Solar power plants must be planned, financed and installed as system solutions. Supply subsidies are no longer the determining factor for commercial success, but rather the production price per solar electric kilowatt hour and the availability of sunlight. Solar power plants compete with conventional coal and nuclear power plants, which can produce electricity around the clock.
Utility scale in competition
Utility-scale PV systems are therefore subject to much greater competition than commercial PV facilities, which usually produce for users in nearby areas – or only for the factory beneath the PV rooftop application. Then the network charges become irrelevant to underbidding central, conventional power plants. An investor who decides to go with residential PV experiences even less competition in covering his own electricity needs, either full scale or in part. The residential PV must simply be less expensive than the (for the most part, relatively high) household electricity costs.
One can interconnect hundreds of thousands of solar modules and install gigantic solar farms that can offer competitive output comparable to fossil or nuclear power plants.
Competition with conventional power plants forces investors to calculate utility-scale PV very carefully. Above all, this relates to the choice of components, especially modules and inverters, surface area needs and operating and financing costs.
Modules: Crystalline silicon or thin-film?
Crystalline silicon has dominated all segments of the PV market for years. Thin-film modules were able to intermittently capture a high market share in utility-scale PV thanks to the high degree of automation and fast mass production. But crystalline modules are making a come-back after volume-based improvements in automation technology and manufacturing have resulted in considerable cost reduction. Higher efficiency rates may reduce the performance gap of thin-film modules.
Crystalline silicon is noted for its high level of efficiency. Monocrystalline silicon (mono Si) is somewhat more expensive than polycrystalline (poly Si), but it produces more electricity per given area due to its higher level of efficiency. The difference is so small, however, that neither form of crystalline technology has been able to outperform the other.
Thin-film PV is significantly more diverse. Amorphic silicon (a Si) has facilitated the creation of long-lived modules for years, but its efficiency rate is clearly lower than in all of the other technologies. Combining the amorphous silicon layer with a microcrystalline silicon layer does produce more power (micromorphic silicon, a Si/?c Si). Modules based on cadmium telluride (CdTe) can be produced at very low cost, and could therefore capture market share from crystalline modules for some time. Thin-film modules based on copper, indium, gallium, and selenium (CIGS) provide the highest efficiency rates; they could end the dominance of crystalline modules in the future.
Thin-film modules basically include three technological benefits when compared to crystalline modules; firstly, they use sunlight more efficiently at low levels of sun irradiation; secondly, they have lower temperature coefficients, i.e., their power output does not drop as quickly when they get hot. Thirdly, thin-film modules may be produced in a continuous manufacturing process, resulting in significant benefits in terms of manufacturing technology. The greater need for space in comparison with modules employing crystalline Si solar cells is a significant reason why thin-film solar modules in utility-scale PV play only a minor part.
The temperature coefficient indicates the percentage by which a module’s power output drops when the temperature increases. For crystalline silicon, this output is reduced by about 0.5 percent with each increase of 1 degree in temperature (°C). In thin-film modules, the value is a mere 0.25 percent. The starting point is the power output under standard test conditions at a module temperature of 25 °C.
Example: A solar power plant made of monocrystalline solar cells with a nominal capacity of 1,000 KW delivers only 800 KW during maximum sun irradiation if the solar cells are heated by 40 deg. Kelvin to 65 °C. The same nominal power plant with CdTe solar modules delivers 900 KW, by comparison.
Inverters: String or central?
Solar generators consist of combined series and parallel switching of modules. Up to 30 modules are switched in a string row, so that the electrical voltage of the individual modules combined may add up to 1,000 volts.
The DC power produced is converted by an inverter into AC current at mains voltage and frequency. The inverters regulate solar voltage and electricity in such a way that the maximum possible power is delivered despite constantly variable irradiation and temperature levels. This Maximum Power Point (MPP) is a point along the power-voltage line of the solar generator, which the inverter continues to meet, while irradiation and temperature levels fluctuate.
Converting solar electricity may occur with a certain phase shifting if necessary (in the case of a mains disruption, for example) in order to save reactive power, and to support the network.
As a rule, solar power plants are connected to only one inverter (central inverter). Since the level of efficiency of an inverter is relatively low in the partial load range, a central inverter is subdivided into several power units. Three to four inverters are therefore used in a hierarchical rank order (master and slave). As long as the radiation is low (primarily mornings and evenings), only the master is active. As soon as its upper power limit is reached, it switches to the first slave.
The characteristic of the master-slave unit is composed of the characteristic lines of the individual inverters, and therefore provides a higher efficiency rate in lower power ranges than a central inverter does. In order to provide the same load to individual inverters, master and slave are exchanged in a certain cycle, sometimes in such a way that the inverter with the lowest number of operating hours is started as the master every morning.
In order to increase yield, MPP controls are adjustable, so that each individual module is fitted with a “power optimizer”, which seeks to identify the module’s MPP, as each MPP of the individual modules varies slightly.
It is also possible to connect one or more strings to an inverter, and then connect this string inverter to a solar power plant. This solution also makes sense if not all strings are arranged at the same angle to the sun (for example, in a hilly landscape), and therefore do not achieve the same level of power at the same time, or if individual strings are overshadowed during part of the day.
Good inverters work with peak efficiency rates of nearly 99 percent, 98 percent is listed as the target value for solar farms with more than 1 MW power output. The efficiency depends upon the power. Mornings and evenings, as well as during periods of cloudy skies, the inverters work in the partial load range with a lower level of efficiency. To be able to compare inverters, the efficiency rate of each characteristic line is weighted accordingly, and depending upon the region, a “European” or a “California” rate is calculated in order to facilitate this comparison.
Due to cost reductions and increased efficiency, string inverters increasingly provide an alternative to central inverters at large PV plants. Using MPP tracking for each individual string, they are particularly suited to plants where the modules are set up at various azimuth angles. Shading that only occurs on one part of the string can also be compensated for using decentralized inverter concepts.
Most solar farms are rigid installations. The yield can be increased if one adjusts the modules to the sun’s course during the day. A complete module string is mounted on a frame which aligns the modules to the sun’s position as closely as possible from morning to night. The mechanical expenditures are relatively complex. A moving system requires more maintenance and must be just as stable as a rigid system, so that it can withstand a storm.
Power yields can be increased by up to 35 percent when the solar modules are adjusted. This applies, however, only to regions with a high rate of direct sun irradiation. In addition, it must be taken into account that installation and operating costs for these facilities can be up to 30 percent higher than for rigidly-installed systems.
Surface area requirements and latitude
Open space system modules are usually installed at a flat angle (20 to 30 deg.) to the earth’s surface in order to capture as much sunlight as possible. This orientation is a compromise, because they should be more vertical in the morning, and flatter at noon. The greater the angle, the more distance is required between each module row, so that no or minimal self-induced shadowing occurs.
In order to use the available space as best as possible, the module rows must be close together. On the other hand, they should have the greatest possible distance from one another, in order to ensure that they do not cast a shadow in the mornings and evenings. A compromise must be found between these two requirements, which can be determined by using an optimization formula, and which depends upon the latitude. The angle is largest in northern Europe, and lowest at the Equator; there it makes theoretically the most sense to install the modules closest to one another and at the flattest angle, i.e. parallel to the earth’s surface. A lower angle is also used here, however, to reduce dirt on the surfaces.
The greater the angle to the earth’s surface, the more distance is required between each module row because of shading.
Most solar farms have been installed in temperate latitudes. They use about half the available area. Trackers take up much larger areas, because when they move, they cast shadows which move around the perimeter.
Solar farms are mostly erected on surfaces that are not usable for agriculture. Suitable areas include, for example, landfills, former military areas, abandoned mining sites, industrial properties and commercial wastelands, as well as steppe and desert regions.
An expert opinion of the surface terrain and the condition and stability of the soil is provided at the start of planning. Local wind and snow loads must be considered when designing the facility, as solar power plants are subject to wind-induced vibrations over open terrain. Frameless modules display different elastic behavior than framed modules.
The most cost-effective foundations are foundations on piles driven or screwed into the ground. If the ground is not suitable, concrete foundations are used. Structural engineering factors must be established to ensure that the substructure can hold the solar panels safely in place in the presence of snow or wind loads.
Mounting systems made of aluminum or steel are used to elevate the modules. Aluminum profiles are lightweight and easy to install. They hardly corrode at all, but thermally-induced length changes in heat or frost can result in mechanical stresses. Less expensive substructures are made of steel, but this material is not corrosion-resistant.
Electrical grid stability
The more power a solar farm feeds into the grid, the more important is protection against mains power disturbances. National guidelines for supplying to mid-voltage networks take this requirement into consideration. The solar power plant must support maintaining network voltage and frequency, must balance the active and reactive voltage, and maintain the phase position at the supply point. The inverters must bridge brief network interruptions without switching off the PV facility (fault ride-through).
New generation central inverters are able to produce reactive power even at night. They take the active power from the grid to subsequently feed it back to the grid by shifting the phases of current and voltage, which is reflected in reactive power.
In the future, large-scale inverters coupled with accumulators will increasingly take over network control in cases of failure. Then the importance of conventional backup power plants will diminish (must-run units).
Large-scale accumulators are needed to buffer peak power for short periods of time. Since storage is still very expensive, there are currently only pilot projects which mainly use low-cost lead-acid batteries. Lithium-ion batteries have higher capacity with longer lifetimes, but due to their higher cost, as with the use of redox-flow cells, these are discussions for future applications, as they have just recently been approaching series production.
In addition, it is not yet clear how much storage capacity is required. If only power peaks are capped and need be buffered for only a few minutes, then a capacity of one megawatt hour may be sufficient for a 50 MW solar farm, for example, in order to cap 10 MWs and feed it into the batteries for six minutes.
The capacity must be significantly larger, if the solar power produced throughout the day must also be available in significant amounts at night. This option is currently far from being economically feasible.
The many years of experience gained from the erection and operation of large solar energy plants has already led to the development of standardized units. The first plant of this type consists of a 3 MW power plant covering an area of 155 m × 215 m. Larger solar energy plants are simply put together using several of these standardized 3 MW units, so that PV power plants of any size can be erected.
In order to realize the most cost-efficient power generation possible, everything is uniform in design. A very specific module type and an assembly system designed specifically for this are used at all standard power plants. With a module inclination of only 8° and very little space between the rows of modules, a high installed capacity per km2 is achieved.
Due to the high input voltage of both central inverters (1,500 volt DC) a standardized voltage level is reached on the output side, so that inexpensive transformers can be used.