PV Power Plants 2013 - Industry Guide: Technology

Technology

Striving for increased efficiency

: Crystalline modules (left) and thin-film modules: Owing to the relatively high efficiency of crystalline modules, a smaller installation area is needed per unit of output compared with thin-film modules.
Photos: Tom Baerwald

Until just a few years ago, PV plants were chiefly built from components that were often only available in limited quantities and PV modules were lacking in particular. This is changing as production capacities are undergoing dramatic expansion. Today, solar power plants are planned, installed and financed as system solutions, and at the end of this chain comes the price per kWh of solar electricity, which competes with that of other technologies. During the first stage of the market launch, the solar energy price per kWh must be lower than the feed-in tariff in order to ensure that the PV power plant is profitable. During the second stage of market penetration, it is sufficient for the price to be less than that of conventional power generated by coal-fired or nuclear power plants. Return on investment is therefore determined by the efficiency of the entire system, from individual modules to inverters and grid feed-in.

Crystalline silicon or thin-film

: Today, thin-film modules are mainly used in large installations.
Photo: Tom Baerwald

Business in MW-scale PV power plants is dominated by crystalline silicon and cadmium telluride. Crystalline silicon solar cells have many advantages: Com mercial mono- and polycrystalline silicon modules now achieve 14 to just over 20 percent efficiency. Owing to the relatively high efficiency of these modules, less installation area is needed per unit of output, which also means that fewer mounting frames and cables are required. New “cast-mono” wafers achieve similarly high efficiencies to monocrystalline solar cells thanks to the particularly cheap polycrystalline silicon wafers employed.

Innovative processes, such as the string-ribbon process, may reduce manufacturing costs further. A thin ribbon of silicon is pulled from a melting crucible between two wires. This technique facilitates the production of ultra-thin wafers just 135 micrometers (µm) thick and eliminates the heavy material losses that ensue from sawing conventional silicon wafers, which are 180 µm thick. Tandem cells that combine crystalline wafers with amorphous silicon coatings achieve extremely high efficiencies with the wafers being exceptionally thin. Novel cells with selective emitters and rear-side contacts are only between 100 and 120 µm thick. Bifacial cells that are also able to absorb sunlight on the rear sides of modules and thereby boast yields of between 25 and 30 percent higher than comparable standard modules have also been on the market since 2012.

As thin-film modules are significantly less efficient than crystalline silicon modules, they need to cover up to 30 percent more surface area than crystalline silicon modules to achieve the same output (cf. fig. 2). This entails increased costs for installation, support frames and cabling. However, thanks to intensive research and development the efficiency of thin-film modules is currently improving at a faster rate than that of crystalline silicon modules. The efficiency of research modules fitted with cadmium telluride cells has now increased to more than 18 percent, while that of cells made from copper-indium semiconductors has risen to more than 17 percent. If these efficiency values are also attained in commercial modules, the fact that thin-film solar modules require a larger surface area will no longer have so such a bearing.

Since it is now possible to manufacture thin-film solar modules in large numbers, the costs of doing so have plummeted. In terms of price per unit of output, thin-film modules now cost the same as crystalline modules. However, the manufacturing costs (and therefore the prices) of solar modules with crystalline silicon solar cells have recently fallen at a greater rate than those of thin-film solar modules, resulting in the current market share of thin-film technologies declining somewhat.

As a general rule, large-scale PV plants equipped with thin-film modules can produce power just as cheaply as those constructed using crystalline modules. The lower manufacturing costs may therefore compensate for the increased outlay incurred for installation.

In addition, thin-film modules perform favorably in diffuse light conditions and at high temperatures. They utilize weak light more efficiently and, compared to crystalline silicon, their output is not so badly impaired when they heat up. This makes them significantly more suitable for the hot climate of the sunbelt, semi-desert and desert regions.

Colossal factories for producing thin-film modules with silicon technology or with compound semiconductors of copper indium gallium selenide are currently under construction. Production capacities for inexpensive cadmium telluride (CdTe) modules have experienced the strongest growth. Both CIGs and CdTe technologies have made the successful leap to bankability, demonstrating their reliability in large solar farms. In spite of this, 2011 and 2012 saw the market share of thin-film technology recede, as the decline in prices for crystalline modules coupled with their high efficiencies led to a resurgence of crystalline silicon technology in large-scale plant construction. Nevertheless, the new factories being built across the world are virtually exclusively intended for the production of thin-film modules. Significantly smaller investment funds are required here than for silicon cell and module manufacturing technology, so it can be expected that a growing proportion of thin-film modules will be found in large-scale plant business over the medium term.

The temperature coefficient

: Measuring the temperature coefficient
Photo: Tom Baerwald/Solarpraxis AG

The temperature coefficient indicates the percentage by which a module’s output will drop as its temperature increases. With crystalline silicon modules, output falls by around 0.5 percent for every degree of increase in temperature. This value is just 0.25 percent with thin-film modules. To calculate this, the output under standard test conditions at a module temperature of 25 °C is used as a starting point. To illustrate, a solar power plant with monocrystalline solar cells and a rated output of 1,000 kW will only generate 800 kW, even under maximum insolation, if the solar cells’ temperature rises by 40 kelvin (K) to 65 °C. In contrast, a power plant with the same rated output but equipped with CdTe solar modules will yield 900 kW.

Quality assurance from factory to construction site

: guarantees long-term yield stability and operational safety.
Photo: Tom Baerwald

The long-term yield stability, operational safety and thus investment security of a photovoltaic system are primarily dependent on the longevity and reliability of the modules and system technology used. These qualities are examined in complex test procedures as laid down in industry standards or technical guidelines. test certificates and additional quality marks provide evidence that a module meets the necessary requirements.

Independent institutes certify solar modules using test samples supplied by the manufacturers. The PV+Test (www.pvtest.de) introduced by Solarpraxis AG and the German Technical Inspection Association (TÜV Rheinland), on the other hand, buys samples covertly so that manufacturers are not able to preselect which items are tested. Increasing importance is being attached to testing for potential induced degradation (PID), electroluminescence cell and module inspection and thermography, as these analyses reveal weak points in production and allow flaws to be detected faster.

In addition to these types of approval tests, spot checks at photovoltaic power plants ensure that the manufacturer data corresponds to the components that are actually delivered. Such quality testing can involve different levels of complexity and cost, from visual checks and performance measurements to electroluminescence and thermal imaging. The purpose of testing is always to safeguard the anticipated long-term yield and minimize the risks of technical failure. Drones are increasingly being employed to fly over large-scale, multi-MW plants in order to gather thermal images of the module arrays. Solar farms with outputs in excess of 10 MW can be inspected using helicopters.

Extra audit

When buying a large quantity of modules, it is always sensible to place the products delivered, their manufacturers, as well as the components and materials used under closer scrutiny. An audit helps to validate the results of various module tests and estimate the risks of a project. As certificates from different test institutes may vary considerably, results are more meaningful if manufacturers subject their products and components to more stringent testing, and not simply to that specified in the standards. There is no such thing as absolute security, but reasonable assessments can be made. Checklists can also prove useful in analyzing the risks of a product.

Standardization

At present, manufacturers supply products with highly specific features in an attempt to make them as distinct as possible from those of their competitors. If the wide variety of solar modules could be reduced through standardization, it would open up enormous scope to bring down the costs associated with systems technology. This is why large project developers are changing their strategy to offer only standardized power plant units comprising just one specific module type and one particular plant design (substructure and string connection). These modular units allow the solar plant output to be scaled up rapidly with the greatest of ease.

Inverters

: String inverter: The electricity is fed into the grid by several, independent inverters.
Photo: Tom Baerwald

Solar generators are a combination of solar modules connected in series and in parallel: Depending on the voltage of a given solar module, up to 30 of them may be connected in series to form a string, so the electrical voltages of the individual modules will add up. Connecting 20 to 30 modules in series produces a string with a system voltage (direct current, DC) of up to 1,000 volts (V).

In very large solar farms, it can sometimes make sense to increase the DC voltage to 1,500 V, as this allows for lower current in the DC cabling. Consequently, smaller cable sizes can be used, which in turn significantly lowers the cost of cabling. At the same time, thermal losses are reduced owing to the low DC current. On the other hand, the junction boxes must be equipped with fuses certified for these voltages, which are more expensive than those for lower voltages. The inverters, too, must be approved for the higher DC input voltages, and thus require power electronics that are designed for such purposes. Suitable transistors are even more expensive. The major advantage of having a higher system voltage is the option it provides to combine solar arrays with wind energy systems. Such hybrid power plants take up the same area and feed power into the grid via a joint switching station.

Inverters regulate solar voltage and solar power such that the solar generator will furnish the maximum possible output, even with constant fluctuations in temperature and insolation. They convert the direct current generated by photovoltaic systems into alternating current that can be fed into the grid. While smaller systems feed single-phase power into the low-voltage grid (grid voltage of 400 V), larger solar power plants with outputs of 100 kW and above feed three-phase current into either the low-voltage or medium-voltage grid, which can have a voltage of between 20 and 50 kV, depending on the national grid standard in a given country.

Not all inverters are suitable for every type of module (depending on MPP voltage window, whether or not a transformer is present and whether or not the DC circuit is grounded). It should therefore be checked during planning if the inverter is suitable and approved for the chosen modules.

Central inverters versus string inverters

: Central inverter housing: Several strings are connected to one inverter.
Photo: Tom Baerwald/Skytron Energy GmbH

In a central inverter design, several strings are connected to one inverter with an output of up to 2 MW. The device is usually made up of several output units of say 500 kW. These units operate in a master/ slave configuration where one is responsible for controlling the system (master) and switches on the inverter’s additional output units (slaves) as insolation and generator output dictate.

As a result, inverter operation under partial load – unfavorable owing to the low conversion efficiencies achieved – becomes less frequent, which increases the system yield by several percentage points. The inverter units regularly exchange roles (master/slave) in order to balance out the operating times of the device parts and increase service life. In a string inverter system, just a few strings are connected to an inverter with a lower output. In the case of solar generators, where strings track the sun on individual tracking units, it can prove beneficial to equip each tracker (i.e. each string) with its own inverter. Assessments of technical and economic viability will determine which option will best improve a given system. When performing such assessments, the impact of an inverter and its efficiency on plant yield is given the same consideration as how the choice of system design will affect the system and installation costs. Ultimately, preference will ideally be given to the system with the lowest levelized cost of electricity, allowing for the plant to be operated in the most profitable way possible (cf. fig. 3 and 4).

Module Level Power Management

Of late, module level power management (MLPM) solutions have increasingly gained precedence. Known as power optimizers, they are called upon to increase the output of individual modules. They enable maximum power point tracking (MPPT) for each module and minimize mismatches arising from production tolerances or shading. Micro inverters, also known as module inverters, take things a step further. They convert direct current from the module directly into grid-ready alternating current. This eliminates losses and avoids a whole series of additional technical problems which can result from complex DC cabling in large inverter systems.

An enhanced version of the power optimizer was recently launched onto the market as module maximizer. This device not only tracks the MPP, it also records the output data of a module at any given moment and sends this to the central monitoring system. This allows drops in the performance of individual modules to be detected straight away.

European and Californian efficiency

Inverters operate less efficiently if low insolation means that they can only feed in a portion of their rated output, as is the case, for example, in the morning or afternoon, or in cloudy conditions – operation under partial load thus results in lower efficiency. Weighted efficiency enables a comparison to be drawn between the efficiency of different devices. European and Californian weighted efficiencies are two commonly used comparison standards that correspond to the differing insolation conditions in Central Europe (weaker, more diffuse insolation) and California (stronger, more direct insolation).

Good inverters operate with peak efficiencies of almost 99 percent. 98 percent is taken as a guideline for solar parks with outputs of over 1 MW. The amount of power consumed by the inverters themselves is an important factor which largely determines the feed-in period over the course of the day.

Above all, high efficiency and high plant availability mean higher yields: If the average inverter efficiency or annual availability can be increased by around three percent, a 1 MW solar park will generate approximately 86,000 US dollars of additional revenue within ten years.

The capacity of inverters should be adequately proportioned to allow regional peaks in insolation to be fully exploited without the devices becoming overloaded. It has become established that the STC rated output of solar modules should correspond to the AC rated output of inverters. Here, too, it is important to carefully plan the optimum technical and economic solution. Recent findings based on insolation readings taken at short intervals provide clues as to how additional yields can be harvested.

The key to the grid

: Increasing conductibility has resulted in an ever greater number of PV plants feeding their power into the medium- and high-voltage grids.
Photo: Tom Baerwald

The greater the output a solar farm feeds into the grid, the more important protection against grid failure becomes. The directive on medium voltage grid feed, which came into force in Germany, takes this requirement into account. If grid stability is threatened, the grid operator can either disconnect a plant or use it to stabilize the grid. This may include maintaining grid voltage and grid frequency, balancing real and reactive power in the grid and phase shifting at the feed-in point. Inverters must also be able to ride through short grid interruptions of 200 milliseconds without shutting down the plant (fault ride through). This capability allows them to support the grid, meaning that large-scale PV plants have great potential for stabilizing power grids. New central inverter models are even capable of stabilizing the grid with reactive power and freeing up grid capacity during the night. To do so, they take real power from the grid and then feed it back into the grid at an efficiency of 99 percent, though when this happens a dramatic phase shift is seen in the current and voltage, which manifests itself as a high level of reactive power. In the future, large inverters fitted with batteries will take charge of grid management in the initial critical milliseconds of instability. These solar inverters are able to stabilize the grid as they can substitute the inertia of the rotating masses found in large-scale fossil fuel power plants by using energy reserves from storage devices. This also results in energy utilities’ conventional regulating power plants, which are also known as must-run units, becoming redundant.

Permanent monitoring of plant operation is essential for investors and operators alike, as it permits faults and failures to be recognized and rectified quickly, keeping yield losses to a minimum. Automatic operation monitoring and error diagnosis systems, which can either be integrated into inverters or installed separately, send alerts to operators via e-mail, text message, smart phone or cell phone and identify potential causes of error.

Grid feed-in guidelines

In the US, IEEE standard 1547 (voltage and frequency tolerance) applies to grid feed-in. There, inverters must also be able to identify when the subgrid is shut down. If power generation and power consumption balance one another out in this subgrid, a photovoltaic plant may continue to work independently – the grid will therefore remain live. In the event of “islanding”, as this is called, a solar installation’s inverter must therefore disconnect it from the grid. In Spain, the technical connection requirements must be regulated by contract between plant operators and grid operators (Art. 16 real Decreto 661/2007).

The economic and technical requirements for feeding solar electricity into the utility grid vary across the European Union, and even within member states. In Germany, the Renewable Energy Sources Act (EEG) provides a legal and economic framework, while the Medium Voltage Directive of the German Association of Energy and Water Industries (BDEW) lays down technical specifications. In fact, it must be possible for grid operators to control all the inverters in a solar park centrally. The BDEW Directive came into force in Germany 2011, and the Low Voltage Directive, containing special requirements on feeding into the low voltage grid, was introduced soon afterwards. Inverter manufacturers must provide certificates to demonstrate that their devices meet the new guidelines, and plant certificates must be issued following the installation of solar parks to confirm that their interaction with the power grid is appropriate and has been simulated.

Since July 2011, large-scale solar plants in Germany must be issued with a certificate proving their compliance with the technical specifications of feed-in management before they can be connected to the grid. The certification process is conducted by independent test institutes. In mid-2012, these test institutes were not numerous enough to process the backlog of applications. Therefore, transitional provisions were implemented until the end of 2012.

Foundation and anchoring

: Fixing modules to the ground
Photo: Tom Baerwald/Parabel AG

Solar parks should reliably generate electricity for at least 20 years. They are generally built on open land, such as former military sites, landfill sites, former mining fields or hitherto unutilized fallow land. Planning starts with a survey of the relief, solidity and the quality of the ground. Local wind and snow loads must also be taken into account when designing a photovoltaic plant.

Just like bridges, large-scale solar installations are vulnerable to wind-induced vibrations, though frameless modules exhibit different elastic properties to their framed counterparts. The simplest and least expensive types of foundation are ones that use piles driven into the earth. Alternatively, piles are also available that are screwed into the ground (screw pile foundations). Concrete foundations made from either ready-mixed or in-situ concrete provide a further alternative for applications such as tracking systems or those where the piles cannot be driven into the ground.

The modules are supported by systems made of wood, aluminum or steel. Wooden structures are comparably light but will warp in the course of 20 years. They must be waterproofed and should not come into direct contact with the soil. Aluminum is also an extremely light material. Systems made of this are easy to install and hardly corrode, but the price of aluminum fluctuates greatly. Furthermore, owing to the thermal properties of aluminum, heat and frost cause greater stresses in the structure. While steel substructures are more cost-effective, steel is not resistant to corrosion and is therefore unsuitable for some locations.

The mounting system must be capable of supporting the solar modules securely for a long period of time. Mounting a freestanding PV power plant is frequently easier than many other types of installation, as the construction area is more easily accessible than, say, a slanted roof. A big disadvantage of free-standing plants, however, is that they lack a truss to which to screw the assembly system. This is why anchoring the mounting frame safely into the ground is a factor which should be given adequate consideration, as it will need to keep the equipment stable for decades.

Tracking systems

: Dual-axis tracking system
Photo: Tom Baerwald/Solarpraxis AG

Depending on their location, crystalline silicon modules can furnish up to 35 percent higher yields if they are able to follow the path of the sun mounted on trackers with one or two axes. Nevertheless, the higher investment costs and additional maintenance required are more likely to pay off in southern regions which receive a high proportion of direct insolation. In northern areas of Central Europe, financing this additional expenditure is becoming less and less worthwhile given falling module prices and higher maintenance costs.

Single axis trackers rotate PV arrays so that they follow the sun’s daily path from east to west. Dual axis (hemispheric) systems also tilt on a vertical axis to follow the sun’s movement.

The use of trackers entails considerable additional preparation work on the foundations; the ground must also be sufficiently stable. Furthermore, the surface area required for tracking systems is larger than that for non-tracking PV installations as, to avoid shading, trackers must be positioned at a sufficient distance from each other.

Cabling

Losses due to cabling are often underestimated. If the plant is badly planned, total energy losses in copper cables can add up – anything over one percent is unacceptable. To avoid high losses, the cable cross section must be relatively wide, while cable lengths should be as short as possible. During installation and operation of the PV plant, the plug connectors on the solar module cables must be checked to ensure that they are water tight and that the connections are not prone to fault voltage or short circuits.

Cables should not be exposed to direct solar radiation, so should be laid in shaded areas. This is because every degree of temperature increase in the copper material increases electrical resistance and multiplies losses as a result. What is more, sunlight (UV light in particular) can degrade the cable insulation material.

Lightning and overvoltage

Large ground-mounted PV plants always need their own protection system against lightning and overvoltage. Overvoltage can be caused by indirect lightening strikes and the associated electromagnetic induction in cable loops. According to TÜV Rheinland, in Germany almost 50 percent of all damage to PV plants is caused by overvoltage. Protection against direct strikes (direct strike lightning protection) or coupling as a result of strikes elsewhere in the grid (indirect strike lightning protection) must be taken into consideration during the initial stages of planning. Shutdown systems should also be integrated that allow the PV system to be swiftly disconnected from the grid in the event of fault voltages or fires.

Storage systems for solar power

: Battery park belonging to a storage system developer: Storage systems capable of storing large quantities of energy for minutes or even hours at most are needed to drive forward the energy transition.
Photo: Tom Baerwald/Younicos

Storage systems for PV power plants are just beginning to be competitive and it remains to be seen if mechanical, electrochemical or electrical energy storage will predominate in large-scale PV plants in the future.

While in 2012, lead-acid and lithium-ion batteries were only initially supplied for small-scale systems to optimize on-site consumption in private households, the first large-scale storage systems with a storage capacity of 500 kWh came onto the market in the second half of the year. In cases where the solar output must be regulated by grid operators, these systems allow the solar yield of a 1 MW power plant to be buffered for 30 minutes.

Synthetic methane (power to gas) is opening up a promising new path for technological development that will allow solar power to be stored on a major scale. The methane, which is generated from surpluses produced by the solar power plant, is known as wind gas or solar fuel. Once grid capacity becomes available again, methane can be converted into electricity in traditional gas turbines and fed into the grid. It is also used to generate peak load power.

Operation and maintenance

In addition to costs of the technology itself, the cost of operation and maintenance is another important factor to consider. These costs do not figure in construction of the installation, but can add up quite considerably over the service life of a solar power plant. For each kWh of solar power generated, between one and ten euro cents fall to the costs incurred during the 20- to 25-year operation of a plant. This large spread is owed to the wide range of potential costs: For example, these can include wear caused by extreme weather or vandalism. Expenditure is also incurred for monitoring plant safety and protection against theft. In the case of off-grid systems, battery costs are the real killer. Currently, investors can expect to pay around 10,000 to 14,000 euros for operation and maintenance on top of the costs for the technical system.