It won’t be Tesla or SolarCity that runs much of the solar panel factory the state is building in South Buffalo.
As the $900 million RiverBend factory gears up to begin production this summer, Panasonic will play a large role at the plant as Tesla’s partner in producing solar cells and panels and as an experienced solar panel manufacturer in its own right.
Tesla will be Panasonic’s customer, agreeing to buy most of the solar panels and solar products that the Japanese company churns out of the Buffalo factory for the next 10 years.
And it will be largely Panasonic solar panel technology that is used to create the solar panels and solar roof shingles that the factory will produce, not technology from Silevo, the California start-up whose high-efficiency designs first attracted the interest of state officials looking to make a splash by attracting a clean energy company to the region as part of the Buffalo Billion.
"Tesla will oversee factory operations in Buffalo, and will manufacture solar roof tiles there," a SolarCity spokesperson said Tuesday. "Panasonic will manufacture solar cells at the factory, with support from the Silevo team, including cells for the solar roof tiles that will be a hybrid of the Panasonic and Silevo architecture. Panasonic will also manufacture solar panels in Buffalo."
The new arrangements reflect deep changes that have taken place, both at the company and within the solar marketplace, since SolarCity first bought Silevo nearly three years ago. The way the solar panel factory will operate has been in flux ever since Tesla said that it planned to partner with Panasonic in the operation of the Buffalo gigafactory in the months leading up to its November acquisition of SolarCity.
Despite the changes, Tesla still is pledging to create 1,460 jobs in the Buffalo Niagara region and attract another 1,440 from suppliers and service providers, with a target of hitting full production by 2019. It’s still working with state officials to finalize orders for some of the factory’s manufacturing equipment, which the Cuomo administration has agreed to pay for as part of its $750 million in subsidies it is providing to lure SolarCity to Buffalo.
Yet documents filed with the Securities and Exchange Commission make it clear that the operation of the 1 million-square-foot factory on South Park Avenue will be very different from what was originally planned.
Solar growth slowed
It was Silevo’s technology – and its potential to lower costs by requiring fewer solar panels on a typical installation – that caught the eye of SolarCity, which acquired Silevo in 2014 and expanded the size of the Buffalo factory to five times its original size. They initially saw Silevo’s high-efficiency technology as a way to make rooftop solar more affordable, and more competitive with conventional sources of electricity produced from fossil fuels. Not anymore.
Back in 2014, SolarCity’s vision was built around expectations that its residential solar installation business would keep growing by leaps and bounds as consumers flocked to its offer to install solar panels on their roofs at no upfront cost.
But that vision also was built around the idea that SolarCity would be able to keep raising billions of dollars in new money to pay for the upfront cost of installing all those new solar arrays. The idea was that all the money SolarCity borrowed would be repaid over time by the modest monthly payments the company would receive from each customer based on the electricity produced by its solar panels.
That wasn’t how it worked out. Growth within the solar marketplace slowed, rather than accelerated. Loans that allowed consumers to buy their solar panels outright gained popularity over the leases that were the backbone of SolarCity’s business.
Along the way, investors grew alarmed that the losses that SolarCity’s leasing model produced – nearly $1.6 billion over the past two years – would be too much of a burden, forcing SolarCity into the arms of Tesla. Tesla CEO Elon Musk, who also was SolarCity’s chairman and the cousin of its CEO, touted the acquisition as a way of building a renewable energy powerhouse that would sell electric vehicles, solar energy systems and the batteries that both powered the cars and stored the electricity generated by the solar panels.
Yet it also was clear that SolarCity’s business needed fixing. Instead of growing, SolarCity deployed 26 percent less generating capacity in the fourth-quarter than it did a year ago. For all of last year, it deployed 845 megawatts of solar generating capacity, less than the 900 megawatts it predicted as recently as November and far less than the 1,200 megawatts it forecast a year ago.
In response, SolarCity has been cutting costs. It eliminated a little more than 3,000 jobs over the past year – about 20 percent of its workforce – and trimmed its sales and marketing spending by 3 percent, while reducing its general and administrative expenses by 9 percent. It cut research and development spending by 15 percent. Musk hopes to trim SolarCity’s stubbornly high customer acquisition costs by selling rooftop solar in Tesla’s stores.
RiverBend on schedule
The Panasonic deal will have other financial benefits for SolarCity, freeing it from much of the burden of making future investments at the Buffalo factory by shifting that responsibility to its partner. Tesla already has a partnership with Panasonic to make battery cells at Tesla’s battery gigafactory in Nevada.
Panasonic will operate much of the factory at RiverBend in Buffalo, not Tesla, which has never made solar panels and already has enormous operational challenges on its plate in the opening of its battery gigafactory in Nevada and the roll out of its mass market Model 3 sedan.
“We anticipate significantly reduced expenditures related to manufacturing operations at the RiverBend manufacturing facility,” SolarCity said in a Securities and Exchange Commission filing. SolarCity executives also said they still hope to increase the capacity of the Buffalo factory beyond 1,000 megawatts by tweaking the layout of the factory and modifying some of the production equipment.
“We believe that the short delay in finalizing the overall configuration will help maintain the long-term competitiveness of the manufacturing facility,” SolarCity said in the filing. "However, this is an aggressive schedule, and we may experience additional delays.”
The filings make it clear that the Buffalo factory won’t rely on Silevo’s unproven “Triex” solar panel technology, which SolarCity executives had hoped would be able to convert 20 percent to 21 percent of the sun’s energy into electricity.
Instead, the factory will be based on Panasonic’s high-efficiency technology, which already is in use, and some elements of Silevo's technology. That’s far less risky than basing operations at the biggest solar panel factory in North America around a Silevo technology that has never been produced in large volume.
The change will save SolarCity $84 million because the shift away from Silevo’s technology means that it won’t owe Silevo’s former owners any additional money. The shift to Panasonic means Silevo won’t hit two of the three performance targets spelled out in its purchase agreement with SolarCity that would have triggered additional payments to Silevo’s former owners. Solar panel production at Silevo’s small factory in China is being wound down because of tariffs that the U.S. government imposed on imported solar panels last year.
So the RiverBend factory remains on schedule. But the way it runs is going to be very different from the original plan.