New system architectures, control methods, and device designs promises much more efficient solar power
BY DAN KINZER
Fairchild Semiconductor, Portland, ME
People are increasingly aware of the dramatic change in attitude concerning renewable energy in political and economic circles worldwide in recent times. Though some of the strong incentives were curtailed last year in the period of global recession, the market for solar energy installations is expected to grow at a 30% to 40% annual rate for the next several years (see Fig. 1).
This growth will be across the range of powers, from the large megawatt central power generating stations to the rooftop residential systems around the world. Efficiency of both the cells and the power electronics are climbing ever higher. At the same time, new topologies and device designs continue to raise the level of performance.
Fig. 1. The market for solar energy installations is expected to grow.
By far the predominant solar cell technology is the single crystal silicon pn junction cell (see Fig. 2). This has been the most widely available material, and offers a good tradeoff between cost and efficiency for a broad range of applications. High-power concentrator systems may choose multijunction cells that achieve upward of 25% efficiency, but at much higher cost. Lower-end systems may choose polycrystalline or thin-film systems that offer lower efficiency, but very attractive manufacturing costs.
Fig. 2. The predominant solar-cell technology is the single-crystal silicon pn junction.
Typical solar technology
A typical power system consists of two power stages (see Fig. 3). The front end is a boost converter that takes the panel output voltage and boosts it to a dc bus voltage that is just high enough to feed power into the line through an inverter.
The input to this system is the solar cell array, which may be a panel, a string of panels, or a parallel and serial combination of panels. Each panel commonly generates a voltage in the range of 50 to 60 V. These are then strung together to arrive at the desired dc voltage prior to the boost.
Fig. 3. A typical solar power system has two power stages.
The system also consists of a maximum power point tracking (MPPT) mechanism. Any solar cell or string of solar cells has an output voltage where the power is maximized; lower the voltage and the current doesn’t rise enough to compensate, or raise the voltage and the current drops too quickly. There is typically a computation system that multiplies voltage and current measurements to determine what that maximum power point is and control the output voltage to that value.
In a string of cells, the output current is determined by the lowest output current in the string. If illumination varies or any one cell is partially shaded or obscured, all other cells will be limited and will not operate at their peak output power capability.
There are many ways to compensate for that, depending on the system design. In large central power-generating stations, the cells are usually arranged in large open areas without shade, and often even track the sun to maintain maximum incident isolation at all times.
In smaller systems, however, the arrays can be arranged at different angles to incident sunlight, can be partially shaded, or may just operate at different efficiencies. In such cases, it is desirable to partition the string and operate each partition at the MPP. Then each dc output voltage can be summed. A controller may actually inject a current into weaker partitions to balance and optimize the output current from the entire unit.
To be continued
