Central inverters convert the DC power collected from an array of solar modules into AC for connection to the grid. They are typically floor or ground mounted, as opposed to string inverters, which are typically installed on a wall or other vertical structure. They range in power from around 50kW to over 1MW and can be designed for indoor or outdoor use.
Architectures and Features
The simplest architecture of a central inverter consists of a single DC-AC conversion stage. Some inverters have a DC-DC boost stage to increase the MPP voltage range. In some cases, a low frequency transformer is provided at the output to boost the AC voltage and provide isolation. However, this decreases the efficiency and increases the size, weight and cost of the inverter. The trend is to use transformerless inverters in commercial installations with a front-end boost stage if required.
While central inverters in Europe have always been 1000V rated, they are available in two DC voltage categories in the U.S.: 600V and 1000V. The 600V DC category is most commonly used in commercial rooftop projects, but also in some utility-scale projects. The 1000V DC rated inverters tend to be larger (>500kW) and have been used in “behind-the-fence” (utility-owned) projects for several years. Recently, smaller 1000V DC central inverters are being employed in commercial rooftop projects in response to code changes.
Central inverters for commercial projects in the U.S. have either 480V or 208V 3-phase outputs. Inverters used in utility projects typically have a lower AC output (determined by the minimum DC voltage) which is connected to a MV transformer. It is common to parallel the outputs of several inverters on a single transformer winding if the arrays are ungrounded. Recently, the wind industry standard of 690V AC is becoming a popular output voltage to take advantage of BOS cost savings.
Central inverters can be either monolithic (using a single power train and MPP tracker) or modular (with multiple power trains). Modular inverters are more complex, but have the advantage of being able to operate at reduced power in the event of failure of one or more modules. They could use either a multi-MPPT or a Master-Slave control approach. The multi-MPPT approach essentially uses a separate converter and MPPT for each floating sub-array, which increases the overall energy harvest under partial shading conditions. In the Master-Slave approach, the Master module is always on and commands other modules (Slaves) to turn on as more power is available from the array, which maximizes inverter efficiency under low insolation conditions.
Solar inverters have historically been required to operate at unity power factor and disconnect from the grid in case of under-voltage or frequency (anti-islanding). They are now increasingly being required to support the grid with features such as low-voltage / low-frequency ride through (LVRT/ LFRT), reactive power support and dynamic power factor control. This requires the ability to control the inverter parameters and not just monitor them.
Challenges and Future Trends
As inverters have grown bigger over the years, manufacturers have been able to address thermal management issues by using liquid cooling and new semiconductor and magnetic materials. The challenges of managing the DC collection of bigger arrays is driving the push towards higher DC voltages (1500V DC inverters are already available). New approaches such as distributing small central inverters around the array are being considered. As the maximum efficiencies of inverters approach 99%, the focus of innovation is increasingly on reducing cost, implementing smart grid features, improving reliability and standardizing monitoring and control interfaces.
By: Tilak Gopalarathnam, Director of Research and Development at REFUsol Inc