On the way to an ‘all-electric society’: Potential for the use of higher voltages in the field of renewable energy

By Michael Geiss
fraunhofer ise
Medium voltage string inverter for future utility-scale PV power plants. Image: Fraunhofer ISE

In late 2023, researchers at Fraunhofer ISE put into operation the world’s first medium-voltage string inverter as part of a project to explore the power electronics for the next generation of large-scale PV power plants. According to the project’s leader, Michael Geiss, the signs are looking positive for the introduction of medium voltage in PV.

The current climate targets can only be achieved with a step away from fossil fuels to the production and use of renewable energies. This step is also a step towards an “all-electric society”. According to the current state of technology, enormous quantities of raw materials are required for this transition. On the one hand, this is necessary to connect a wide variety of consumers and producers to our grids intelligently and with low losses using power electronics.

This article requires Premium SubscriptionBasic (FREE) Subscription

Unlock unlimited access for 12 whole months of distinctive global analysis

Photovoltaics International is now included.

  • Regular insight and analysis of the industry’s biggest developments
  • In-depth interviews with the industry’s leading figures
  • Unlimited digital access to the PV Tech Power journal catalogue
  • Unlimited digital access to the Photovoltaics International journal catalogue
  • Access to more than 1,000 technical papers
  • Discounts on Solar Media’s portfolio of events, in-person and virtual

Or continue reading this article for free

At the same time, however, large quantities of copper and aluminium cables are also required to connect all the components in this grid. The raw materials used must therefore be used as efficiently as possible, as our planet’s reserves of raw materials are finite.

One way of saving raw materials is to use higher voltages. With the same power, the current and therefore the required cable cross-section can be reduced. This has long been used in transmission and distribution grids to keep losses low. Especially in the field of utility-scale PV power plants, the use of higher voltages offers enormous savings potential.

figure 1
Development of primary energy distribution until 2050. Credit: Fraunhofer ISE

The graph above shows the required expansion of photovoltaics to meet humanity’s hunger for energy. By 2050, 63TW of installed PV capacity will be required worldwide. A technical solution to save materials therefore has a major impact here.

Motivation for the development towards higher voltage levels

The PV industry is subject to enormous cost pressure. This is one of the reasons why considerable investment has been made in the further development of this technology in recent years. Technological advances and economies of scale in production have brought down the price per installed kWp by around 14% per year in recent years.

figure two
Relative development of the costs of PV modules and BOS components starting from 2006. Credit: Fraunhofer ISE

The graph above shows that the reduction in PV module costs has made an enormous contribution here. The installation costs and the BoS hardware costs now account for 40% of the total costs, as the pie chart below shows. In the future, savings in this field will therefore have a much greater impact on the overall price.

figure three
Breakdown of component costs according to their shares in a utility-scale power plant. Credit: Fraunhofer ISE

This is where the aforementioned increase in output voltage comes into play. The currents can be reduced by increasing the system voltage. This leads to savings at various points: the most visible effect is the significant reduction of cable cross sections, as shown in the image below.

figure four
Minimum cable cross sections for 250kW at different AC voltages. Credit: Fraunhofer ISE

This is because the required cable cross section increases approximately quadratically with the flowing current. In other words, doubling the output voltage (i.e. halving the current) leads to a reduction in the conductors’ cross section of approximately 75%. Only the pure current-carrying capacity of a single conductor is considered here. If the cables are laid in bundles in the main lines of a power plant, the current-carrying capacity must be corrected downwards due to the poorer heat dissipation.

This means that a larger cross section is required for bundled conductors in order to transmit the same current. By reducing the cross sections, the bundles also become smaller, which in turn means that smaller reduction factors can be applied for the bundling. As a cable is not only made of copper, the monetary savings in bill-of-materials costs will not be the same due to the increased insulation.

The increase in raw material requirements mentioned at the beginning and the associated price increases must be taken into consideration here. If the price of copper rises sharply in the future, the proportion of reducible costs in the overall cable will also increase. In the case of cables, the savings potential lies not only in the pure material costs; the laying and connection of smaller cross sections is also much simpler, thus reducing installation costs.

But the move to medium voltage does not only affect the cables; the increased voltage can also increase the output of the subsystems. Today’s power plants usually use subsystems between 3 and 5MVA in size. A higher output is difficult to achieve in a low-voltage transformer due to the large copper cross sections. If the voltage is increased, a winding with the same cross section can transmit a higher power. With a voltage of 1,500V, 10-12MVA is already possible in a transformer and thus in a substation.

This results in a smaller number of transformers and switchgear, while the power plant size remains the same. This also means fewer construction works and lower installation costs. At the same time, there is more space for PV modules.


As the graph below shows, the above-mentioned potentials have been utilised in the past. As the power of the inverters increased, the output voltage also continued to rise. Since 2018, the increase in output voltages has stagnated despite further increases in output power.

figure five
Development of the output voltage and power of string inverters in recent years. Credit: Fraunhofer ISE

The main reason for this development is the definition of “low voltage”, which is limited to 1,500Vdc or 1,000Vac. Above this value, the “medium-voltage” range begins, which is internationally part of the high voltage range. However, all PV-specific standards are currently only available for low voltage. A further increase is therefore associated with a significantly higher standardisation effort. This is because not only do the PV-specific standards themselves have to be adapted, but they also refer to other basic standards from the medium-voltage field of application.

At this point, a chicken-and-egg problem arises: without increased interest from the industry, the standardisation committees will not tackle this major standardisation issue. On the industry side, however, the lack of a normative basis and test specifications is an obstacle to the move to medium voltage.

However, the standardisation situation is not the only hindering factor here. For the construction of a medium-voltage PV power plant, a wide variety of new components from different manufacturers is needed. Here too, one manufacturer has to lead the way. But all beginnings are difficult because the introduction of a new technology lacks any economies of scale, which makes pricing more difficult. This is a particular hindrance in a price-sensitive market such as the PV sector.

Added to this is a rapidly growing market with increasing demand. The resulting certainty in sales has reduced the pressure on component manufacturers to innovate.

Another reason was the challenge of producing a highly efficient inverter for medium-voltage applications. For the step up to medium voltage, semiconductors with blocking voltages above 1.7kV are required. Silicon components have switching speeds that are too low for this application and therefore lead to increased losses. Highly efficient inverters for medium-voltage applications can only be built with silicon carbide (SiC) components with high blocking voltages.

State of the art in research and technology

In the early days of SiC semiconductor development, there was great euphoria around the technology, and the advantages of wide bandgap devices were demonstrated in a wide variety of components. These included components with blocking voltages of up to 25kV. Even back then, higher voltages were being considered in the field of PV and the first research projects were underway. However, the costs of the new semiconductors were still too high and the technology was still plagued by various teething troubles.

This meant that components with high blocking voltages in particular were still too far away from being ready for serial production. With the rise of electromobility and the increasing demand for <1.7kV SiC components, all manufacturers focused on optimising these voltage classes. The further development of components for higher voltage classes came to a standstill. SiC has now fully arrived in electromobility.

More manufacturers are once again focusing on the higher voltage classes. The 3.3kV class is available on the market from various manufacturers and the 6.5kV class is well on its way to serial production. In the meantime, various research projects have shown that the construction of highly efficient inverters based on SiC is technically feasible.

Researchers at Fraunhofer ISE have developed the world’s first medium-voltage string inverter and successfully put it into operation on the grid. It was developed as part of the publicly funded MS-Leikra project. The inverter has an output voltage of 1,500Vac with an output of 250kVA. It has a two-stage design. The step-up converter with a PV input voltage of 1.7kV to 2.4kV is based on 3.3kV SiC semiconductors. The inverter section was constructed using hybrid ANPC modules. Four silicon and two SiC semiconductors are used here. This topology allows the major advantages of SiC to be utilised at only slightly higher costs.

Higher voltages are also possible. A three-phase converter with 3kVac/250kVA was also developed at Fraunhofer ISE, as shown below. The converter unit can be connected in parallel on the DC and grid side, each with its own LCL filter.

figure six
A 250kVA inverter stack with 3.3kV SiC transistors developed at Fraunhofer ISE. Image: Fraunhofer ISE

It is an ANPC topology with 3.3kV SiC MOSFETs, using three half-bridge modules per phase. The switching frequency at the transistor is 16kHz. The PWM-pattern of the ANPC generates a ripple frequency of 32kHz for the alternating part of the current. This helps to decrease the filter effort. Depending on the design of the inverter, up to 2MVA can be realised in a control cabinet with a floor space of 80x80cm.

The University of Texas has also developed a medium-voltage inverter based on SiC. The nominal voltage here is 4.16kV with an output of 1MVA. New power plant concepts were also devised in this project in order to increase efficiency at the overall system level.

The current situation and the market

The projects mentioned show that the technological course has been set for the transition to medium voltage. At the same time, the technical innovation potential for photovoltaics in the low-voltage range is heavily saturated. Technological advances and thus unique selling points are only possible to a very limited extent, making price the main selling point. This poses major challenges for European manufacturers in particular.

The market situation and the rising raw material prices in recent years have increased the pressure to innovate. In the meantime, there have been initial advances in the direction of higher voltages. One example is the Mengjiawan project in Yulin City in the Shanxi province of China. Here, a pilot plant with a DC voltage of 2,000V was built and put into operation.

The researchers at Fraunhofer ISE are now convinced that it is no longer a question of whether the technology will take hold, but who will be the first players on the market and thus determine the technology. As already shown above, the global demand for installed PV power will be 63TW in 2050. There is a huge market here. According to a DNV report, the share of utility-scale power plants, and thus the area of application for medium voltage, is between 40% and 60%.

This is a great opportunity for Western manufacturers to regain technological leadership, at least in the area of utility-scale power plants. If a powerful consortium can now be founded, in which all the important suppliers of a utility-scale PV power plant are covered, the remaining hurdles can be tackled together. This approach has several advantages:

  1. The joint approach reduces the risk of developing a product that does not make it onto the market due to a lack of other components.
  2. If the power plant is viewed as a whole with all the required technologies and their strengths and weaknesses, the entire power plant can be optimised. In this way, higher yields can be achieved over the service life.

The issue of standardisation

The problem of the lacking standardisation is now also being addressed. During the MS-LeiKra research project, a standards search was also carried out for the AC side. By drawing on standards from other areas such as electrical machines, all aspects would be covered. What is missing is the transfer of this knowledge into specific product and test standards.

The standardisation committees have now recognised the increased interest in the topic of medium voltage and are shedding light on the subject. As far as the DC side is concerned, a proposed standard for voltages up to 3kV is currently being developed at IEC level. This means a doubling of the maximum 1,500V possible today.

The first test specifications also exist in the field of testing. TÜV Rheinland, for example, has developed first internal test specifications (2PFGs) for PV modules up to 2kV and offers testing in accordance with these.

Whether an increase in voltage to 3kV is sufficient, or whether the economic optimum point is slightly higher, must be determined by further investigations. From the point of view of SiC semiconductors alone, higher voltages could also be implemented in the near future with 6.5kV components.

One frequently mentioned point of criticism of the move to medium voltage is the significantly increased safety requirements and the resulting need for additional training for employees. Those requirements are justified. If you look at the scope of applications of medium voltage, which is part of high voltage internationally, this extends to over 100kV. Historically, the main requirements come from the grid sector. In principle, however, the same physical laws apply directly below and above after the lower limit of 1,500Vdc.

The introduction of a further voltage range of “low medium voltage” could be considered in terms of standards. The exact limits remain to be discussed but could be between 1.5kV and 10kV, for example. This would allow the necessary requirements to be reduced to an appropriate level. With photovoltaics as the only field of application, this would probably be a step too far. However, as considered in more detail in the next section, this lower range of medium voltage is also very interesting for other large fields of application in which similar savings potentials arise.

Applicability to other areas: PV, wind, storage, charging

In addition to utility-scale PV power plants, wind power plants, industrial grids in heavy industry and the charging of electric vehicles, ships and aeroplanes are interesting areas of application for low medium voltage.

In very large industrial plants with large dimensions and high demand for process heat, for example, local renewable energy generators could be connected to large heat pumps using an MVDC bus with lower losses and lower material costs.

In the field of mobility, the electrification of trucks, ships and planes continues to progress. Due to the large battery capacities, high charging capacities are required to enable fast charging. Here, too, simplifications could be achieved by increasing the voltage and thus lowering the current. Lighter cables and plug contacts make the system easier to use. In addition, the current heat losses in cables and plug contacts can be greatly reduced. Even when charging normal electric cars, medium voltage can bring significant potential savings – not directly in the vehicle, but in the underlying infrastructure.

Many new charging parks are needed to enable electromobility. In Germany, there must be a charging point for cars and trucks every 50km along the highways. To meet the necessary demand, the stations will have a grid connection capacity of up to 33MVA, which is equivalent to a small town. To relieve the load on the grids, it makes sense to use a combination of on-site renewable generation and storage. Through the roofing of the parking lot areas, there is very large PV potential on site.

However, the connected load must also be distributed on site and routed to the individual charging points. To achieve a high level of efficiency and limit the use of materials, the move to medium voltage is also a very sensible approach here. As PV, storage and charging technology are DC-based, an MV-DC bus system would probably be the most efficient solution in the future.

The same approach is also conceivable in the area of hybrid power plants. Photovoltaic and wind power plants, batteries and electrolysers could be combined very efficiently with power plant-internal MV-DC bus systems.

As already described, the future costs of various individual components are important. The use of medium voltage in all of the areas mentioned will expand the market further. This means that more significant economies of scale can be achieved in terms of component costs. In addition, a broader field of application provides increased market security for developers of components for this field.


Contrary to all the hurdles in the past, the signs are currently looking good for the introduction of medium voltage in PV. The technical feasibility has been demonstrated several times. The first steps have already been taken by various manufacturers and the first pilot systems are already in operation. The problem has also been recognised on the standardisation side, and the committees are working on the important normative basis for this step towards resource-saving renewable energy generation. At the same time, there are other areas of application where risks can be minimised and costs reduced. What is now missing is a powerful consortium that can jointly remove the remaining hurdles and thus pave the way.

After training as a state-certified electrical engineering technician (2010 to 2012), Michael Geiss studied electrical engineering and information technology at Leipzig University from 2012 to 2017, specializing in energy technology. After completing his master’s degree, he has been working at Fraunhofer ISE in the High Power Electronics and Systems Engineering group. His focus in various research projects is on medium-voltage applications. Since 2019, he has been project manager of the “MS-LeiKra” project, in which the world’s first medium-voltage string inverter for photovoltaics was developed.

    11 March 2025
    Frankfurt, Germany
    The conference will gather the key stakeholders from PV manufacturing, equipment/materials, policy-making and strategy, capital equipment investment and all interested downstream channels and third-party entities. The goal is simple: to map out PV manufacturing out to 2030 and beyond.

    Read Next

    June 10, 2024
    The researchers adds that building a solar cell manufacturing plant in locations with less water availability will be feasible in the future.
    June 5, 2024
    Leapton Energy has launched a series of tunnel oxide passivated contact (TOPCon) modules with a maximum power conversion efficiency of 22.95%.
    June 3, 2024
    FTC Solar has launched its Automated Hail Stow Solution, an automated programme to stow PV modules ahead of hailstorms.
    May 29, 2024
    As the unexpected happens, accurate forecasting becomes a vital tool to sustain grid operations, writes Tigo Energy's Evgeny Finkel.
    May 27, 2024
    PV Tech Power spoke with Lukas Jakisch of PV testing company TÜV Rheinland about the developments and challenges facing the sector.

    Subscribe to Newsletter

    Upcoming Events

    Solar Media Events
    July 2, 2024
    Athens, Greece
    Solar Media Events
    July 9, 2024
    Sands Expo and Convention Centre, Singapore
    Solar Media Events
    September 24, 2024
    Warsaw, Poland
    Solar Media Events
    September 24, 2024
    Singapore, Asia