Energy From Thorium Discussion Forum

That is true as far as it goes but keep in mind that 50% here is just the coupling efficiency - not the system efficiency end-to-end. It doesn't take into account losses through the rest of the system. Much of what I have read this evening on the subject - and there is a fair amount out there - glosses this over, or invokes superconductors and precision network control via computers; all possible in theory, but not yet ready in practice.

And even if it were as efficient as electric rail (which is very high) the capital costs cannot be justified for RPEV if it cannot be justified for rail (which in fact it cannot.)

Yes but - if you have no battery you must generate and transmit the electricity when it is needed.
You compared the cost of the battery to the cost of extra electricity.

To make a full comparison, one should add the cost of peak generators, and transmission lines to the RPEV and deduct the savings achieved by not having to lug the battery around all the time.

Yes, I agree completely.

By the way, do you have any pointers about why electrification of rail is not economic? I'm curious about that for other reasons.

-Iain

The main costs of electrification are in the construction and maintenance of infrastructure; as well as the need to purchase different rolling stock (or at least locomotives). In one recent case an exorbitant $115 million for 10km. At prices like that it is difficult to justify the investment.

There are some routes that the cost is recoverable over a reasonable time, and indeed there are plans to convert them, but mostly these are commuter-type light-rail trains that have the traffic.

Now there is some indication that changes in attitude and regulation are on the horizon and things may improve, but at the moment electrification is going to be slow.



This is the whole point of these systems, it is why they have been designed. However it would seem that the very people that have worked on these ideas are glossing over these very losses in the literature. When this happens, you can be damned sure that the raw numbers don't look that rosy. Any time someone proposes inductive power transfer, idling power loss is going to be one of the first questions asked - that there isn't a ready answer is telling.

Before I start in on more calculations, I realized there was one interesting feature to this system I proposed. Since the car is following the cable in the road, it doesn’t really need you to steer! You put some radar in the front bumper to maintain a proper following distance and you can just have to car drive itself and notify you when you get close to your exit. I suspect this would result in significant decreases in highway fatalities, considering the distraction levels for the modern commuter, especially if auto drive was required for freeway vehicles.

I've put all my references at the end.

The next part is to calculate how many cars are on the freeway at rush hour. Surprisingly, I couldn’t find a ready number for that, so I went back to the envelope. I found a reference that the lane capacity of a freeway is around 2,300 cars per hour. At 60 MPH, that works out at about 138 cars per mile or 826 cars on 6 lane freeway (3 lanes each way). The 4 lane road would be about 551 cars. Since I previously found out that there are 75,000 miles of freeway in the US, I assumed 10,000 miles of 6 lane freeway and another 10,000 miles of 4 lane freeway and then threw in another couple of million cars for the rest of the freeway system which comes out to around 15.8 million cars.

But what do I do about freeways that have more cars than their capacity? As near I can tell, the cars would actually use less electricity because they are moving slower and generating less wind resistance. Read “Sustainable Energy – Without The Hot Air” by David MacKay if you are interested in the math.

There are around 235 million cars and trucks in the United States, not including larger commercial vehicles and stuff. This means that 15.8 million is only 7% of the total. I figured not everyone commutes at the same time and a lot of them are on surface streets and some people actually car pool and besides if more cars are on the freeway, then they don’t use more electricity and if someone knows the actual number, then they will correct me.

Next I had to figure out how much electricity each car would use. After looking at the specs on a bunch of EV, I came up with 10KW per car by assuming significant weight savings from a smaller battery pack. I suspect the actual number is higher, since I want the cars to cruise at about 60 MPH. If people in Los Angeles could average 60 MPH during their freeway commute, they would start naming their kids after me. Come to think of it, with the auto drive function, their kids might be conceived during their commute. I’m sure it has been managed a few times, without the auto drive.

Anyway I multiplied the two numbers together and came up with a total of 158 Gigawatts. At this point, I had what might be referred to as a Doc Brown moment, and used some words not suitable for a family forum. Remember afternoon rush hour actually overlaps the peak electricity consumption point of the day. If I want 158GW, then I have to generate it. The whole nameplate capacity for the United States is around a Terawatt and actual average power generated is less than half that (464GW).

I’m pretty sure I can’t get peaking power for less than a $1 a watt. Actually after adding in the rest of the plant and the power distribution, I’m certain I can’t do it for a buck. And I need another 160 billion. Suddenly the power plants cost 6 times as much as the whole freeway system and probably over 10 times and I’ve got a system that costs over 200 billion and probably over 300 when the usual pork is added in. This might still be worth work doing, but isn’t exactly cheap.

This annoyed me, since in my previous post, I eliminated Global Warming and put a dent in International Terrorism for 25 billion. How could I be so far off.? I thought I remember reading that we could convert all our cars to EVs without adding any electric generating capacity.

I found the paper and skimmed it. What it actually said was that we could have one EV per household without adding additional generating capacity. They assume that the charging would use 1.625 KW for 8 hours to add 13KWh to the battery. They didn’t actually calculate it for the whole country, since they were using Cincinnati and San Diego for their examples. There are around 100 million households in this country, which works out to 163GW to recharge the batteries for 100 million vehicles, which would be 43% of the total cars and light trucks in the country. Technically this doesn’t require new generating capacity, but tends minimize that whole peak/off peak differential , so the electricity gets more expensive or we build a lot more off peak power. If we want to convert most of our cars and trucks to electric power, then I’m not sure if peaking power is even an option.. I’d don’t think there is enough Natural Gas to do it.

I guess my next part will be to calculate if most of the electricity comes from Natural Gas, then does it make sense to use an electric car versus running one on CNG?

References:
http://www.state.nj.us/transportation/eng/documents/RUCM/Section3.shtm
http://www.inference.phy.cam.ac.uk/withouthotair/cA/page_254.shtml
http://en.wikipedia.org/wiki/Toyota_RAV4_EV
http://en.wikipedia.org/wiki/General_Motors_EV1
http://www.eia.doe.gov/cneaf/electricity/epa/epat2p2.html
http://www.eia.doe.gov/cneaf/electricity/epm/table1_1.html
http://www.euec.com/documents/pdf/Paper_5.pdf

Now remember that the coupling losses are ~50% for an inductive system and a 5% transmission loss from the generator to the coil. Then losses in the coil itself, most of which will be reactive with the angle fluctuating rapidly (the load is dynamic after all) needing even more available power to overcome, and this just isn't going to happen without superconductors and a very smart grid.

Not a very practical idea at this time.

You seem to be into doing apple and orange comparisons.

There is no cost for the 'rolling stock'. People have the option of buying an EV when they normally would buy a new car. Their old 'rolling stock' still works fine. Of course, if you are trying to get most people to switch to EVs, then the system has to work without subsidies, so you need to make the EV as cheap as possible. Who would subsidize them? Farmers? People who don't own a car? Our current system taxes the poor, so yuppies can buy EVs, but that hardly makes sense if you want everyone to drive an EV. Robbing Peter to pay Paul is unfair, but robbing Paul to pay Paul only makes sense to a politician.

I have no idea where the $115Million/10km number comes from, but I don't need to discuss it. I noted that KAIST and I came up with numbers in the same ballpark. I estimated $80,000 a lane-mile. Laying fiber optic cable in a rural area uses similar type systems but obviously a much smaller cable. I've seen estimates of as low as $16,000 for laying fiber optic cable in rural areas. An estimate 5 times higher to do something similar hardly seems absurd and even one 15 times higher would do all the freeways for $75 billion.

It's is possible that EVs don't make sense with or without a big battery pack. I did some math in the last long post. I suspect that if a lot of people supporting EVs suspected that they would require hundreds of new power plants, they would change their mind.

You may be interested in this ORNL study:
http://www.ornl.gov/info/press_releases ... 0080312-02

If you read the post with more care you would see that it was answering a previous question on electrification of railways and of course if one wishes to take advantage of an electrification program on a railway one would need to replace the rolling stock, in particular the motive power units. The $115Million/10km number comes from a local project that I am familiar with, again involving rail.

The point I was making is that electrification of that mode is prohibitively expensive in most cases despite the fact that efficiencies are much higher.

Before you gulp that 158 GW is a lot of electrical power, consider the petroleum power used...

3.290e9 bbl/yr
x 5.6e6 btu/bbl
x 1 kwh/3412 bbl
x 1 day/24 h
x 1 h/365 day
= 616 GW

Also, before concluding that the price is sky high consider that we consume some 20 million barrels of oil per day. Oil is currently around $70/barrel. So it is worth a lot to replace a substantial portion of this with electricity. Suppose your system suffices to displace 1/2 of the oil used for transport, and that 1/2 of the oil consumed is used for transport, this would mean we save roughly $350M/day. If I take your $200B then you system has paid for itself in roughly 2 years.

You point is valid. I was just shocked at the capital cost of constructing that many power plants. Also I suspect that 158GW is way too low. It is based on each car drawing 10,000, which should be ballpark if the cars are driving at 30MPH, but is probably too low for cars driving at 60MPH.

I agree with that part. I think achieving independence from unfriendly foreign countries for our fuels would also also be good. I need to calculate if I'm using natural gas electric power, if it makes more sense to just run the cars on compressed natural gas directly.

Speaking of peaking power, I was reading that a BWR can go from 50% power to 100% power in 1 hour. Is this true and can other reactors do that? I grew up in TVA country, so having the federal government build power plants doesn't offend me, but I rather they built nukes.

Notice they are talking about plug-in hybrid electric vehicles. They are probably talking about something like the Chevy Volt, that has a battery pack, but still has a gasoline engine. The Volt can only go 40 miles on battery. Most pure EVs has a range three times that on battery. The full charge range is 8.9KWh. The battery has a capacity of 16KWh, but GM limits the maximum charge/discharge to maximize the service life of the battery.

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

Is there other variations on Li-ion batteries that are less finicky?

In 1998, Mercedes Benz were on the point of launching an all-electric version of the A Class small car. Powered by a 30kWh Zebra battery weighing 370kg (including control system), the vehicle was claimed to have demonstrated a real world range of 120 miles. A fleet of 16 A Class cars tested the Zebra battery in all weather conditions - some of these vehicles are still being used by Mercedes today.

Improvements in Zebra technology since then have reduced the weight of a 30kWh unit to 270kg (120Wh/kg). This specific energy is superior to any automotive LiIon battery available or under development today.

The Zebra battery is suitable for pure BEVs but it is not suitable for hybrids.

The history of Zebra development dates back to the 1970s. The technology was first developed in South Africa. By the 1990s, there were some 8 companies developing sodium beta batteries and 4 pilot production plants were in operation. Today, only MES-DEA in Switzerland manufacture the Zebra battery, although NGK made some for stationary applications in Japan for a number of test installations.

As a "hot" battery, ambient temperatures have no effect on battery performance. In sub-zero winter temperatures, a Zebra powered EV will deliver as much power and energy as in mid-summer. This temperature independent capability is a unique feature of "hot" batteries such as the Sodium beta variants (Sodium Sulfur and Sodium Nickel Chloride). It also can supply cabin heat, another concern in colder climates. They also can be charged at high currents without damage.

The Zebra battery NaNiCl technology makes far more use of cheap and readily available materials than LiIon and NiMH. The major active materials are nickel, iron and common salt, along with some aluminum. The separator is ceramic beta alumina, a very inexpensive material. The case is made from stainless steel. The only potentially limiting active material is nickel, although MES-DEA state that less than one third as much nickel is required per kWh as NiMH (1.53kg/kWh compared to 6.8kg/kWh for NiMH). Cycle life of over 1400 nameplate cycles has been demonstrated which amounts to a 10 year calendar life in an average automotive profile.

If nickel availability becomes constrained, Zebra technology has the potential to be developed into an even lower cost variant that would use little or no nickel - the Sodium Iron Chloride battery. This has an open circuit voltage of 2.35V against 2.58V for NaNiCl, but could be manufactured in unlimited quantities from very cheap and ubiquitous active materials (iron and common salt). Specific Energy would only fall by 9% and battery operating temperature could be reduced.