Self-Driving Car Liability Limits, Fuel Cells Following Solar’s Footsteps … (Cleantech Talk 43)

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

---

Episode #43 of Cleantech Talk is here! This episode covers accident liability for self-driving cars, the EV tax credit, and fuel cells’ exponential ramp.

You can listen to this episode and subscribe to Cleantech Talk on iTunes or Soundcloud. You can also listen via the embedded player above or download this episode. Jump into the show notes below for more goodies.

We started off noting the Thomas Fire in California, which continues to burn through late December, and hoping our listeners in the area remain safe. It’s always seemed odd not to simultaneously wish that listeners in other areas remain safe, or wish that non-listeners remain safe too, so we’ll go for the all-of-the-above approach, even if it turns a specific hope into a hopelessly non-specific one.

Eric Holthaus has written two excellent reports for Grist (here and here) about how climate change has set California up for a year-round fire season, not in the years 2217 or 2117 but in 2017. It’s sobering stuff. So I’ll round this out with some hopeful news that (California-based!) research on ever-cheaper sensors and drones could help us counter this challenge. While fire has a place in the cycles of nature, we may be able to use technology to detect fires sooner, allowing us to make better-informed decisions about which ones to allow to burn, and which to extinguish before they become dangerous.

Liability Limits for Self-Driving Cars

Nicolas noted in a mid-month article that the state of California decided that if autonomous vehicles crashed, the automakers could be held liable, even if the AV system wasn’t maintained as recommended. This isn’t unexpected — most product liability laws work in this way. If you get electrocuted by your toaster, the company is still on the hook even if you didn’t de-crumb it on a regular basis.

GM: the leader in self-driving cars?

In North America (and probably elsewhere), the legal precedent is that if you make a product, you’re liable for consumer injury even in the event of misuse or modification. This is the overriding reason why GM crushed the EV1. It didn’t want some would-be MacGyver injuring himself (or worse, his kids; or even worse, someone else’s kids) if he tried to DIY some upgrades.

We didn’t have a chance to discuss the possible legal liabilities of a neural network approach to AVs. The problem with neural networks is that the computer “learning” is a black box; you can’t point to lines of code that say “this is how we programmed our system to do X.” Neural networks are powerful, but if they’re built using non-representative data, they can be powerfully wrong. I strongly suspect that the first major OEM rollout of AV technology will involve geofencing to specific routes (perhaps highways) and will only come on if enough of the sensors watching the road and the driver are functioning properly.

EV Tax Credits Are Safe. Yay!

The EV tax credit in the United States was saved. That’s fantastic news for the pro-electric vehicle coalition, but probably bad news for Tesla in the medium term, because it means that at some point in the future, it’ll have a price disadvantage of up to $7,500 against the competition.

How Plug-in America celebrated the survival of the U.S. EV tax credit

To quickly recap the tax credit legislation currently in place, for one quarter after the one in which the automaker has reached 200,000 US EV sales, its EVs will qualify for up to $7,500 in tax credits; the maximum credit is then $3,750 for the next two quarters, $1,875 for the two quarters after that, and then falls to $0.

Think about Toyota, the carmaker EV advocates love to hate. Over the years, it has sold a bit fewer than 70,000 tax credit–qualifying vehicles (those being EVs, PHEVs, and fuel cell vehicles), with perhaps 25,000 of those coming this year. Suppose Toyota sells a bit more than 100,000 in the next 3 years. That would bring the company up to 175,000.

In 2021 — roughly when Toyota thinks Panasonic will be mass producing its solid-state batteries — Toyota could sell 25,000 fuel cell and/or breakthrough solid-state battery-based EVs with a full $7,500 tax credit at a time when Tesla’s tax credits will have probably ramped down to $0, while the tax credits for GM and Nissan (and perhaps others) will also be ramping down. EV volumes in 2021 will be higher than they are today, which means that Toyota will probably also sell more EVs during its six-quarter ramp-down of incentives than Tesla will be selling. That’s an enormous competitive advantage, and while Toyota is an extreme case, every other automaker will have this pricing advantage over Tesla for some time. That’s never good for a business.

So, in this sense, while there would have been a bit of turbulence at Tesla if the EV tax credit had been cancelled, it might have actually worked in Tesla’s favour — especially since its base is probably less price sensitive than the folks buying EVs from the world’s less-inspiring auto brands!

Of course, that’s just a mental exercise. Hypothesizing of second-order effects aside, it’s a great thing the EV tax credit survived!

Fuel Cells Today = Solar in 2002 = Wind in 1994

I contributed to the Fuel Cell Industry Review this year (free download here). The quick summary is that fuel cells are currently where solar was in 2002 and wind in 1994. This applies both to annual shipments in MW and to growth rate trends.

As such, I think dismissing fuel cells right now is short-sighted. It’s like dissing solar 15 years ago or wind 25 years ago. At the time, each technology was a rounding error, but a steady exponential ramp based on learning curve effects made wind and solar significant. In time, the same will happen for fuel cells, too. You can get on the train early, or get on the train late!

One knock against fuel cells is that they’re less efficient than batteries at turning renewable energy into forward motion. That’s true, but focusing too much on efficiency can be counterproductive. Recent research shows that men — American men, at least — perceive “green” products as effeminate and unmanly. Which means that appeals to choose EVs over fuel cell vehicles for environmental reasons … could backfire. This is another reason for Tesla’s success: it grew the market of EV buyers beyond the “eco” demographic.

The bigger objection most people have to fuel cells is that most hydrogen is generated from natural gas. That’s true for now, just as it’s true that more than half the world’s electricity is still generated from fossil sources. It’s well known that the declining costs of wind and solar will displace ever more fossil power from the grid (and not a moment too soon) — but it’s less well known that the declining costs of wind and solar (aided by carbon prices and clean fuel standards) will start to displace ever more natural gas from hydrogen generation. That transition will be slower, but it’s just as inevitable. (I’m happy to report that an old CleanTechnica post of mine actually looks foresighted for once!)

Mexico’s recent utility-scale solar auction winners proposed to sell their electricity at an average 2.1 cents/kWh (one bid came in at 1.77 cents/kWh). And earlier this year, the head of research at Engie predicted that in sunny areas, solar could hit 1 cent/kWh by 2025. Unsurprisingly, this makes him hugely bullish about hydrogen, which he thinks could become a cheaper fuel per energy unit than LNG. This is fantastic, because fugitive methane emissions can cancel the climate benefits of natural gas over coal, or make it worse.

As further evidence, Shell recently partnered with electrolyzer company ITM Power to test a 10 MW electrolyzer system to supply hydrogen for a massive petrochemicals complex in northern Germany. (ITM Power claims it has already hit the U.S. DOE’s year-2020 cost targets for electrolyzers.)

A 10 MW electrolyzer would consume 10 MWh per hour, and assuming 50 kWh per kg of hydrogen (to make the math easy), that means it will produce 200 kg, or 0.2 tonnes, of hydrogen per hour. Given that there are 8760 hours per year, that makes for 1750 tonnes per year, which is roughly 1% of the plant’s annual consumption of 180,000 tonnes. If the technology proves out, you can bet Shell will bring a lot more online, if for no other reason than regulators are looking at the aforementioned clean fuel standards (which require a reduction in fuels’ carbon intensity). Renewable hydrogen can help them move in the necessary direction.

Separately, Norwegian electrolyzer company Nel Hydrogen (which is partnering with SunPower for solar-powered electrolysis in California, as the state requires that 33% of the hydrogen fuel be produced from renewable resources) has divulged that they’re doing early design work for a 400 MW electrolyzer installation with a major industrial partner, specifically to undercut fossil hydrogen. That’s as cool as it is big!

As for learning curves, BNEF has assumed a 19% learning rate for batteries and wind in its latest “State of the Industry” presentation, with a roughly 25% learning rate for solar. And Plug Power’s announcement of a 25% learning curve for its fuel cells was from a November presentation to investment analysts.

It all adds up to fuel cells being worth paying attention to going forward. They’ll be a niche technology — as I mentioned in the podcast, maybe a “special teams” player in American football, or a baseball “role player” — but they’ll be part of the solution nonetheless. Seeing as there are so many people already covering efficiency, wind, solar, and storage, we’re happy to report on the findings they uncover. For those who caught on to wind and solar a bit late, and who want to track the birth and infancy of a newer climate solution, fuel cells are one of the places to be, and we’ll be delighted to highlight the milestones along the way. That said, we’ve been a bit fuel cell heavy as of late, so we’ll make sure to shine a well-deserved spotlight on some other sectors for a while!

Best wishes to everyone for 2018 — and for those of you wanting to reach out on Twitter, Nicolas is @ElectricExaminr and Matthew is @ElectronComm!