Higher voltage, lower cost


A mere three years ago the solar benchmark for electrical potential was 600 volts DC. Then 1,000 VDC systems came of age, and now in less than three years, 1,500 VDC will be the new notch carved into the bench. Sooner rather than later, the benchmark could hit 2,000 VDC. Yet, despite the rapid U.S. development of new technology in solar, the time required for product certification and regulation inevitably delays the adoption of the next technological phase.
This year, the slowly accelerating march of higher voltage toward the 1,500 V norm is being regarded by some engineering, procurement and construction (EPC) companies as a promising new means of cost reduction and higher yield. First movers gain market share. Reduced numbers of inverters and shorter runs of wire save balance of system (BOS) costs, and higher voltage increases DC yield.
“We are seeing an upward trend in 1,500 V interest by EPCs, but I can’t point to any large projects yet. If you look at the volume of 1,500 V systems installed in the U.S. today, it is minuscule. But by the end of 2016, the volume of those systems that we will be installing will far surpass everything else we do in the world,” reckons Mahesh Morjaria, Vice President of Systems Development for First Solar, the Tempe, Arizona-based manufacturer that is also ranked as the second largest EPC in the U.S.
However, for many other EPCs, particularly smaller players, a wait-and-see attitude may be the typical response to 1,500 V adoption over the next two years. After all, the U.S. utility-scale market is still booming, and expected to continue until the last stroke of the clock at the end of 2016, when the existing federal income tax credit will expire. However, should that tax credit not be reintroduced in at least a diminished form, even a small technological advantage – like the estimated 3% total system cost/savings offered by 1,500 V adoption – may be critical to survival.

Advantages of 1,500 V systems

The advantages of a large 1,500 V system over a 1,000 V system seem to be a no-brainer. “From an EPC standpoint, with 1,500 V systems, the impetus is that they can reduce installation time, reduce string size, reduce wire costs, and potentially reduce other costs,” suggests Brian Grenko, the CEO of newly-formed PV test engineering firm Amplify Energy, based in San Francisco.
More specifically, “Raising the system voltage to 1,500 volts allows for 50% longer strings, which eliminates 33% of the combiners and wiring in a system,” said Mark Kanjorski, the Director of Marketing at Seattle-based Ampt, in a session presentation at Intersolar North America in July. “This also enables inverter manufacturers to increase the rated output power of inverters by 10 to 40%,” he calculated. “However, to do this, all of the system components will need to be rated at 1,500 volts, which means there will be a delay before these components are widely available, and they are also likely to cost more on a per unit basis,” said Kanjorski. “Even so, the cost per watt savings potential will likely drive higher voltage systems to be adopted over time,” he concluded.

Early U.S. movers follow Europe

U.S. EPCs have not kept pace with their European counterparts in the move to 1,500 V. “In Europe today more utility PV is being installed at 1,500 V than at 1,000 V,” observes Grenko.
In one recent installation, EPC Belectric with GE and Kofler Energies, installed a hybrid 1,500 V PV and combined heat and power (CHP) plant at a GE roof site in Berlin. The facility is being touted as “the world’s first roof-mounted 1,500 V solar PV system” by Kolitzheim-based Belectric, which also has a facility in Newark, California. The 600 kW hybrid power plant, developed by Kofler Energies and GE’s planning office BLS Energieplan, combines PV with natural gas-fired Jenbacher CHP technology and a battery solution. A Belectric announcement noted that “the system is based on a scalable business model that also can be adapted for larger units.” Belectric has been installing 1,500 V PV systems since 2012, it notes.
One large U.S.-based EPC that has been early to move into 1,500 V systems is First Solar. In March 2014, First Solar announced that it had partnered with GE’s Power Conversion unit to develop a more cost effective and productive utility-scale PV power plant design, combining First Solar’s thin film CdTe modules with GE’s ProSolar 1500 volt inverter/transformer station system.
“We have developed a U.S. utility platform based on GE’s 4 MW 1,500 V inverter, the largest in the industry, where the benefit comes in typically, and it makes more sense,” says Morjaria. He previously announced First Solar plans in December 2014 to install 1,500 MW of 1,500 V large-scale solar PV arrays in the U.S. over the following two years, based on its thin film glass-on-glass panel. The company now operates two 1,500 V test plants and intends to build more this year and in 2016, Morjaria says.
“The use of 1,500 V architecture in larger PV arrays and inverters results in lower installation and maintenance costs, since it requires fewer inverter stations to convert the plant power from DC to alternating current AC power,” Morjaria adds. “In the case of our Series 4 FS CdTe thin film modules, 15 modules are normally xAdvertisement

At a glance

  • Not long after 1,000 VDC systems rose to prominence, 1,500 V components are now entering the field.
  • The cost reduction benefits of higher voltages are well known, but few EPCs have begun installing 1,500 VDC systems in the U.S.
  • Europe has led the way in this transition, with First Solar at the forefront of the U.S. charge into this space.
  • Conflict with NEC building codes in the U.S. has served to restrict the adoption of 1,500 VDC components in rooftop markets.
  • Many 1,500 V components are not broadly UL-listed either, although the industry expects that to change in the coming years.
  • Alternatives to 1,500 V components include DC-DC power optimizers and 2,000 V bipolar applications.

connected in series per string as opposed to 10 modules in the case of 1,000 V. This reduces the number of strings (and BOS cost) for the same amount of power,” he explains. “The 1,500 VDC also enables increase in the power of the inverters (which are typically current limited) and thus an increase in DC voltage; the use of large-scale 4 MVA inverters leads to a significant increase in the DC array size,” Morjaria calculates. “Thus, the typical 1,500 VDC system design will incorporate between 5 and 5.3 MWp DC of PV capacity, corresponding to a DC/AC ratio between 1.25 to 1.33,” he concludes.
Off the U.S. utility fields, 1,500 V may take longer to adopt. “U.S. rooftop applications of 1,500 V will get more conflicted with National Electrical Code (NEC) codes for buildings,” Morjaria predicted.

EPCs await market demand

Smaller EPCs are aware of the 1,500 V campaign, but not yet pressured to become involved. “Some of our larger supply partners like SunEdison (the seventh largest U.S. EPC) are more up to speed on it, and there is a lot of buzz in the solar community that it has the potential to drive down costs, maybe a nickel a watt,” says R. Stanley Allen, the CEO of SolAmerica Energy, a regional EPC based in Atlanta. “Still, it will be a couple of years before that becomes the reality,” he adds.
SolAmerica has performed most of its work in the U.S. Southeast, but is moving north into other Atlantic Coast markets, he noted. The company, which opened an office in Washington, DC in July, has a base of about 12 MW installed, he says. Similarly, Borrego, the 30th largest EPC in the U.S., has not yet installed a 1,500 V system, but expects to soon. “1,500 V systems are not broadly UL listed, nor are there proper provisions in NEC to cover non-utility owned systems,” says Senior Project Engineer Randy Batchelor, based in Oakland. “It may take one to three years – perhaps under the 2017 Code cycle – before UL-listed products become sufficiently available. That will have to take place before we will be comfortable with it from a code perspective and perspective of the permitting authority having jurisdiction (AHJ),” he notes. “However, it would enable Borrego to gain market share if we are able to adopt the new voltage class sooner than other EPCs,” he observes. “Certainly the 1,500 V systems will enable some savings for EPCs that will be passed on to the customer,” he adds.

1,500 V component entrants

With new codes emerging chapter by chapter, 1,500 V components are hitting the show floors with increased frequency and much ado. “The recent announcement by UL to adopt ANSI/UL 62109-1 as the American National Standard for Safety of Power Converters for Use in Photovoltaic Power Systems enables U.S.-based certification of 1,500 VDC inverters, and is a step in the right direction,” Morjaria notes.
Among panel makers, Trina’s 1,500 V Duo-Max won special recognition at the MIREC conference this past May in Mexico for its innovative glass-on-glass technology. Shanghai-based JA Solar also released a 1,500 V panel in March of this year. Yingli won wide recognition for its metal-framed 1,500 V panel, introduced at Intersolar North America in July. And Irvine, California-based S-Energy also unveiled its new UL-certified 1,500 V panel at Intersolar – the company calculates that the panel will reduce the amount of total PV electrical materials by 14%, leading to a 10% decrease in the BOS cost.
Among inverter makers supporting the 1,500 V solution other than GE, Power Electronics in May unveiled its new utility-scale HEC1500V inverter with capacity ranging from 1 MW to 3 MW. Then, at the Intersolar North America exhibition in San Francisco in July, KACO announced its blueplanet 1500 TL3 with Ampt Mode, a 1,500 kVA transformer-less solar inverter with protection class IP 54/NEMA 3R for outdoor use, notes Ben Castillo, a technical sales and marketing lead for the Rocklin, CA-based company. The inverter is available as part of a 3,000 kVA integrated power station. Used in combination with Ampt String Optimizers, the blueplanet 1500 TL3 achieves a 50% increase in rated output power, lowering the specific cost of a system inverter solution by 33%, the company claims.
And among BoS suppliers moving into 1,500 V territory, ahead of the pack is ABB, which in October 2014 introduced a new line of 1,500 V disconnect switches, molded case switches, contactors, surge protection devices, and voltage/current sensors. Some of these components are rated up to 3,000 amps at 1,500 VDC and carry various UL and/or IEC certifications, the company indicates.
Similarly, Shoals Technologies Group, based in Portland, Tennessee, unveiled its 1500 V SlimLine Combiner Box at Intersolar NA, the latest of a four year old line of 1,500 V PV products.

U.S. regulatory morass

It can take months, if not years, for a PV component manufacturer to gain certification for its product, once national codes have been set. Comprehensive testing of new components naturally precedes both steps. In the case of 1,500 V products, little is yet known about performance. “Both 1,500 V modules and inverters have limited field history. The regulatory challenges in many places, particularly outside of North America, are lower due to existing International Electrotechnical Commission (IEC) standards that (already) address 1,500 VDC design and safety,” says Morjaria.
“Greater challenges are faced in the U.S., where UL standards and interactions with the NEC are not fully aligned xAdvertisementto IEC standards, accepted both in the U.S. and around the world,” Morjaria continues. “So the lack of alignment in standards that address 1,500 VDC applications often make it challenging and confusing to obtain plant construction permits from local authority having jurisdictions,” he concludes.

Uncertain performance from PID

One concern with the performance of 1,500 V components regards potential induced degradation (PID), in which ions from the cells migrate through insulation into the panel frame and beyond. “Voltage is one of the biggest influences of reliability and durability, and it is yet to fully be seen what the effect of 1,000 V will be on panels,” reckons Grenko. Clearly, panel resistance to PID will become increasingly critical as array voltages increase to 1,500 and beyond, at least for framed panels. Some electrical engineers postulate that in arrays of 1,500 V or more, the higher positive potential is capable of causing new failure mechanisms. A special coating used by Yingli for its 1,500 V panel “ensures high resistance to potential induced degradation,” according to a statement by Yingli Green Energy CEO and Chairman Liansheng Miao at the time of the release.

Future voltage bumps

There may be more than one way to skin the voltage cat. Prior to its July exit from the solar inverter business, a white paper by Advanced Energy, of Bend, Oregon, observed that “1,000 V is still the standard, but 2,000 V bipolar is an alternate and viable option, since it provides a 1,000 V listing with a 2,000 V collection capability. This is the best near-term path to meeting the requirements of both large utility plants and smaller commercial projects with readily available product and code compliance.” Another potential avenue of voltage gain might be from the use of distributed power conversion (DPC), suggests Kanjorski. “At the same time these new 1,500 volt architectures are being considered and deployed, there are 1,000 volt architectures using distributed power conversion (DPC) approaches that offer similar advantages to 1,500 V systems,” he says. “String-level power optimizers are DC-DC converters that manage power out in the array.
With string optimizers, system designers can build strings that have twice the number of modules, which cuts the number of combiners, disconnects and homeruns in half,” Kanjorski said. And importantly Kanjorski added, “the longer strings enable a higher inverter input voltage, which supports an increase in the inverter’s rated output power by 40% to 70%.” Kanjorski went on, “This can all be done today using 1,000 volt system components, so the industry doesn’t have to wait for higher voltage components to become available.” One key to DPC gains is MPPT decentralization, he pointed out. “With string (DPC) optimizers, you get MPP tracking on every 10 modules, which is 50 to 300 times more granular than typical central inverters,” he said.
The bottom line message to take from all this buzz seems to suggest that we toss out our old voltage meters that stop at 1,000 V and start shopping again.

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.