This article was updated on April 10, 2017, and originally published on Nov. 19, 2016.

It's not that crazy hard. -- Elon Musk

That's exactly what Elon Musk said when discussing the manufacturing of solar roofs during November's investor meeting on Tesla's (TSLA -3.55%) acquisition of SolarCity Corp. The merger went through in the fourth quarter if 2016 as planned, and now the solar roof, which will be made in Buffalo, New York, with the help of Panasonic, is expected in the summer of 2017. 

Musk seems to think creating a new product category in both solar and roofing will be a piece of cake with Solar under Tesla's umbrella. I hope he's right, but reality may present more difficulties than he implies.

Rendering of the solar roof on a home with a Tesla car in the garage.

Rendering of the solar roof. Image source: Tesla.

Solar roofs are the biggest test Tesla has faced

When Tesla Motors began making vehicles, it changed the auto industry as we know it, but it used a lot of existing technology along the way. A Model S looks and feels a lot like a traditional car, the batteries Telsa uses are standard lithium-ion batteries Panasonic was already making, and there's a century of history of building electric motors. Tesla put the pieces of the car together differently, but it didn't reinvent the wheel altogether. 

The solar roof is different. Tesla is using solar cell technology that's not proven (Silevo) and already needs help getting off the ground. To help, it's partnering with Panasonic, a company that shut down one of its largest solar manufacturing plants because the product wasn't cost-effective. The solar roof they're making is headed for a market that has left dozens of companies in bankruptcy -- or with major financial losses. They're using a different (and better) method to make solar roofs than previous companies, but the product isn't proven, hasn't been field tested, and it's not clear who is going to install it or how much it will cost. 

Musk's response to the uphill battle he faces was: "It's not that crazy hard." 

Bold predictions on the solar roof's cost

A lot of the discussion at the shareholder meeting centered on how cost-effective the solar roof would be. Musk kept saying it will be comparable to a "normal roof" in cost, and the energy would essentially be free. 

This would be amazing if true, but Musk later started to compare the solar roof's supply chain to ceramic or concrete tile roofs, which are the most expensive options customers can buy. They can range anywhere from $400 to $2,000 per square (100 square feet). By comparison, a standard asphalt shingle costs about $90 per square to install.

The market for high-end roofs is small, too. Data is a little scattered, but according to the publication Roofing Contractor, all non-asphalt or single-ply residential roofing options account for about 13% of the market in the U.S. More than half of the U.S. roofing market is asphalt shingles, and there's literally no way the solar roof will be competitive with that product on a cost basis. Here's a little math to back up that statement. 

Based on the data sheet from a solar panel that has similar efficiency to what Tesla claims for the Silevo panels, you can fit about 19.6 watts of solar cells in a square foot. The best-case scenario for a Silevo solar cell (not the whole solar roof, but just the cell) is a cost of about $0.30 per watt, based on the price of solar cells today. That means a square of solar cells would cost about $588 before we even consider the cost of manufacturing the roof tile, installing the tiles, wiring to connect tiles, the inverter, and other components and costs that go into roofing and solar systems. 

If we just use SolarCity's installed cost of about $3 per watt, we get an even more alarming cost of $5,880 per square. And I can't imagine the solar roof will be easier to install than a standard solar system, so this is probably a better ballpark figure to use. 

It's possible the solar roof is competitive with the most expensive roof options available to customers, but saying it's competitive with a "normal roof" is misleading at best. 

Remember, we don't even know if this product can be manufactured at a mass scale yet. Of course, "It's not that crazy hard."

A long road ahead

The solar roof is an undoubtably beautiful product, and it could give Tesla a great new product with which to excite customers. But it won't be comparable to the cost of a normal roof, and we don't even know if Tesla can manufacture the product yet. 

As Tesla develops its solar strategy the solar roof will be a key product for the company to bring in customers. But the path to success for the solar roof will be extremely difficult. And it wouldn't be surprising if it's much harder to make than Musk is telling investors today.