Japan's Northernmost Mega Solar Plant Fighting Against Snow

5-year struggle over tilt angle, height of mounting system

2013/12/22 19:34
Kenji Kaneko, Nikkei BP CleanTech Institute

"Milk Road" is one of the main roads from Wakkanai Airport to the city of Wakkanai. Every morning, trucks loaded with fresh milk use the road, driving from local farms to milk processing factories. As the road supports the regional economy, in winter, snow is removed preferentially from this road. As you drive along Milk Road, you will see a group of solar panels neatly standing close together in the midst of the wetland and green hills. This is the 5MW "Wakkanai Megasolar Power Plant" (Fig. 1).

Polycrystalline silicon panel chosen from 5 types

The construction of this power plant commenced in 2006 as a verification facility of New Energy and Industrial Technology Development Organization (NEDO). The construction cost totaled approximately ¥5 billion (approx US$48 million). The project was aimed at operating a large-scale solar power generation system under severe weather conditions such as snow cover and frigidity, and studying grid stabilization technologies that leverage large-sized secondary cells.

The facility was transferred for free to Wakkanai City in March 2011 following the five-year verification test. Along with the 74 wind power facilities with approximately 76MW output in the city, the solar power plant is now a showpiece attraction for visitors as a renewable energy facility that represents the "environmental city Wakkanai," which is aiming for energy self-sufficiency.

Starting as a verification facility, the power plant gradually expanded as different kinds of solar panels were installed in four phases. In phase 1, in 2006, five types of solar panels from nine manufacturers were set up to verify the amount of power generation during the cold season.

Introduced then were the single- and poly-crystalline silicon panels manufactured by Sharp Corp, Kyocera Corp, Mitsubishi Electric Corp and SunPower Corp of the US; the multi-junction silicon (HIT solar) panels manufactured by Sanyo Electric Corp (now Panasonic Corp); the amorphous silicon (non-crystalline silicon) panels manufactured by Kaneka Corp and Mitsubishi Heavy Industries (MHI) Ltd; and the CIS panels manufactured by Solar Frontier KK and Honda Soltec Co Ltd.

After monitoring the amount of power generation for a year and calculating the "performance ratio (PR) rating," which indicates to what extent the expected power generation yield has been achieved, the crystalline silicon type showed relatively better results.

In solar cells, power generation loss increases in accordance with a rise in temperature. The loss at high temperature is large in the crystalline silicon cells though their conversion efficiencies are higher than the other types of solar cells. Their PR ratings were higher probably because such flaws of crystalline silicon cells did not show much thanks to the low temperature in the verification project in Wakkanai.

Based on the assessment, crystalline silicon panels and polycrystalline ones, in particular, whose cost efficiency is higher among those tested, were installed in and after phase 2. As a result, the polycrystalline silicon panels account for roughly 90% of the solar panels with 5MW output.

(Continue to the next page)

Measures to make snow slip off even at 30 degrees

When it comes to measures against snow cover, the "tilt angle" of the panels on the mounting system and the "height" between the back of the panels and the ground are significant factors. The tilt angle needed to be acute enough for the snow on the panels to slip off while the panels have to be set at a height so they would not be buried in snow even after heavy snow.

Snow on the panels would slip off more easily as the tilt angle sharpens. But the amount of power generation would be lower as the sun would shine over the panels at a slant. The higher the panels are mounted, the less they are likely to be buried under snow. But the installation cost would rise.

In addition, the panels' shadows grow longer and shade other panels more in the morning and afternoon as much as the tilt angle is sharpened and the mounting height is raised. If spaced far apart enough not to shade other panels, however, the number of panels that can be installed on the site would decrease.

In the verification test, NEDO explored the optimum balance in the paradoxical relation between the amount of power generation and measures against snow cover by changing the tilt angle and the height from the ground depending when the panels were installed.

In phase 1, the tilt angle was 33°, and the mounting height was 2m. The panels were tilted at 33° because they can generate the most amount of power at that angle in light of the sun's culmination altitude. And it was estimated that snow would not accumulate to 2m given past records. However, the snow on the panels did not slip off at an angle of 33°.

Consequently, in phase 2, the tilt angle was sharpened to 45°, and the height was lowered to 1m. Then, the snow on the panels slipped off, but the fallen snow accumulated to over 1m under the panel and covered the lower area of the panels.

In phase 3, the tilt angle was unchanged (45°), but the height was raised back to 2m (Fig. 3). As a result, the snow slipped off the panels and once fallen did not cover the panels.

In phase 4, the panels were tilted at 30° with an "arrangement to make snow slide off more easily" added to the panel surface to verify its effect (Fig. 4). The arrangement included, for example, a measure to fill in the slit between the panel frame and the glass surface by injecting silicon caulking (Fig. 5). This was because it had turned out that snow tended to stay on the panels because of the gaps and slits on the solar panel's surface. The arrangement had a significant effect with snow slipping off most panels even at an angle of 30°.

On the other hand, the raising of the solar panels, a measure taken in parallel with the caulking, did not work efficiently. The measure was meant to space the lowest row of panels from the upper panels by raising the lowest row when mounting them. Snow accumulating on the upper panels would slip beneath the mounting system from the space with the lowest row, instead of accumulating only under the lowest row.

Exactly as expected, the snow on the upper panels slid down from the space to the inner side of the mounting system. But the accumulated snow beneath the mounting system caused pressure on and distorted the panels on the lowest row from the ground. Hence, NEDO gave up applying this measure.

(Continue to the next page)

Utilization rate even on par with national average of 11.8%

When studying grid stabilization technologies using large secondary cells, NEDO installed NGK Insulators Ltd's sodium-sulfur (NAS) cells with 1.5MW output in conjunction with the 5MW solar power plant (Fig. 6). The system was operated primarily focusing on two technologies: "output fluctuation control" and "programmed operation."

In solar power generation, the output drastically fluctuates in a short time should clouds block the sun on a sunny day. If such fluctuation could be controlled, the load on the power grid can be alleviated. Setting a target of controlling the range of fluctuation by about 80%, a control effect almost equivalent to this target was confirmed in the verification test.

"Programmed operation" is a method of controlling the operation to achieve the programmed load (trend in the amount of power generation) by predicting the hourly output of solar power generation based on the weather forecast and formulating a power generation program in advance in accordance with the discharge and charge of the secondary cells. Although the reproducibility is significantly affected by the accuracy of the weather forecast-based prediction of the power generation volume, improvement in the prediction accuracy was confirmed through the verification test.

After the verification test, Wakkanai Megasolar Power Plant became a facility for the power selling business run by Wakkanai City. Following the implementation of the feed-in tariff scheme, the plant now brings about power sales of ¥140 to 150 million to the city every year. The plant is also the focus of studies where the aging impact can be assessed since it is one of the first large-scale solar power plants in Japan.

In fiscal 2010, its yearly utilization rate was 11.8%; in fiscal 2011 the rate was 10.1%; and in fiscal 2012 the rate was 10.1%. Despite the slight fluctuations caused by the weather, the rate has been stable since operations started in 2006.

Cold regions do not seem to be suited for solar power generation, but this plant can achieve a facility utilization rate equivalent to the national average (12%), as seen in fiscal 2010. In the amount of power generation by month, its facility utilization rate was relatively higher in Japan during the cold period without snow from April to June (Fig. 7).

From this verification test, key factors and challenges to large-scale solar power plants in snowy regions surfaced; that is, (1) how to boost the amount of power generation through measures against snow cover during the period between November and February where the amount of power generation decreases greatly and (2) how to generate as much power as possible in spring and summer, when it can generate more power than in other months.