power under Antarctic use. First, a lead-acid battery will deliver significantly more than its nominal capacity when discharged at a very slow rate. The Deka 8G31 is rated for 97 amp-hours at a 5 amp rate, however a GPS system will only draw ~40 mA from each cell in a 1000 amp-hour battery bank. At this discharge rate the battery will delivery ~140% of its nominal capacity. Second, the batteries are charge during summer and autumn while "warm", then discharged cold during winter. Such a battery will deliver substantially more charge than one charged and discharged while cold. For this reason, a 30% reduction of the battery's low-current capacity is assumed for latitudes below 64, and a 50% reduction for latitudes above 64.
Last modified: 2019-12-24 02:12:28 America/Denver