[sien]

This is a good reason why Tesla cars are folly.

This is why electric vehicles won't work. I'll copy in most of a comment I made a few days back.

Let's say tomorrow some grad student gets fusion going at a very low price. The best way to use this to power cars would be to use it to create a fuel with a high energy density. If you had 'free energy' you'd extract C02 from the atmosphere and turn it into a hydrocarbon.

For more info look at:

http://en.wikipedia.org/wiki/Energy_density

and this interview with Nobel Prize winning Physicist Robert Laughlin

http://www.econtalk.org/archives/2010/08/laughlin_on_the.htm....

the key quote is: "The ones that are technically trained get it right away: hydrocarbons, which we burned today have the greatest energy density possible of all fuels. Things that have carbon in them. Will people fly airplanes? Usually people say yes for the same reasons. Well, how are you going to make the airplanes fly? Battery. Batteries are pretty heavy. Oh--you can't have airplanes unless you have hydrocarbon fuels. You could in theory do it with hydrogen, but it's highly dangerous, noxious fuel. Quantum-mechanically, we know the energy content of those fuels is optimal. There will never be anything that beats them." A massive breakthrough in energy density for batteries might be possible but it's unlikely. Huge resources have been put into improving batteries and while they have improved it's not been enough to get near the energy density of hydrocarbons.

[ta12121]

No.

Electrolysing water to make H2 and extracting CO2 from the environment, and then synthesizing hydrocarbons from them, is extremely energy inefficient.

Sure, theoretically, it can be done, and maybe one day it could even be done efficiently. But Tesla is making battery-electric cars that work today.

The military does not care about efficiency because they have nuclear reactors on their ships and their goal is to not to have to transport liquid fuel.

[sien]

If they can hit $3 / gallon why does it matter?

Tesla's cars are wildly price inefficient because they cost $100K. For 10K you can get a decent car that has 4 times the range. http://en.wikipedia.org/wiki/Tesla_Roadster

How many gallons of fuel can you get for 90K?

They are nice toys for rich people. Perhaps there is a market for that. Good luck to them.

But for mass transit unless you can get the price of the batteries to plummet it just does not work.

[ta12121]

$3 a gallon is what the fuel costs today when it comes straight out of the ground. Claiming to be able to build and run a nuclear reactor and then synthesize the fuel through multiple energy inefficient steps all for the same price is a pipe dream.

[danielharan]

It's not competing against the price straight out of the ground; it's competing against fuel that's been refined and delivered to a moving ship somewhere potentially very far away. Instead of $3 a gallon, it could be up to 10X more.

[antidoh]

And, dollar cost of the fuel is not the only cost of maintaining a "long supply tail." There's also the dollar cost of all the ships and sailors on that tail, and there's the logistical opportunity costs: we have to protect that tail with military resources that might otherwise have gone to more directly military purposes, and the head of the tail (the military activity at the front) is held back and slowed down by a "heavier" supply tail.

This doesn't eliminate the tail (we still have to deliver ammunition, lubricating oil, food etc. to the fleet) and it doesn't eliminate replenishment at sea (we still have to get that stuff from supply vehicles to ships), but it does lighten the supply load and create more military options.

[danielharan]

Exactly. Less replenishment means more flexibility.

Lots of people killed by IEDs on long supply lines in Afghanistan is an extreme example of the human and military costs. Some of those died to fuel A/C for uninsulated tents... madness.

[ta12121]

The amount of energy needed to refine crude oil is much less than the energy it takes to synthesize it. Think about it: people refine the crude, transport it, and use the end product in an energy net positive manner. To synthesize hydrocarbon fuel, you need to put in at least the amount of energy you're going to store, and with current technology probably 2 times as much at least due to inefficiencies. Then you need to still account for building and running the nuclear reactor. Try to imagine the systems as a whole, the number of steps and losses at each step. The scheme does not make sense.

1) uranium -> nuclear power -> expensive synthesis -> local transport -> fuel

2) uranium -> nuclear power -> transmission lines -> charging batteries

3) crude oil -> refinement -> transport -> fuel

EDIT: Found a reference from another comment: http://en.wikipedia.org/wiki/Fischer%E2%80%93Tropsch_process... It says that the input energy is 2-4 times the stored energy for synthesized hydrocarbon fuel.

[danielharan]

Asking me to imagine a system as a whole doesn't prove your argument.

Instead of the energy system, consider the cost of military supply lines. There's more than the financial cost of delivery; long supply lines are vulnerable to attack and disruption. You don't need to imagine an example: consider IEDs in Afghanistan. Many of those were trucks delivering food and fuel to bases. Efficiency (i.e. insulating tents so less fuel is needed for A/C) results in less deliveries and less deaths.

The same principle applies at sea. Oil ships are a vulnerability and another thing to plan, as well as a major cost that can be more important than the energy efficiency issues.

[wpietri]

You're cheating.

Telsa, like many startups, is pursuing a low-volume, price-insensitive market first. That helps pay for all sorts of startup and R&D costs, and lets them work out issues before they scale. Their long-term goal is, per Wikipedia, "eventually mass producing fully electric cars at a price affordable to the average consumer".

There are several fully electric vehicles in production that cost $35k.

http://en.wikipedia.org/wiki/List_of_production_battery_elec...

They have a substantially lower cost of operation than gas-powered vehicles, so they are plausibly cheaper to operate over the lifetime of the car:

http://en.wikipedia.org/wiki/Nissan_Leaf#Operating_costs

And over the long term, expecting batteries to get better is a pretty good bet. Everybody from Apple to Toyota is eager for battery improvements.

[wisty]

Telsa cars are expensive for the same reason Ferraris are - they target people who want bling.

A mass-produced commuter could be much cheaper, but there's a huge risk. Tesla aims to reduce the risk by validating the market, solving technical problems, creating infrastructure (recharge stations) and solving social problems (convincing people that range doesn't always matter, and educating them on the strengths and weaknesses of electric cars).

A low total-cost (including fuel and maintenance) electric car can probably be produced, but it's not the Tesla's top model.

[rhizome]

In non-currency terms, this sounds a lot like Unix-Windows arguments 15 years ago. If you can bundle Windows, why does it matter what computer you get? For 1/10 the price you can get a great Windows PC with 10 times more software. Tesla Roadster being like an SGI or Solaris box, Linux/FreeBSD is just for snobby nerds who like to tinker. Unless you can make it easier to use it just does not work.

[jnw2]

Mass transit has some advantages with predictable trip lengths: no transit agency is going to insist that they need a bus that can go 200 miles on a charge if their average bus only travels 50 miles each morning. Meanwhile, the average consumer ``needs'' car with a battery that is several times what their 99th percentile trip length is.

[iwwr]

It's energy inefficient, but you're thinking in terms of energy scarcity we have today. What if we had 10x more electric energy capacity (most likely due to nuclear resurgence)?

[ta12121]

Energy scarcity is a fundamental property of technology and economics. Aside from a few countries like France, nuclear power serves a fraction of our electricity needs, and electricity is a fraction of our total energy needs. Where are all these nuclear power plants going to come from? The technology described in the article is a way to transform energy. Not to produce it.

[ommunist]

I think you are right. Also that last mile Boosted Desk surfing transport is quite a thing. Batteries are forever.

[comicjk]

Lithium-air batteries (an immature technology, certainly) have an energy per gram 3.5 times less than gasoline's. But electric motors are about 3.5 times more efficient than internal combustion engines. And the technology to produce synthetic fuels from CO2 is pretty much as far from mass scale as lithium-air batteries.

I agree that we will need carbon-based fuels for jet engines, but as for cars, the physics and economics are by no means as favorable as you represent.

[dalke]

"Let's say tomorrow some grad student gets fusion going at a very low price. The best way to use this to power cars would be to use it to create a fuel with a high energy density."

That is incorrect. If the fusion source is compact, and can produce a lot of instantaneous electrical power, and is quick to throttle up and down, then a "Back to the Future"-style Mr. Fusion would be the best way to power cars.

Energy density isn't the only factor. There's also a question of infrastructure. Electrical distribution has few moving parts, while extracting C02 from the atmosphere to make fuel and distributing the fuel has many more mechanical parts. A fusion plant which could produce 50 kW, weighed 2 tons, and could be installed in the back yard of a home would mean that a house could be off the grid and still have power left over to charge the car, while using intermediate hydrocarbon storage would mean trips to fill the car, or heating oil, or cooking gas.

So while I completely agree that airplanes will not be powered by batteries, I don't think that energy density is the only factor to consider in the economics equation.

[stcredzero]

> That is incorrect. If the fusion source is compact, and can produce a lot of instantaneous electrical power, and is quick to throttle up and down, then a "Back to the Future"-style Mr. Fusion would be the best way to power cars.

There's also safety to consider.

[sophacles]

On top of that, reliability. To have a large plant with a high mechanical complexity which can justify dedicated maintenance workers to help manage complexity and the results of part wear etc is likely able to achieve the same or higher reliability than a backyard unit, in TCO terms anyway.

[dalke]

So I think you both agree with me that it's not necessarily the case that a hypothetical fusion power source or free energy source is best used to produce hydrocarbons for distribution and downstream use.

Regarding the comment of sophacles, power distribution is part of the economic factor. It may be that the central plant is much more reliable than a backyard plant, but the power grid - subject to thunderstorms, ice, tree falls, backhoes, curious squirrels, and so on - makes the overall power supply system less reliable than a backyard fusion plant.

[sophacles]

No. I will agree that a distributed power system may provide overall reliability but this condition must be true:

There is still a grid. If my power source goes out, I want backup to come from other nearby sources - the timeline of power restoration from the current delivery system is on the order of minutes or hours for over 80% of outages, and on the order of a couple weeks for over 99% of the rest of outages. If my power plant breaks, I need restoration numbers that meet that. (additionally, I need plant repair bills to be lower than however much money having the backyard plant would save me. TCO considerations again).

Further, these two assumptions are built into your "better" assessment:

* It is cheaper to have a power plant in my back yard than buying it from the grid.

* The backyard source can be made safe.

Combining these two assumptions is a big deal. If both are true, I will agree that it is a good option (with the caveat listed above). However, there is a HUGE amount of R&D to get there, including a massive set of efficient production runs for parts to build all the systems to make it happen. The economics of this points to it not being likely that everyone has a backyard fusion plant.

It is far more likely to see big fusion plants in greater number scattered around the power grid to provide higher reliability in the cases of line loss etc. Further, with energy now being much, much cheaper to produce, you'll likely start seeing more reliable distribution channels for electrical power. Overhead lines would be reasonable to replace with underground ones, which are less efficient, but are also more reliable as they are less likely to be damaged in weather events. You'll also probably see a reduction in star-topology distribution - more redundancy in distribution paths, at the cost of some efficiency, because the complex equipment will be cheaper to manufacture (you know, because energy to do so will not factor into costs anymore).

[dalke]

The original premise was already unrealistic. I made it even more unrealistic. If there is a "Mr. Fusion" device which can produce 1.21 GW, on-demand, safely, and it small and light enough to fit in your car, then there's no need for a grid. You would just have several of those devices in your house.

My hypothetical was to show that there could be cases where it does not make sense to use a Mr. Fusion type device to produce hydrocarbon fuel which is then used as the energy source. Everything I said takes place in the original fantasy world. Under the original premise -- "some grad student gets fusion going at a very low price" -- then it must be using some principle we haven't yet thought of. And with that premise in place, almost anything goes.

Once I put realism into place, then the original hypothetical is not sustainable. The long term solutions for real life are decreased energy use, fission, hypothetical fusion, and renewable. None of the last three can exist without a grid, at least for most people. The only way to be without a grid is greatly reduced power use, a less concentrated population, and switch to local renewable resources. That isn't going to happen.

[jnw2]

Several strings of solar panels + lithium ion batteries might very well be cheaper than paying your local electric monopoly for transmission line capacity in 20 years. And if the non-redundant parts (the inverter, perhaps) fail, it might not be all that different from your water heater failing today.

(Although if we had that technology in cheap enough form, some of the major loads in your house may switch to DC to avoid conversion losses to AC, since solar panels and batteries are both inherently DC technologies, and that might make the inverter less important.)

[lumberjack]

Your argument is only balanced on that one premise of Fusion or free energy, both of which are still vaporware and will be for decades if not centuries to come. Have you considered that maybe electric cars will fill the gap between now and then if ever? And you consider free energy a possibility but advancement in battery technology too far fetched. I don't agree at all.

[pigboy]

Energy density is a critical issue for aircraft because lift/drag ratios are capped at around 15 to 30. For every 15-30 N of fuel weight, you must supply at least 1 N of additional thrust (which translates to power, once you multiply by speed).

Cars, on the other hand, experience a combination of aerodynamic and powertrain drag and rolling resistance, of which only (mostly?) the rolling resistance depends on mass. The typical coefficient of rolling resistance is about 10x better than the typical L/D ratio for aircraft (if I believe Wikipedia). Hence, road vehicles are 10x less sensitive to energy density than aircraft.

So a statement that electric vehicles won't work is nonsense, without considering how sensitive the different vehicles are to fuel mass.

[Mvandenbergh]

Most people miss the difference between hydrocarbon fuels as energy source* and as energy carrier. If you make HC fuels from atmospheric or oceanic carbon, you have a nice closed cycle.

(*) Technically fossilised solar energy of course.

[Dn_Ab]

Very good point. But it doesn't make Tesla cars a folly so much as put a strong constraint on their long term economic viability: the invention of fusion power or some other method of making energy abundant turned to making hydrocarbons effectively renewable. But that could be decades away, timeline of practical viability is effectively unknown. Considering how speculative this notion of really cheap energy is (for now), I still think Tesla is a good bet. On the event of the floor dropping off the prices in energy markets, there will be few that are dependent on large energy barriers to entry that will emerge unscathed.

It is also effectively win-win. If such a thing becomes possible the costs in much of production will cut very drastically and the potential of things now open would be just incredible. So much of expense is due to energy constraints.

[confluence]

You are wrong.

Electric engines run at 92% efficiency. Car engines run at ~15%. So energy density need not reach equivalence.

That physicist is wrong too (Nobel prizes don't mean you aren't wrong - it just means people listen to you more) - jet engines on planes require oxygen to work - hence capped at ~10 km altitude with massive drag causing ineffecient travel.

10 years till battery tech reaches complete parity for lowest denominator cars, and 7 years before we start to see hyper sonic electric passenger jets being tested way up in the upper atmosphere (no oxygen needed, go as high and fast as you want - London - LA in a few hours - space views - cheap power).

[jrabone]

What? 10 years until battery technology has advanced to the point of parity?

Gasoline has an energy density of 132 MJ/US gal, or say 1500 MJ on a full small tank. Modern engines hit 30% thermal efficiency, but then there are drive train losses etc, so I'll give you the 15% as an absolute worst case, AND give you a 100% efficient electric motor / drive train. Therefore your car needs to provide 225 MJ of stored energy for parity.

The best commercially-available Li-ion battery has an energy density of around 245 Wh/kg, or ~ 880 kJ/kg. Therefore you need at least 255kg of conventional batteries for motive energy parity (ie. not counting heat/AC/light etc.). The equivalent in gasoline (0.77 kg/l), weighs 35 kg.

You are asking for nearly a ten-fold increase in usable energy density in battery technology in ten years, at price-parity with gasoline, and with the charging infrastructure to support it. I don't think it's going to happen - assuming lithium-air batteries ARE commercialised on that time scale, the cost of charging that kind of energy in reasonable time, safely, is going to be the real problem here. Every gas station is going to need its own nuclear power plant!

[confluence]

I'm assuming nuclear fusion is viable with the development of the ITER and DEMO plants in the South of France.

If they aren't viable - I'm calling time of death on the world.

[DennisP]

Don't rest all your hopes on ITER. Alternative approaches that could well be cheaper and come to fruition a lot sooner include NIF/LIFE, polywell, focus fusion, General Fusion, Tri-Alpha, Helion, levitated dipole, petawatt picosecond laser fusion, and Sandia's new approach to magnetized inertial fusion.

If none of them work, advanced fission designs like LFTR or IFR could be almost as good, with better safety and a hundred times less nuclear waste than conventional reactors.