That is 3 billion USD wasted on a tech that will be worthless when the fully driverless BEV taxi services are a reality in a few more years. Tesla will not even spend 3 billion USD for combined R&D for model S, X and 3. They spent more than that for their production equipment but not on their R&D.

The old auto-industry will not change strategy until a newcomer like Tesla starts to take serious market share away from the old industry. Hopefully that will happen with model 3.

They may want to own the whole distribution network rather than sharing with Toyota, etc. After all, they will need some revenue stream when auto sales drop by two-thirds.

Improved ICEVs (45 mpg), improved HEVs (65 mpg) and improved PHEVs (100+ empg) will be around for another 20 to 30 years.

Phasing out those three technologies, in favor of a mix of extended range BEVs and FCEVs, will be very progressive.

Improved (3X to 5X) lower cost batteries and much Lower cost H2 and distribution networks are required.

@Roger Pham

We (in the United States) do not have grid excess renewable energy. Last year about, we got about 13.5% of our electric energy from "renewables" which was mostly hydro electric. Only about 4% was wind and less than 0.6% solar. It is far better to replace some electric power that would be generated by burning coal than trying to make hydrogen by electrolysis. I do not know why you persist with the excess renewable energy to hydrogen fantasy.

Internal combustion engines will be with us for quite a while -- especially diesel engines for high continuous load applications such as trucks, tractors, construction equipment, etc.

Here's another way to look at it.
Mercedes are bringing out new generations of of diesel and gasoline engines that will be more efficient and cleaner than the previous generations.

This is possible because Mercedes are premium cars and the customers pay more for them, thus they have the margin to afford the extra money it takes to develop and deploy proper aftertreatment for diesel and GDI petrol engines.
It will also benefit many commercial trucks using Mercedes engines.
I don't think that is too bad.

@sd, that could change reasonably quickly.

Renewables alone powered 90% of German electricity for several hours on a recent Sunday this year, and the entire nation of Portugal was also powered completely by renewables for 4 days in a row this year (i.e. not dipping below 100% of demand).

The economics of utility scale solar, now at <3 cents per kWh without any subsidies, must surely push even the US to a tipping point some time soon.

@Clett, there is nothing too great about getting 90% of your electricity when the sun shines and the wind blows, the question is where do you get it when the sun does not shine and the air is still. Take any still night/evening as an example.
You'll still have to keep a dispatchable source of power, typically thermal or hydro.
So what to do with the excess power on your sunny Sunday afternoons (sounds like a Kinks song).
Ideally, you would store it all in batteries or hydro storage, but this would be too expensive and there just aren't enough battery factories to make them all.
IMO, you are better storing it in batteries (ideally batteries in EVs) than h2, unless, you use the H2 straight away for some chemical process, like cracking heavy oil.
Making H2 is bad enough, but storing it and transporting it is even worse.

Agreed, storage will be the key test of the Energiewende. In my view, batteries will play only a small part in this, buffering supply for perhaps several hours or so once the EV fleet is up and running. The requirement to buffer week-to-week and interseasonal supply will depend on alternative solutions.

I think the requirement for longer term storage will most likely be met largely by power-to-gas and large-scale thermal storage. Crucially, both will make use of the existing powerplant infrastructure - natural gas plants will be run on e-gas and coal plants will be replumbed to serve as thermal stores, and occasionally burn biomass. I think this will be crucial to win the acceptance of the incumbent generators, who would otherwise be forced to retire large capital investments early. The inherent round-trip inefficiency of these processes won't matter too much when only one in three kWh supplied comes from storage. The costs should also be manageable since the gluts of power from solar and offshore wind, which are soon to be frequent and enormous, already have a spot price of less than zero. There's some more discussion of this in chapter 10 of "Science for the worried: An optimist's guide to the future" for anyone interested in this topic.

HarveyD, clett, et al

I do not know where you got your data from. Cherry picking highly select sources? Look at https://en.wikipedia.org/wiki/Cost_of_electricity_by_source

Natural gas combined cycle is consistently the lowest cost while photo-voltaic is consistently the highest and wind and new nuclear are projected to be about equal with on-shore wind a little lower and off-shore considerably higher. But this does not take into account that you need backup power for either wind or solar.

But my real comment was to replace burning coal with what renewable power we do have and not use use it for the hydrogen fantasy.

Also, when you quote some of the European data on renewable power consider that some of this comes from burning peat which in my opinion is worse than burning coal and some comes from wood pellets imported from the US.

Clett

The data is for Dubai which certainly has sunny skies and who knows what the cost of land and labor was along with the cost of finance. The lowest cost in the US was about twice that.

One way for USA to lower local solar power plants (below $0.03/kWh?) would be to:

1) import very large quantity of high performance lower cost PV panels from Asia (mainly from China)

2) use much lower cost Mexican installation labor (before the wall is built) much the same way as many other infrastructures are currently built in USA.

3) borrow $$$B, at very low rates from the GREEN FUNDS and/or from the Federal Bank.

4) develop lower cost ways to store excess REs, to the level required for 24/7 service and/or find ways to match consumption with production with BEVs quick charge facilities nd/or H2 stations.

The maths supports even lower prices soon. Modules are now selling for 50 cents per watt, which means 1 kW of modules costs about $35 per year, amortised over 25 years at 5% interest.

That means the "module cost" of delivered electricity is just 1.6 cents per kWh in a sunny location like Dubai (6 kWh per kW per day incident). The Germans have shown us that the other costs (labour, connections, inverters etc) can amount to much less than the module cost.