How does an solar inverter work?
Solar inverter wellsee WS-P2000
Solar inverter wellsee WS-P3000
Sine wave inverter wellsee WS-P4000
Sine wave inverter wellsee WS-P5000
Sine wave inverter wellsee WS-P6000
One of the most incredible things about photovoltaic power is its simplicity. It is almost completely solid state, from the photovoltaic cell to the electricity delivered to the consumer. Whether the application is a solar calculator with a PV array of less than 1 W or a 100 MW grid-connected PV power generation plant, all that is required between the solar array and the load are electronic and electrical components. Compared to other sources of energy humankind has harnessed to make electricity, PV is the most scalable and modular. Larger PV systems require more electric-cal bussing, fusing and wiring, but the most complex component between the solar array and the load is the electronic component that converts and processes the electricity: the inverter.
In the case of grid-tied PV, the inverter is the only piece of electronics needed between the array and the grid. Off-grid PV applications use an additional dc to dc converter between the array and batteries and an inverter with a built-in charger. In this article we discuss how inverters work, including string, or single-phase, and central, 3-phase inverters; explore major inverter functions, key components, designs, controls, protections and communication; and theorize about future inverter technology.
KEY INVERTER FUNCTIONS
Four major functions or features are common to all transformer-based, grid-tied inverters:
• Inversion
• Maximum power point tracking
• Grid disconnection
• Integration and packaging
Inversion. The method by which dc power from the PV array is converted to ac power is known as inversion. Other than for use in small of-grid systems and small solar gadgets, using straight dc power from a PV array, module or cell is not very practical. Although many things in our homes and businesses use dc power, large loads and our electrical power infrastructure are based on ac power. This dates back to the early days of Edison versus Tesla when ac won out over dc as a means of electrical power distribution.
An important reason that ac won out is because it can be stepped up and travel long distances with low losses and with minimal material. This could change in the distant future if more of our energy is produced, stored and consumed by means of dc power. Today, the technology exists to boost dc electricity to high voltages for long distance transfer, but it is very complex and costly. For the fore-seeable future, ac will carry electricity between our power plants, cities, homes and businesses.
In an inverter, dc power from the PV array is inverted to ac power via a set of solid state switches—MOSFETs or IGBTs—that essentially flip the dc power back and forth, creating ac power. Diagram 1 shows basic H-bridge operation in a single-phase inverter.
Maximum power point tracking. Te method an inverter uses to remain on the ever-moving maximum power point (MPP) of a PV array is called maximum power point tracking (MPPT). PV modules have a characteristic I-V curve that includes a short-circuit current value (Isc) at 0 Vdc, an open-circuit voltage (Voc) value at 0 A and a “knee” at the point the MPPT is found—the location on the I-V curve where the voltage multiplied by the current yields the highest value, the maximum power. Diagram 2 (p. 70) shows the MPPT for a module at full sun in a variety of temperature conditions. As cell temperature increases, voltage decreases. Module performance is also irradiance dependent. When the sun is brighter, module current is higher; and when there is less light, module current is lower. Since sunlight intensity and cell temperature vary substantially throughout the day and the year, array MPP current and voltage also move significantly, greatly affecting inverter and system design.
The terms full sun or one sun are ways to describe the irradiance conditions at STC (1000 W/m2). Sunlight intensity varies from nothing to full sun or a little more than one sun in some locations and conditions. This means that PV output current can vary from zero to full array rating or more. Inverters need to work with arrays at their lowest voltages, which occur under load on the hottest days, as well as at their highest voltages, which occur at unloaded open circuit array conditions on the coldest days. In some climates, temperatures can vary by 100°F or more, and PV cell temperatures can vary by 150°F. This means array voltage can vary by ratios of nearly 2:1. A string of 22 Evergreen ES-A-210 modules, for example, will reach a Voc of 597 Vdc with a cell temperature of -30°C (-22°F). The MPPT voltage (Vmp) can get as low as 315 Vdc in an ambient temperature of 50°C (122°F). In most cases, the maximum power point voltage operates over a 25% variation. However, this number is lower in regions with more consistent year-round temperatures, such as San Diego, California, and is higher in regions where temperature varies more, such as the Midwest and Northeast. Finding the array’s MPP and remaining on it, even as it moves around, is one of the most important grid-direct solar inverter functions.
Grid disconnection. As required by UL 1741 and IEEE 1547, all grid-tied inverters must disconnect from the grid if the ac line voltage or frequency goes above or below limits pre-scribed in the standard. Te inverter must also shut down if it detects an island, meaning that the grid is no longer present. In either case, the inverter may not interconnect and export power until the inverter records the proper utility voltage and frequency for a period of 5 minutes. These protections eliminate the chance that a PV system will inject voltage or current into disconnected utility wires or switchgear and cause a hazard to utility personnel. If an inverter remained on or came back on before the utility was reliably reconnected, the PV system could back feed a utility transformer. This could create utility pole or medium voltage potentials, which could be many thousands of volts. A significant battery of tests is performed on every grid-tied inverter to make certain that this situation can never occur.
Table 1 shows the voltage and frequency limiting values and the time periods that the inverter has to be offline, referred to as learning times. Notice that some values are different for inverters under 30 kW and those over 30 kW. Tree-phase commercial inverters over 30 kW have limits that can be adjusted with the permission of the local utility. This can be very useful in an area with a fluctuating grid, which often results in a significant loss of energy. Long utility lines, areas with heavy load cycling or an unstable island of power grids can all contribute to grid fuctuation. If a PV system signifcantly underperforms as a result (beyond nuisance tripping), adjusting the inverter limits can be benefcial.
UL 1741 and IEEE 1547 also require that inverters not create a power island. This means that if the utility goes out, the inverter cannot remain on, producing power to any load or portion of a building load, including rotating or oscillating loads. For example, even if the building’s load is similar or exactly balanced with the output of the PV system, the inverter may not remain on if the utility is not present. Algorithms for detecting anti-islanding must constantly check to see that the utility grid is really present. A specially tuned“resonant load” set up to mirror the utility tests this inverter function. The resonant load is made up of a very specific inductive, capacitive and resistive network with many settings. Its goal is to attempt to trick the inverter’s anti-islanding algorithm into thinking that the utility is really there, at many different prescribed power levels called out in the UL 1741 standard. This load is connected to the inverter operating at full power, and the grid is connected. Te resonant load is set to the exact output power of the inverter. When the whole system is stable, the utility is disconnected while the resonant load maintains voltage and frequency within the inverter’s limits. The inverter has a maximum of 2 seconds to successfully recognize that the grid is disconnected and shut off.
Manual ac and dc disconnection means are designed into inverters or PV systems so that the inverter can be disconnected from the grid and the PV array if service technicians, installers or other qualified personnel need to turn of the inverter or access the main inverter enclosure. Automatic ac disconnection means—such as an ac contactor—are used to minimize or totally eliminate nighttime tare losses and reduce susceptibility to damage from nighttime power surges and lightning strikes.
Disconnecting power supplies, chips and components of all types at night also extends their service life. The power electronic module, such as this compact 55-pound power module from a 100 kW PV inverter, is just one of the many components that are integrated into the final inverter package.
Diagram 4A 60Hz, transformer-based, single-phase inverter circuit.
inverter packages, for commercial and utility scale applications, may ship as more than one enclosure.) Packaging also protects the inverter from the outside elements and keeps unintended guests, human or otherwise, away from the equipment. The use of high quality materials and finishes is necessary to meet the needs of the application. The service life of a PV inverter, for example, requires the use of corrosion-resistant fasteners, like stainless steel screws, to ensure that individual components can be accessed and serviced over 25 years.
KEY INVERTER COMPONENTS
In this section, we discuss key inverter components. As a starting point, basic inverter operation is illustrated by looking at a single-stage, single-phase, 60 Hz transformer-based inverter. Additional inverter topologies are explained subsequently.
Solid state switches. All inverters today use some combination of power semiconductors—IGBTs, MOSFETS or both in some cases—to invert dc to ac power. Other key components in the main power inversion circuit are inductor(s), capacitors and a transformer, either 60 Hz or high frequency. The latter is used in transformer-based inverters to adjust voltage levels as needed by the topology and to provide galvanic isolation between the solar dc input on one side and the inverter’s ac output to the grid on the other. Single stage products like 60 Hz transformer-based string inverters typically use an H-bridge for inversion from dc to ac, as shown in Diagram 1 (p. 69).
Diagram 4 shows all the key components in a single stage inverter, including the H-bridge circuit. The switches at the far left represent the power semiconductor switches. By alternately closing the top left and bottom right switches, then the top right and bottom left switches, the dc voltage is inverted from positive to negative, creating a rectangular ac waveform.
In a grid-interactive, 60 Hz transformer- based inverter, however, the output current needs to be a sine wave form. This requires a more complex operation. The H-bridge puts out a series of on-of cycles to draw an approximated sine wave shape. This is known as pulse width modulation (PWM). With a 250 Vdc to 600 Vdc input, the H-bridge circuit for a typical 60 Hz transformer-based string inverter will put out an approximated sine wave with an ac voltage of about 180 Vac. The role of the components after the H-bridge is to smooth and change the magnitude of that approximated sine wave.
Magnetics. Te string inverter in Diagram 4 contains several pieces of equipment that are referred to as magnetic or magnetic components. These include the inductor and the transformer, shown to the right of the H-bridge. These magnetic components filter the wave shapes resulting from the PWM switching, smoothing out the sine waves, and bring ac voltages to the correct levels for grid interconnection. The magnetic also provide isolation between the dc circuits and the ac grid.
Note that the ac wave-form entering the inductor is raw and triangular; but on leaving the device, it is clean 180 Vac sine wave. Because 180 Vac cannot be directly connected to the utility grid, it goes through a 60 Hz transformer. Te resulting smooth, sinusoidal 208, 240 or 277 Vac inverter output is connected to the grid. Grid synchronous operation is made possible by grid sensing feed-back. Grid voltage information is provided to the inverter’s digital signal processor (DSP) or microcontroller, the device that controls the H-bridge.
Magnetics are labor and material intensive, and their costs are tied to expensive commodities like copper and iron. They can also take a costly tare on system performance, and careful design is needed to achieve maximum efficiency. There are two main loss components associated with the use of magnetic. The first component includes core losses,
which involve the magnetic material (such as iron laminations, magnetic ribbon or sintered powdered material) and gap losses, which result when the magnetic components have a gap between core halves, for example. The second loss component is the conductor or coil loss. These are simply the resistive losses in the many coils of wire around the magnetic cores. Good inverter design needs to minimize both types of losses.
Minimum dc input voltage. A wide dc input voltage window is beneficial to PV system designers and installers, since solar arrays operate over a wide voltage range. An even wider voltage window is required to enable designers to select between a wide range of PV products and string configurations. Achieving this wide dc input voltage range is not easy, because inverter designers have to balance concerns like efficiency, circuit complexity and cost.
The laws of physics also limit inverter designers. The inverter’s dc input bus voltage needs to be greater than the peak of the ac voltage on the primary side of the transformer. In order to maintain this relationship at all times, an additional control and safety margin is required. With a minimum PV input voltage of 250 Vdc, for example, the highest amplitude ac sine wave you can create is about 180 Vac, as illustrated in Diagram 5Te PV input voltage, of course, will greatly exceed 250 Vdc in many array layouts or temperature and light conditions. If 250 Vdc is selected as the inverter’s lower dc voltage design limit, then the H-bridge will always create an ac sine wave with a magnitude of 180 Vac. This is true even when the dc voltage present is 300 Vdc, 400 Vdc or higher. This is because the rest of the 60 Hz transformer-based inverter needs to operate on a relatively fxed ac voltage. Te voltage on the utility side of the inverter’s transformer—the second-ary side—is fxed within a small range of variation. Inverter designers must set the voltage on the primary side of that transformer accordingly.
Capacitors. The most important use of capacitors in the inverter power stage is for filtering ripple currents on dc lines. Ripple is an undesirable phenomenon caused by power semiconductor switching. Capacitors are also used to keep the dc bus voltage stable and minimize resistive losses over the dc wiring between the PV array and the inverter, since the resulting current from the array to the inverter dc bus is constant. A relatively smooth dc voltage and current at the input of the inverter allow good PV voltage regulation, which results in an MPP tracking algorithm that works well and has high accuracy.
Lower frequency capacitors, typically electrolytic types, make up the main capacitance on typical inverter bus structures. These have very high capacitance and the ability to filter large ripple currents. High frequency capacitors, typically film capacitors of various types, are also used for filtering out high frequency noise and spikes from power semiconductor turn-on and turn-of cycles.
Capacitors, particularly electrolytic types, are susceptible to failure from long-term operation in hot environments. Inverter designers have to carefully select the proper capacitors to ensure that they can take the heat and absorb the high ripple currents that are possible. Selecting high-grade versions and using more than are required to minimize heating of the capacitors caused by the ripple current are typical approaches.
A different and newer approach, seen in some commercial 3-phase inverters, involves the use of high capacity film capacitors. Unlike electrolytic capacitors, film capacitors cannot dry out and will therefore last longer. They are also less affected by temperature. Te trade-off is that film capacitors are more costly and take up more space. However, that trade-off can be very worthwhile, especially on large inverters where space is less an issue. Commercial and industrial PV systems will produce large quantities of electricity over 25 years or more, so the inverter needs to be as reliable and long-lasting as possible.
Maximizing efficiency. Optimizing efficiency, or reducing loss, is an important part of inverter design and component selection. The goal is to optimize the inverter for maximum efficiency, while maintaining high reliability and delivering a product at a good price. Because there are losses associated with each, components of efficient design include the choice of IGBT or MOSFET power semiconductor switches, switching frequency, method of switch control and turn-on and turn-of method. Many of these choices require a careful balance of waveform smoothness, noise, reliability and efciency.
Because the installed cost for PV systems is high, requiring subsidies to make fnancial sense, the benefts of high efciency are compelling. A 1% increase in inverter efciency translates into immediate and long-term savings, a result of increased energy harvest and increased compensation for that energy. An even more powerful way to look at efciency on a 100 kW PV system is that a 1% gain in efficiency means you could install 1 kW less of PV. This results in upfront savings of $6,000 to $8,000.
As the installed cost for PV decreases, inverter efficiency may become less critical than it is today. However, it will always be better to convert as much PV power into ac power as possible. To do otherwise results in waste heat. With greater inverter efficiency, less energy and fewer materials are needed for the inverter’s cooling system, resulting in prolonged inverter life. Thermal performance. As far as thermal performance is concerned, the first goal of inverter design is to minimize loss. The next goal is to minimize temperature gain for maximum inverter component life. The last step is to minimize power requirements and energy consumed for cooling system needs. Where a significant temperature differential exists between inverter components, different types of components are often separated into temperature zones within the inverter’s overall enclosure. This is especially useful when lower temperature components are also more sensitive to higher temperatures. This approach is utilized in many 60 Hz transformer-based inverter designs, both string and central types. Te 60 Hz transformers can operate at much higher temperatures than semiconductors, capacitors and other electronic components.
Software and monitoring. To reliably control the inverter, the software designed to run on the inverter’s digital signal processor or microcontroller is developed over years of code writing and debugging. Te most critical control is the one driving the power stage. This creates the PWM waveforms that generate the sine waves ending up on the utility grid and at the building’s loads. Software also controls the inverter’s interaction with the grid and drives all the appropriate UL 1741 and IEEE 1547 required controls and events. Another part of the software controls the MPPT function that varies the dc voltage and current level as required to accurately and quickly follow the moving MPP of the PV array. All of these major functions, as well as a multitude of others, are carried out in unison like an orchestra. Software is used to drive the contactor that places the inverter on the grid in the morning and of the grid at night. Software controls temperature limits and optimizes cooling system controls. Software, its development history and robustness, is a critical element in any inverter.
A completely different aspect of the inverter’s software is related to communication with other inverters, PCs and data onto the Internet, as well as building management systems. Sometimes this is done by separate devices, such as data monitoring devices or Internet gateways that gather data from the inverter or integrated software and hardware within the inverter. Data monitoring is an important part of a PV system since it lets owners and installers know its status and provides quick alerts if there are inverter or PV system faults. Since an inverter already measures and calculates much information regarding ac and dc sides of the PV system, it is typically a convenient place to gather this data, process it and place it on the Internet, for example.
Some inverter manufacturers provide PC or Web based monitoring options; some inverters are compatible with third-party monitoring systems; and some inverters have both options available. Monitoring is one area of inverter development that is evolving quickly. Several inverters manufacturers or third-party monitoring providers offer advanced features, like revenue grade monitoring, PV string and subarray monitoring, weather monitoring and inverter-direct monitoring.
ALTERNATIVE INVERTER TOPOLOGIES
So far we have discussed the design and operation of a single-stage, single-phase, 60 Hz transformer-based string inverter. Other common inverter design topologies and applications include 3-phase inverters, high frequency inverters, bipolar inverters, transformer less inverters and battery based inverters. In some cases a single inverter product may incorporate several of these features. This is the case with Solar on inverters from Advanced Energy, which are 3-phase, bipolar and transformer less products. While we do not address this exact combination of features in this section, we do explore the most common alternative inverter topologies for PV applications.
3-phase, 60 Hz transformer-based inverters. The operation of a 60 Hz transformer-based, 3-phase inverter is very similar to that of the string inverter. Te difference is that a central inverter has three phase outputs instead of two. In order to generate three phase outputs, 60 Hz transformer-based central inverters typically use a 3-phase bridge. This is a bridge with a 6-switch design.
