brimic, the one benefit I could see to having each local fill station in a city be set up with a 10MW reactor would be increased capacity to the power grid.
How many fill stations are in a city of a million people? Perhaps 500 or so?
That would be 5 gigawatts of power generation added to a metro area. Our big nuke reactor here, Palo Verde, only generates about 3 gigawatts.
There is a reason why there is a resurgence in efforts to build small, inherently safe reactors (btw, the tidbits design sucks, and don't even get me started on hyperion)--distributed power generation.
However, for electric cars, the fast charge issue is a tough nut to crack. Even the current high power nano phosphate based lithium ion batteries are limited to 5-10C charge rates (6-12min). While there are technologies being developed that could push it to 20-50C (what would be needed for a true "gas station charge" the issue then become heat dissipation in the battery pack. Due to internal resistance issues, charging 10kwhr of battery capacity in 3 minutes (200kW) would result in the generation of nearly 25-50kW of waste heat in the pack...assuming typical liquid cooling, that would require a radiator twice as large as a typical sportscar, or result in the boiling of over a gallon of water!
What I See in the future, should the electric vehicle become more commonplace, is the following:
1. Series hybrid diesel vehicles (small diesel with very high pressure turbocharging at low revs with a modified miller cycle) for "unlimited range, but "plug in hybrid" size (e.g. Volt/leaf) battery packs with liquid cooling and advanced nano phosphate battery technology
2. Charging stations using a custom fast charge connector at 300V/600A with on-connector quick-disconnect coolant lines for high flow chilled liquid cooling during charging (with the above battery technology, a lifetime with proper design of 500-1000 fast charge cycles should be possible, or a battery lifetime of 50-100k miles if you only use that method)
3. Distributed power generation stations with local storage for peaking (e.g. Flywheel) at "gas stations". There are ~120,000 gas stations in the US, or roughly one for every 1500 homes. In the US approx. 40 billion vehicle-miles are driven every year, assuming an all-electric range of 50 miles, and half of all miles using quick-charge, and all of those charges being used in a 12-hour period, the average station would need to charge 1-2 vehicles per hour, or about 10-20kW on average. Assuming a "worst case" station load of 50 cars in a single hour, and a total of 200 in a day, a local storage of 1000kWhr (within the range of medium size flywheel storage), and an average load of 100-200kW would be reasonable. If each station had an efficient, combined cycle, 500kW class micro turbine, the surplus electricity from those 120,000 stations would be (on average) 42Gw, or enough to power about 20 million homes (10% of the residential load in the US)...since people are either at home/work, or driving, this would also allow for significant load leveling and more steady state performance. The resulting overall efficiency would be 1.5-2x conventional hybrids from a co2 perspective (or more, depending on turbine fuel, as their efficiency is 1.5-2x a typical engine, and natural gas results in far less co2 than the heavier hydrocarbons).
So that is an interesting approach. HOWEVER, my preferred approach is to use synthetic hydrocarbon (methanol and/or Dimethylether, which replace gasoline and diesel, respectively) generation (which can be done at high efficiency from water and atmospheric CO2) at large facilities localized at nuclear power plants. This way all of the biggest problems with alternative transportation fuels are eliminated.
1. The infrastructure stays virtually the same (pipelines, trucks, gas stations)
2. Minimal to no vehicle modifications are required
3. Carbon-neutral
4. Oil independent!
5. Allows for off-peak use of base-load capacity (makes the power plants more efficient)
6. Allows for economies of scale
With this method, the capital cost required to transition the infrastructure (which we have trillions invested in already) and vehicles (again, more trillions in already sunk costs) is minimal, the technologies are already in hand, and the investment strategy is well defined--beginning with current power generation facilities, add fuel generation equipment, then start building new nuclear plants until over a decade or three, the entire system is transitioned. Btw, even at current huge installed costs for nuke plants, ($5-10/watt), transition of the entire Petroleum usage of the US (21 million bbl/day) to synthetic fuels in this fashion would require the installation of ~2TW of new plants ($10-20trillion, or about $500billion per year amortized costs). While that seems like a lot, that means the equivalent "oil" price for profitable break-even is $99/bbl (assuming 60% efficiency on the conversion, and a 1.5:1 cost ratio of total to amortized capital cost).
Since oil is already above this, and likely to stay there, doesn't a truly long-term, carbon neutral, all domestic (note, all of that $500b/yr is now domestic jobs and investment) alternative that uses all the existing infrastructure sound good? I mean, reversal of the trade deficit (switching $300 billion out of the import column, and replacing it with $200 billion of oil EXPORTS!), elimination of the co2 produced from transportation, and an overall reduction in fuel and energy costs seems like an obvious choice doesn't it?
Now, why aren't we going down that path?