(ppps - wow. I just looked at the incentive bill for Comprehensive Energy Legislation. It puts nuclear power under the same umbrella as wind, solar, etc.
It doesn't give the same credit to nuclear as it does to wind - ie: getting a flat-rate 1.2 cent/kwH production credit to wind, plus a 1.8 cent/kwH tax credit, but it *does* extend the 1.8 cent/kwH tax credit to nuclear for the first 6,000 MW.
Not quite the same, but whoever builds the first advanced nuclear plants is going to get a windfall - they truly are going to be able to 'print money'. If the next gen plants generate at 1 cent/kilowatt hour operating cost, then as I see it, at 40% tax rate, they are going to be able to generate electricity for free and sell it for sheer profit...
Now if they can only streamline the hurdles for getting approved.
said that the operating costs for nuclear energy are about 1.5-2 cents a kilowatt hour.
Right, so the way you prove that nuclear is cheaper than the alternatives is by ignoring all the costs except for the one that makes your argument look good.
the '12 cent/kWH' figure that you get factors in a lot of the original R&D cost
Now you're making stuff up again.
Ok, lets talk some figures here. I'm not 'making stuff up' - my argument is that the statement 'Nuclear is Expensive' is a inherently useless, dumb statement. Its like saying 'cars are expensive' or 'computers are slow'. And that the 12 cents/kwH figure that you cite lumps everything together into one big basket, rather than looking at things critically.
Is nuclear expensive? If you're talking about 1st generation nuclear power I'd agree with you (it being done primarily for military reasons). If you're talking about 2nd generation (ie: 40 year old technology) nuclear power, it really depends on what you include in your scope. And how long the plants last. If you've got a second generation nuclear power plant (like the ones at Vogel) that are producing electricity at a cap of 2 cents/kWH, and that cost about $5000/KW to build, the economics of running the plant become pretty straightforward:
where.9 is the capacity factor and 5 billion dollars is the startup cost, and 1 GW is the amount of energy produced. Calculate this, and it gives approximately 10.25 billion dollars cost for running that power plant. Divide that by the number of kilowatt hours produced:
(10.25 billion dollars) / (1 GW *.9 * 30 years)
and you get about 4.33 cents / kilowatt hour (if you don't believe me, believe frink). If they last 60 years:
that's approximately 15.5 billion dollars for the running of the plant. And at a total cost of about 3.3 cents/kWH.
So - where does Diesendorf get his 12 cent/kwH number? Well, first of all, there is the insurance 'subsidy' that you talk about (which a large part is the 'Price Anderson act' which protects nuclear producers with a cap on disasters) and the R&D 'subsidy'. If you factor in that research used to develop those plants, plus failed reactors (ones that started and never completed), plus the past R&D going back to 1945, you get (I'm guessing) about 12 cents/(kilowatt hour). And that's the number that you are apparently attracted to.
But that's my point - these costs (failed reactors, botched administration) are unlikely to re-occur. The nuclear plants *themselves* are competitive. And the R&D isn't useless - like say, the cost of supporting a war to keep ourselves hooked on oil, or the cost to cleanup the environment and the damage to agriculture. Its not money wasted, its a one time charge spent for plants to come that are now being built in (ironically) india and china with our R&D. If you take the millions of man-years used in us perfecting nuclear power and put it into engineering practice, you drop the startup cost from $5000/KWH to approx $1000/KWH, and the running of the plant from 2 cents/kwH to about 1 cent, which is cheaper than your average coal station. Which is what is being built in india at this very moment (the advanced westinghouse reactor).
So, in *further* answer to your question, 'is nuclear expensive'? No, not *modern* nuclear power. I'd also argue that most of the 'subsidies' that are mentioned in these calculations aren't really subsidies; that they lead to bigger and better things, and to a point where the 'subsidies' end. Sort of like the subsidies for wind and solar energy, which you seem so fond of.
Ed
(ps - I didn't mention Lovins because he bolstered my argument; I mentioned Lovins to show that I had read more than
And when you turn them on, its like printing money - it costs on the order of 1.5 - 2
cents/kW to maintain and run them.
It costs 12c/kWh for nuclear power. I've no idea what hole you pulled your figure from, but I could take a wild guess.
First of all, your "original comment" was 'Nuclear is expensive and nobody wants a nuclear
reactor in their backyard. End of discussion'
I'm glad you've finally noticed. I've no idea why you've put "First of all" in front of that
sentence though, because you're the one who should be apologising for not paying attention to what I said.
I'm feeding a troll here, but here goes -
1) I said that the operating costs for nuclear energy are about 1.5-2 cents a kilowatt hour. As for the base cost, like I said its approximately $1000/kWH. And I got that figure from both an operational plant (the Vogel twin plants), and the profit figures from nations who rely on nuclear electricity (france, sweden).
2) the '12 cent/kWH' figure that you get factors in a lot of the original R&D cost which is not necessary to repeat, as well as a lot of the political costs that were associated with that R&D. It also has to do with the assumption that you are only going to be running the plant for 20 or so years, and it looks like most plants will be running for a lot longer than that (about 60 in fact). In practice, the true operating cost is about 1.5 cents/2 cents an hour. And - unlike wind - there is no subsidy for the operation of nuclear plants. Do you really thing the Vogel plants would be selling nuclear power at 5 cents per kwh, at a 7 cents loss? Do you really think that *france* would be selling 1GW-9GW nuclear power for 8 cents/hour to businesses (or to other countries like italy) for a 4 cent loss/kWh?
And that doesn't even count 3rd generation/4th generation nuclear reactors, which have orders of magnitude improvement over current, 1970s technologies.
Why isn't there more nuclear power? Again, its politics. Its ironic, because so much effort has gone into fighting nuclear power buildup that plants that to *real* damage, like coal, are sailing through and the greens are actually *retarding* sustainable development. Its no wonder that true greens (like James Lovelock) are amongst the strongest supporters of nuclear power.
3) what you *say* is worse than useless when you don't back it up with figures, references, reasoned argument and discussion. The fact of the matter is that I didn't see your original comment at first, just your second mouthed platitude (the more you look into it the more.. etc. etc.). So, you tell me to go to the first comment, which I dutifully do so, and see that its another useless mouthed platititude. Doesn't do me - or anyone else - any good.
Read some books. Please. Start with those I mentioned, and then read Vaclav Smil, Amory Lovins (although he's a bit 'pie in the sky', he's got some good ideas on efficiency), and maybe deffeys. I'm not 'nuclear zealot' - I am not in the nuclear industry, and have no vested interest (other than wanting to keep living with the comforts of civilization). And I think that wonders can be done with solar and wind - to shave load off of peak power. But *replacing* peak power is a totally different thing - Denmark is finding real barriers in doing this, even though they are under political mandates to try to totally rely on renewables - and their spot price for electricity is twice as high as the countries with supposedly 'too expensive' nuclear power.
Why would I do that? This is why I am not wasting my time discussing this with you. My primary comment - as demonstrated in the SUBJECT - is that nuclear power is expensive. That is an indisputable fact as even the nuclear power lobby admits that it is only profitable in the US when subsidised, and is best suited for countries that do not already have cheap plentiful supplies of fossil fuels (such as France).
First of all, your "original comment" was 'Nuclear is expensive and nobody wants a nuclear reactor in their backyard. End of discussion', which is just as fatuous (and empty headed) as the comment that I responded to.
Second of all, I could argue with that comment (as devoid of info as it is) in three ways. First, if you look at the countries that have large amounts of nuclear power, their energy cost per-kWh is on par with those who rely on coal, etc. Source here(look at France && Sweden && Germany for examples). And those countries which rely heavily on renewables tend to have *twice* as high an energy cost (look at the table for Denmark for example!)
Second, I could say that we already are heavily subsidizing coal, natural gas and oil in the form of military encounters, damaged productivity in agriculture, damaged health, loss of life, and so on - and that with large-scale solar power that this subsidy would continue - although not as great as it would be with coal. For it takes coal/oil/etc to *build* solar panels and solar towers, for an energy source which is highly disperse and needs a lot of infrastructure to put 'online'. That's where the figure of 200 g/kWh capacity is a killer. Its 25% that of coal - which means if we scaled up solar, and used 4 times as much electricity as we did now - which is not a stretch considering the amount of population in the world and how hungry they are for energy - we'd still be generating the same amount of greenhouse gases and associated emissions we are today. Which is not an acceptable solution.
And third - and finally - I'd say that you have a woeful misunderstanding of the economics of power plants. All the pain that we've caused ourselves on disposal and lots of the overruns on the building costs of current power plants are caused by politics, pure and simple. Its easy to make anything you want as expensive as you want if you throw hurdles, lawsuits and FUD at the problem. Take away the FUD and the things are relatively in-expensive to build - current LWR around $1000/kW to build in comparison with coal being $1100/kW. And when you turn them on, its like printing money - it costs on the order of 1.5 - 2 cents/kW to maintain and run them. The source of fuel - uranium and its preprocessing costs - is close to free because it is so energy dense. And the important thing - especially in comparison with coal/natural gas/solar/wind is that their capacity factor is *great*. Which means they run almost all the time - circa 90% around the year 2000 in the US, as compared to circa 60% for coal and circa 30% for wind. Which means they can be used easily for industrial base load.
And fourth - finally,finally, there are magnitudes of improvement left in new designs. As I said in my previous post, self-contained, automatic, passively safe breeder reactors such as the sstar are designed to slash the amount of source material by 70 times (by using U-238 instead of U-235) and to cut the capital costs in manufacture by about 2 to 3 times due to new materials manufacture.
So yes, the economics of nuclear power plants is quite cheap once you get the politics out of it. As for costs of decommission and storage, those are greatly reduced if you simply let the radioactive actinides disintegrate and then move the relatively inert radioactive waste - or do what the french do and simply reprocess the fuel, and embed the residue in volcanic glass. We produce about 2300 metric tons of waste a year to deliver about 1/5t
I used photovoltaic cells because that is the most commonly known technology.
In the commercial space, superheated water and thermal towers are the most commonly used technology.
In the commercial space, solar provides far far less than one percent of TPES, and when it does it is in niche areas. Why do you think that is?
Now, I would suggest stopping the mouthing of empty platitudes and do some research of your own.
I have done far more research on this topic than you could imagine.
Great.. then spill it. Show me a counter-example study showing g/CO2 figures from cradle to grave for different solar energy technologies, and I might believe you. Otherwise, what good is your 'I have done far more research on this topic than you could imagine' statement?
Its a discussion group, so you know.. discussion might be nice...
Just curious, I looked at your website, and you don't seem to mention 3rd/4th generation nuclear power.
In particular, I'm interested in the use of nuclear power *not* for electricity generation, but for heat generation, to replace natural gas for such activities as:
* ethanol production from cellulose
* hydrogen production from coal
* steam production for oil sands development
To me, the biggest story about nuclear is not going to be the electrical plants (although those are important too) but in the creation of portable, off-the-grid thermal generators. For example (as seen on slashdot), the sstar, which I *hope* they are planning on converting to use for thermal use. 100 MWe == 350 MWth == a *lot* of ethanol produced.
I used photovoltaic cells because that is the most commonly known technology. Thin-film PV cells (eg - copper indium diselenide and cadmium telluride) use less materials/square meter, but their conversion efficiencies are less and the infrastructure needed to support them is more; the amount of CO2 emissions per watt-hour (200 g/Wh) is pretty much a constant amongst the different technologies.
In short, the amount of C02 that even circa 1950s nuclear power plants emit is 5 times *less* than wind, 50 times *less* than solar (where do you think all that material to make solar cells comes from anyways?). As for cost, new nuclear power plants are about the same as natural gas turbines (which are as cheap as it gets) and have orders of magnitude left for improvement with new materials. As for safety, coal with its byproducts kill about 50,000 people yearly. Since its the extraction of materials blamed for premature deaths, if solar was scaled up to the same level as coal was, it would kill about 1000-2000/yr.
Face it; every source of energy has its risks. With nuclear, most of those risks can be mitigated because you get so much bang for your buck with the fuel that you can watch the risks carefully. Passively safe systems make things even more secure. Its one of the biggest irony in history that the Greens are fighting tooth-and-nail probably the most environmentally conscious enery source of all. (outside of conservation)
I hope you are being ironic here; Saying "But those evil arrogant Romans got their comeuppance is sort of like saying the "Tzars sure got them comeuppance by the communists and their secret police".
Yeah sure.. but lots of people had to die in both cases, of course.
I really don't want a fractured internet. It would suck, and it *could* happen.
1) the sales of the DS and of touch-games are pretty impressive.
2) the controller itself has an 'expansion' slot, into which you can put a standard controller (for backwards compatibility to play existing games)
I've got to hand it to nintendo. What a coup. I'm definitely getting one of these.
Ed
(ps - to hear it from the horses' mouth.. go to
here
Very informative presentation by the CEO of nintendo.
)
> Your calculations are reasonable; you don't make > your mistake until the end when you come up with > 15 grams (2200 Curies) of Sr-90 and figure > that's just great for personal electronics with > a small form factor!
> I don't know if you move in Sr-90 circles, but > 15 grams is a lot of pure Sr-90 to have. It > would emit 30 W of heat.
well, to be totally fair it would emit (30 W *.2 == 6 watts of waste heat (the rest being translated into electricity, which eventually would end up as heat, but in a diffuse sense). And you did start by saying that you would need kilograms of the stuff to run a laptop, which isn't true.
Anyways, 6W of heat is a bit, but not anything a decent cooling system couldn't handle (even air convection would do it). You would need to leave your laptop on all the time, however.
> Its radiation would kill you in a few seconds, > if you were directly exposed to it. It would > deposit in your bones like calcium if it ever > escaped its lucite container. Its weight would > be the least of your concerns.
well, the trick is not to get directly exposed then, isn't it..;-)
Anyways, the curie is not the unit of measurement about health, but the sievert. You're right though, I didn't do my homework about this and if the battery escapes its containment you are pretty much toast.
I guess its a question of how much faith you have in the ability of people to make an 'spill proof' container.. Given a system where the battery was bolted to the unit (and hence undetachable), and not requiring aqua chemistry simplifies the engineering problem. The working conditions would be difficult though (you'd probably need a completely automated factory to do it)
I did the math however, and it would be feasible to mass produce these - the nuclear industry has produced 40,000 metric tons of high level nuclear waste since its inception, and of those 40,000 tons, approx 500 out of every 10000 particles is sr-90.. So, we've got:
40000 * 500/10000
or 2000 tons of the stuff. Divide that by 15 grams and we could make 130 million of said batteries..
Increase the share of nuclear power by a factor of 10 and you get 1.3 billion batteries. Hell, at that level you could probably provide enough electric power for cars..
> You obviously haven't done the calcs. Efficiency > is not even the major problem here. The problem is > that ionizing radiation is an extremely deadly > form of energy- a median lethal dose in humans is > about several watt seconds per kg of body mass. > And to get even a tiny little bit of heat, you > need an enormous amount of radiation.
I'm not sure, but I don't think you know the difference between alpha decays, beta decays and gamma rays. As per:
"1 cm of lucite is enough to shield the most active beta emitters (32 P, 90 St)".
Beta decay rapidly breaks up when it encounters matter.
As for tritium, its beta decay is so weak that it could be stopped by the single layer of dead skin cells in your body.
And your math is off. To power a 15 Watt laptop, using Strontium-90 as a source, and assuming 50% efficiency, you need:
15 Watts * 2 efficiency factor / 2.3 MeV
equals
81 x 10 ^12 decays / second.
Given Strontrium's half-life is 28 years, this means:
8.83 x 10 ^8 secs = ln 2 / lambda
lambda = 7.84991 * 10 ^ -10;
N0 - 8.1 x 10^13= N0 * e^-lambda * 1 sec
Solving for N0, we need
1.03695 x 10 ^23
atoms of the stuff, or.17 mol
Given Strontrium's molecular weight is 90 g/mol, this equals approximately 15 grams. And given that batteries weigh in excess of a pound, this is highly doable, if it weren't for the cost of strotrium.
In fact you mentioned it yourself - Pu-238 (with a halflife of 87 years) gets a half-watt a gram, which means - if you had a 'nuclear battery' that was 80% efficient, you'd need:
15/.5/.8 = 38 grams
of plutonium to run a laptop.
However, of course you wouldn't use this, because plutonium is an alpha emittier, and hence is a lot more damaging than the betas. Hence, the use of betavoltaics in places like pacemakers, etc.
.. that's the old technology, which is at.1-.5 % efficient (that's right.001-.005)
Because of this inefficiency, there is lots and lots of waste heat, which causes the problems that you speak of.
If the nuclear battery mentioned on slashdot truly reaches its supposed 200 times efficiency, this is *80%* or thereabouts efficient, which means that there is a lot less material to radioactively decay.
Second of all, the batteries studied operate off of beta decay, which essentially means that they give off an electron. Electron == electricity, so the chances of needing lots of cooling equipment are probably not true.
So.. I haven't done the calcs yet, but nuclear batteries would probably be more feasible than you are thinking from an engineering standpoint.
On the other hand, tritium and strontium-90 are *damned expensive*, and would be so unless there was a circuit that connected supply (the nuclear power industry) to demand (the batteries) and even then, its doubtable that the nuclear power plants could produce enough strontium and tritium to keep everybody happy...
no, plutonium is not 'highly radioactive and among the most lethal toxic substances known to man'. See:
http://www.fortfreedom.org/p22.htm
where bernhard cohen (sp?) offered to eat as much plutonium as nader did caffeine on a TV show, just to disprove this DAMN DESTRUCTIVE MYTH/PIECE OF MISINFO for all time.
> It just occured to me you don't know how > transformers work. They step up one direction, and > step down the other. Even if I generate only 240 > volts AC with PV using a small residential > inverter it goes through my transformer and steps > up to medium voltage distribution lines (4-20kV), > it can go from there through a distribution > station transformer to 230kV+ for long distance > transmission power lines. Transformers are TWO > WAY. There is no low voltage transmission > requirement here.
Wrong. Transformers *can* be two way and be converted to be two way, but by legacy they are primarily either 'step-up' transformers or 'step down' transformers. And by far, the transformers that you see on the top of telephone poles (in america at least) are 'step down only'.
It is up to the *utility* to provide the infrastructure that you need in order to feed back solar into the power grid and *only on demand*. And of course they are fighting this tooth and nail. Hence they hardly have done *any* of this work.
To this end, they usually convert the pole transformers to be dual and rig it so that if you generate excess power, your metering runs backwards. (since hardly anyone has solar, this power then usually makes it to the next house down the block before being stepped down)
Note that this is a far, far cry from using solar to satisfy base demand. You need to:
a) complete the upgrade of the power grid to
get the power from many small residential
systems to one or multiple
distribution station transformers
b) figure out how to dynamically route this
power to where it is needed.
amongst other things (like quality control, billing, etc.)
These steps are not easy. Like I said, right now #b is a heavily manual process, even when the load stations are right next to the consumers of that electricity - and even the slightest mismanagement can cause blackouts and other system failures.
The problem grows exponentially as the power is intermittent and spotty - you really need to have the end-points which need the power be able to talk to the transformers to see exactly how much current they have at one given time. And of course this hinders centralized control - if a power line needs to be worked on, how do you guarantee that no current is flowing in it? And how do you tell the network that that line is down?
The problem reminds me a lot of the internet - and notice that the internet was a trillion+ dollar investement. And, after 30 years, we *still* can't rely on the internet for crucial services.
Let me ask you a simple question - do you really believe that this is going to be a dead simple process, and that the solar market will sustain 40% a year growth for 30 years?
Not at all. You're still underinformed about PV > technology. PV itself is low voltage, however > its converted to AC using an inverter (same > thing for fuel cell BTW) and transmitted over > the grid just like normal electricity. Done all > the time, every day, all over the world. 93-98% > efficient (I know, we haven't accounted for that > yet in our land calcs). PV very nicely > integrates into the current electricity > structure - no change needed The EROI numbers > given already account for the inverter energy > inputs, and its conversion efficiency. The EROI > stands as it is.
no.. I'm quite aware of how the PV cells work, thank you very much. However, it sounds like you are uninformed of how 'the grid' works.
Like I said, what you state up above is perfectly fine when it comes to local generation and electricity use. In fact, that's what the calcs were tailor made for - they hold up for solar as a LOCAL energy source, not on a high-power grid. Its sort of assumed everywhere that solar's going to be used locally, not for base-level power. And that's what I was asking for EROEI studies on - *base level* solar power. Fine - you disagree with Odum. Show me a study that contradicts his findings, with the assumption that you are using solar power for base-level power.
For the 'grid' is a misnomer - its NOT a two-way street - its designed for centralized distribution, and has been so for approximately 120 years. And that's where the transformity of solar falls down.
"the grid" is designed for high voltages over long distances - both for efficiency reasons, and for facilitated control. Hence, you *can* send electricity 'back to the grid', but its not going to get very far unless you have the infrastructure to support it.
FYI, here's a small primer on how large scale generation works: first there is a generation source, producing low (5-20 kV) voltage electricity, which is then 'stepped up' by a transformer to above 100 kV over alternating current. This is then sent through power wires to a substation which then 'steps down' the current to residential levels (there may be multiple transformers up and down). That's where the 7% efficiency rating comes in.
So - how does this relate to solar? The dimensions of the wires are low, the voltages are low. Hence the efficiency is when transferred through the network is low.
This inefficiency (the one you quoted) is probably accurate for local area networks, but transferring across towns, cities, or even states is bound to be much less efficient unless a boatload of transformers - are used to both step up the excess power, and to step down the excess power on the remote locations that might need it. It just makes common sense - you can't expect to transfer 100 volts cross country without serious inefficiencies - unless of course, superconducting, exceedingly cheap wire comes about - which I'm not holding my breath on.
Then of course, there is the question of routing. How do you know where to send the energy? If say, Minnesota has a huge snowstorm (covering all the solar cells), how does new mexico know that its supposed to send Minnesotans power, rather than Minnesota getting it locally? And if a snowstorm does happen, how do the solar cells get maintained and cleaned?
Right now, its relatively easy - and we *still* can't get it right. Base power load stations are assigned certain areas, and each routing is done via physical switch (that fans out the power generated from a power station to a given set of transformers). Currently, the op centers are heavily manual. Changing to the solar, diffuse way would require:
a) a proliferation of op centers
b) an autonomous power routing network
Of course #b is the preferential way to go, but has that even been started? How does the accounting happen? How do people get paid for their excess electricity and from whom? How do you figure out who to bill?
> What's the end result? EROEI calculations beyond > first or second order become quite tricky and > controversial.
of course they are tricky and controversial. But that doesn't mean that they don't have value. And - Cleveland's quibblings aside - it doesn't mean that Odum's emergy calculations are meaningless. Cleveland really is counting the angels on a head of a pin here - what's important about Odum is not that he gives absolute truth, but that he gives a good FIRST ORDER APPROXIMATION of truth.
In part, solar got such a low score when providing base power because its 'transformity' with our current power grid was low. Every single time you burn a gallon of gasoline, or oil, you are leveraging trillions of dollars in sunk costs, simply because our power system evolved from concentrated power sources. And so, when the power networks arose, they radiated outward from centralized sources. And hence they are designed to transfer large amounts of power in a centrally distributed fashion. Oil gets a high emergy score LARGELY BECAUSE IT FITS WELL INTO THIS INFRASTRUCTURE. That's why people are still bothering with oil 'even though' the EROEI is 3.
Like it or not, you have to take this into account when dealing with solar. Solar goes directly against our current energy infrastructure's grain. It is a) low power and b) intermittent. Not only are we going to have to build the solar panels and collectors, but:
research and design and build low impedance
transmission wire.
design, build low-voltage transformers,
design, build low voltage transmission
and load baring towers,
upgrade existing networks to
minimize transmission costs,
research ways on storing solar energy into
energy carriers
develop the infrastructure to handle these
energy carriers
Hence we'd need to reinvent ourselves in a big way in order to use solar energy in load-based systems.
Now, I don't have all the answers, but a) doing all of this sounds like a century long project, not something that happens overnight. b) if you prorate the energy costs of this research, development and infrastructure, and balance it against the amount of energy we get from solar cells, I'd bet you a boatload of money that it puts the overall EROEI of solar below 1 for some time to come no matter how you calculate it.
And if you are right about the EROEI of oil being 3 in the US, the time for doing this is DEFINITELY short, far shorter than what it will take.
What we need is a band-aid to keep going, one that works with our current infrastructure, one that provides base level power in a scalable way. We'd still need to find an easy way to make oil, and build the infrastructure to make that oil (thermal depolymerization springs to mind) and that's where nuclear power and especially breeder reactors come into play. Its really the only thing that we've got left (well... excepting coal. But I shudder to think of the environmental costs of an entire society built on coal).
For breeder reactors fit *very well* into our current paradigm. Right now, the de rigeur choice of load baring power systems are gas turbines, which are 60%+ efficient, can do co-generation, and are portable. They are from 10-150MW in power and go for about $300/kW.
There is no *technical* reason why these couldn't be swapped out for inherently safe breeder reactor turbines, manufactured for the same (or less) money.
They then could be shipped to any location that needs base-level power - especially mining locations and places 'outside the grid' - where they are especially attractive because they don't involve a huge amount of fuel shipments. Newer ceramics technologies could drive costs down, and ultimately lead to higher power densities and greater profit per unit.
Now, I've got to ask you - do you *really* think that solar is going to bridge the high power infrastructure gap and provide us with a viab
I agree with you that simply putting up solar cells for gathering non-base load energy is a viable proposition, even perhaps a profitable one.
But you are making a far more bold claim: that not only can we use the distributed solar cell technology for non-base, localized power, but that we can concentrate it and use it to run cities and factories, refineries and ships and cars.
That is a *far more difficult* claim to make, and sorry, but I don't believe you.
In the LCS of Alsema's study that you gave me, he SPECIFICALLY STATED that he was only talking about production costs, and didn't go into collection costs, efficiencies of storage, or transmission. He basically says 'we're producing x amount of power, we assume its going to have 100% usability, hence we have an EROEI of X'. Which is what I don't want to hear.
What I want discussed is the whole package:
1) design installation 2) collector materials cost 3) collector cost 4) administrative cost 5) operation and maintenance 6) concrete and other building materials 7) structural aluminum and other supporting materials 8) rebar costs 9) wiring costs 10) operational facilities 11) storage costs and efficiencies 12) transmission and transformer costs
*These* are the numbers that I want to read, and I want to read them from an actual, operational site that provides *base load power*. Something like Odum did but updated (btw - the figures were for 1991-1994). If you have such a document, produce it (perhaps Kato has done one).
Until you do so, the figures you mention are all theoretical. Of *course* land use matters, and of *course* storage and transmission costs matter. You *do* know that transmission efficiencies go down by the voltage generated and by distance, don't you? What about the infrastructure for transformers and switches for power generation?
And if you run out of roof space, are you really going to put thin-polymer photovoltaic cells onto parking lots or on the sides of roads? How long do you think those will really last? Who maintains the facilities for power switching? What about breakage?
No - I don't have blinders on, I'm just a tad more skeptical than you. You seem to have this odd idea that its going to be a breeze to replace the centralized infrastructure of the last 100 years with a decentralized untested method of power generation, without any fallback just in case it doesn't work out as rosy as you think.
So - how about it. Do you have an updated, true, end-to-end document on efficiency for a solar installation providing base power that shows a decent end-to-end EROEI? *That's* where I've seen EROEIs of anywhere from 1 to 2...
Ed
(ps - yes, I did make a mistake about the 14% efficient cells, and thanks for catching it. But it still wasn't clear in the entech doc exactly how much area per module there was, so I'm still unsure on what the land efficiency is..)
( pps - and yes, you still are comparing apples to pears with EROEI. I admit my mistakes, you should admit yours. And I'm still interested in whether or not you think its worthwhile to support breeder reactor tech.. )
First, of course it matters whether or not the 'overall (land) efficiency' is 10-40%. It matters by a factor of four. It turns your estimate of 7% of texas (which is IMO too low) to about 28% of texas, just for collection.
Your whole argument is based on the fact that 'we can use solar without the further use of land' - if we need to cover the equivalent of texas/alaska with solar cells, of course this simply isn't true.
Second, look at the entech doc that you gave me, specifically figure 4:
PVUSA long term performance data:
Array DC efficiency (ranges from 2% to 14%)
So, the '30% efficient cell' becomes 14% DC efficiency, before any storage costs are considered, and before DC => AC costs are considered, and hence you need to at least double your land estimates.
Third, look at the EROEI document that you gave me, and look under 'solar':
Note that Odum has given a value of.41 (as in less than one) to solar voltaics. Cleveland gives 1.7 (improvable to 5.1) to 10.1 (improvable to 30.1).
Now, why are the numbers different? It depends on *methodology*. Your paper - the one that cites the EROEIs of 30-75 - is fairly limited in scope when it comes to energy and environmental impact. It even says so in the LCA scope: that they are only going to be dealing with *production* energy costs (its not clear about whether mining costs are included).
Hence, the EROEIs that you give are tailor made to be more rosy than the overall numbers. Cleveland tends to be out in lala land, because he doesn't base his analysis on economic study rather than scientific study(?) which inherently makes upcoming new technologies more rosy than they actually are (ie: all that hype surrounding them). It also didn't help that the study you cited was done right in the middle of a economic bubble, further distorting it.
Which is why I favor Odum's numbers (from 'Environmental Accounting'), which include:
1) design installation
2) collector materials cost
3) collector cost
4) administrative cost
5) operation and maintenance
6) concrete and other building materials
7) structural aluminum and other supporting
materials
8) rebar costs
9) wiring costs
10) operational facilities
And is where he gets the overall EROEI of.41.
Furthermore, they are based on a real, voltaic power installation in Texas, which operated throughout the 1990s, and are based on an audit performed by the operational manager. (he's done a couple more audits like these, and they've all come out about the same - things are slowly improving, but institutional efficiency always lags)
*Thats* where the EROEI comes down - when you take in account all the supporting infrastructure that surrounds the collection source. After all, if you only look at the EROEI of the wellheads, whilst forgetting the refineries, pipelines, and so forth, oil's 'EROEI' comes out to over 200, too.
So when you say that the EROEI of Solar is 75 versus the EROEI of oil being 10-30 and light-water nuclear being 4, that's exceedingly misleading. You really are comparing apples to pears here.
What is more fair is if you do the same analysis for each source of power in the same way, like Odum has done. And in this case, solar comes out fairly poorly (which of course is *why* its.1% of our energy source, even after 130 years of research).
Now - that's not to say I completely agree with Odum. The EROEI he gives -.41 - was on an installation that was intent on providing baseline power (which solar does horribly unwell), and installations that are directly on top of houses for local power supply will probably perform much better (even if there is still the question of storage penalty).
But all this really points to is that solar has its nic
Tell you what. You find me an example of real-world solar cell efficiency that approaches 40%, as well as pointers as to producing them with an EROEI greater than 2, and I'll concede the point.
Not theoretical efficiency, not 'calculated based on baseline', but real and large-scale usage.
My bet is that can't find any. For what its worth - Smil (who has 30 years of energy research under his belt) knows about the non-polymer solar arrays that you talk about, and talks about them in Energies at the Crossroads.
He too mentions the 40-50% efficiencies you mention, but then goes on to mention that when that gets translated into real world use the efficiency goes down to 10-20%, and mentions the upkeep costs.
He then goes on to say that - reasonably, and before any storage costs are considered, that you can get 30 W/m^2 with current technology (as of 2003).
So, anyways, the proof is in the pudding. So perhaps we could come to some agreement if you had some real-world pointers and real-world examples, as well as EROEI studies.
> Hmmm. I take that post as agreement. Obviously > Smil is massaging the truth (and you are starting > to look silly defending this guy, who in 7 posts > is proved wrong again and again - time to read a > broader group of authors?). I'm looking forward to > getting past your FUD, and discussing your other > questions, but first we must get pass the > nonsense.
Why don't you read the guy instead of denigrating him? Using one of your data sources and an off-the-cuff calc, I get 170 W/m^2 - the one you just gave, I get 200. I looked back at the ERBS and it gives me about the same. Its not a show-stopping difference.
What a lot of people don't like (and don't understand) about Smil is that he *doesn't* spin the facts, and hence steps on a lot of people's toes (including probably yours).
He also looks at the big picture, rather than the upfront figures like you are doing. Which is why I like him, because it gives me context when judging an energy source.
First of all - YES I'm aware that W/m^2 is a power measurement per unit area. If we amortized usage of all energy sources across the world, and averaged them out, it would come out to about 12 TW continuous usage.
And 87 PW is the amount of continuous power that the sun outputs that we can use. I like using it better because its simpler and you can get a good generalistic feel about an energy source using it - kwH tends to lend itself towards electricity itself.
Now - what I mind is that you are still double dipping. More than I at first realized. You mention 40% efficient solar collectors, but they are *multijunction* collectors, '500 sun' collectors, that are getting the energy. These are 40% efficient because solar collection is inherently more efficient at greater concentrations of energy.
These both are concentrators, *and* have to swing out a larger surface area to do so at the same time.
Lets use your document to see how much area they need to swing out:
"SOLAR RADIATION FOR FLAT-PLATE COLLECTORS FACING SOUTH AT A FIXED-TILT (kWh/m2/day) Average: 4.3.
"SOLAR RADIATION FOR 2-AXIS TRACKING FLAT-PLATE COLLECTORS (kWh/m2/day) Average: 6.6.
Now, there is only so much radiation per m^2, so it takes approximately 150% as much area to do tracking collectors. So the average *real* efficiency for these things is going to be 40% * 4.3 / 6.6 = 26%. At least before we consider the next part.
For I look later on in the doc and get:
DIRECT BEAM SOLAR RADIATION FOR CONCENTRATING COLLECTORS (kWh/m2/day) Average: 3.4
Notice how the average, 3.4, is less than 4.3? That's the efficiency penalty for the concentration - and for going through the lenses (second law of thermodynamics gets its share). It doesn't tell me whether or not the concentrating collectors were tracking or not (my guess is they were because you need to follow the sun in order to concentrate the light effectively) but we'll assume that it is not (and hence they aren't halving the energy available) So in the best case that's a further penalty. 26% * 3.4 / 4.3 == 20%.
So we are back at 20% efficiency overall. Even if your numbers are correct, 200 W/ m^2 (although that's still questionable) and 200 W/m^2 * 20% = 40 W/m^2.
And this doesn't address moving parts storage, etc, which of course are further drains on the efficiency.
So as I said, you are optimistic by at least a factor of two, most likely more than 2.5 here.
Ok, now as to your second point. You have a point about electricity (storage, transmission, and EROEI issues excluded), but transportation and heating is a different matter.
Given the infrastructure costs that we have invested in the ICE, its not very likely that these are going to go away any time soon (40+ yrs). So we are going to need to produce either natural gas or oil in large amounts to feed the current fleet of cars and various industries (including the photovoltaics industry which would need the plast
I avoided anything that was tracking, took all the fixed rates, added them up together, averaged them,and multiplied by 365.
Unlike you, however, I'm trying to take into account the various inefficiencies and penalties that you get from using solar power - from penalties for storage, penalties for conversion into other energy carriers, penalties for the infrastructure involved in this conversion, penalties for transmission, and penalties for handling peak load.
And of course penalties for all the energy used to build and maintain the solar cells and the architecture behind them in the first place. I've seen EROEI's from anywhere less of one to about 3, taking the whole infrastructure of solar cells into account.
So yes, we disagree pretty much from the get go. Trackers are a red herring - there is only *so* much power available to gather from any given spot, and trackers just save on material costs.
I could calculate again, but we are essentially calculating different things. Your experiment is quaint at best, and misleading and false at worse. Money is a problem, yes, but so is transmission, storage, inefficiencies in capture, inefficiencies in maintenance, etc. etc. etc.
*That's* where I get an infrastructure the size of texas. By your naive calculations you are already up to 7% - double that, and that's more realistic IMO.
For example, the solar concentrators which you talk about have to have moving parts to get that 40% efficiency because they need to track the light source to concentrate the energy. These parts, whilst efficient, take an energy penalty of their own, and due to the need to directly track the sun point source in the sky, require a larger amount of area per device than you imply.
Average this out, and you're back to getting about 20-60 W/m^2 of energy for solar - which is the reference figure that I've seen everywhere except for you.
This of course is *before* any of the efficiency penalties that I've talked about.
You also are naively using 3 * 10 ^ 13 kW as our total energy source that would need to be replaced - the difference is that 90+% of that energy is in a form that we can directly use - natural gas for heating and gasoline for burning, coal for making steel, etc. You therefore take a penalty if you want to convert the solar into these forms of energy.
Lets take a different tack - overall, the earth intercepts at its surface about 87 PW of energy average. We use 12 TW (source: 'energies' by smil). That means that we use 1/7200 of the total solar flux of the planet. At 100% efficiency, therefore we would need 1/7200 of the earth's surface covered in solar cells - and since over 70% of the earth's surface is covered by water, approximately 1/2200 of the continents.
Since most of the solar needs to be converted to usable forms, I've been using a 60% penalty, which means that about 1/1360 th of the earth's continents would be needed to be covered by solar cells if everything was 100% efficient.
This equals 225,000 km^2. At 40% efficiency, this equals 562,000 km^2.
Since the current infrastructure is approx 290,000 km^2 - and of that, around 2% of that is energy production - the costs for the infrastructure in just collecting the solar power are bound to be more expensive than the current costs.
Just collecting is - even at your rather optimistic calcs - over 100 times more costly in terms of real estate than our current scheme!
And that is not even considering the *other* infrastructure involved, or the inefficiencies that I've talked about (and you haven't responded to). Or the low EROEI.
Anyways, look, I have NO problem with solar power. I hope that it goes gang-busters, I really do. You have yet to convince me. You just happen to be in solar technology before scalability issues become a factor.
But go ahead, lets stop with this area issue. You said that you'd like to respond to the other issues, so please do.
I think I see where things are off, and where our numbers differ. They do in fact reconcile. You CANNOT USE A TRACKING MECHANISM TO DETERMINE SOLAR POWER DENSITY.
For if you do, you are basically double-dipping your calculations.
Figure: If a plate is tracking the sun, it is swinging out an area greater than the plate's area itself (for as it tracks, it gathers energy that would otherwise miss). In the process IT BLOCKS OTHER PLATES THAT WOULD HAVE OTHERWISE GOTTEN THE SAME ENERGY.
If you take, say the numbers for Birmingham, AL over the last 30 years,
You'll notice that the flat collectors get an average of 3.5 kwH/m^2/day, which turns into about 165 W/m^2.
Its the dual tracking collectors that get the large amounts that you are talking about. However, the tracking collectors can't be placed directly next to each other, because they cast a SHADOW on each other.
My guess is that if you take this shadow into account, the benefits you get from tracking are greatly reduced (not eliminated because of better conversion efficiency) and the total that you CAN get is approx, on average 170W/m^2 (which goes along with the satellite data)
This is the only explanation that makes sense. My satellite numbers aren't lying, neither are your numbers - they are just double dipping when you make the assumption that they can be applied to large areas in 100% coverage.
Hence, the 100*100 sq mile solar concentration is way too small, by an order of magnitude or so, and we are still talking about an infrastructure the size of texas.
1) your first link doesn't work (its broken) In fact, I couldn't even reach medc.nrel.gov.
2) second, the number I gave (140 W/m^2) is from the Earth Radiation balance satellite. I'm not sure what's going on here, but I'd hardly expect that group to lie.
3) the storage, transmission, and maintenance costs (especially from EROEI) makes it *much* more inefficent than just a 'third'.
Here's another example of efficiency loss - the sun is highly variable in the year, you hardly capture any power at all during the winter months. Hence, the need to store the electricity captured from the summer for months on end. Hence, high inefficiency.
You want to prove that solar is a decent replacement to me, address the efficiency concerns AFTER you have collected the solar energy, and the EROEI that you get from making solar cells. All it takes is an overall efficiency of 10% (which is not at all hard given the cost of storage, maintenance, and manufacturing, battery and or energy carrier construction, and solar cell replacement ) and you are back to an infrastructure the size of texas.
And that is even with your numbers, which I think are *highly* optimistic.
Elsewise you really are looking at the world through rose-tinted glasses.
Anyways, I'll look for a third party that can corroborate your numbers, since I can't check them for myself. I did look at the NREL and sharp links, but they didn't go into *any* of these issues. They seem stuck on the 'look at us catching 40% of the sun's energy, isn't that cool!!!' or 'huge potential market opening up!!!' stage of things. You have to consider the entire energy life cycle, or your analysis is meaningless.
Slashdot Mirror
(ppps - wow. I just looked at the incentive bill for Comprehensive Energy Legislation. It puts nuclear power under the same umbrella as wind, solar, etc.
It doesn't give the same credit to nuclear as it does to wind - ie: getting a flat-rate 1.2 cent/kwH production credit to wind, plus a 1.8 cent/kwH tax credit, but it *does* extend the 1.8 cent/kwH tax credit to nuclear for the first 6,000 MW.
Not quite the same, but whoever builds the first advanced nuclear plants is going to get a windfall - they truly are going to be able to 'print money'. If the next gen plants generate at 1 cent/kilowatt hour operating cost, then as I see it, at 40% tax rate, they are going to be able to generate electricity for free and sell it for sheer profit...
Now if they can only streamline the hurdles for getting approved.
Ed
1 GW is the amount of energy produced.
oops.. make that the 'amount of power produced'. The calculation is still correct, though.
Ed
said that the operating costs for nuclear energy are about 1.5-2 cents a kilowatt hour.
.9 * (30 years) * (1 GW)
.9 is the capacity factor and 5 billion dollars is the startup cost, and 1 GW is the amount of energy produced. Calculate this, and it gives approximately 10.25 billion dollars cost for running that power plant. Divide that by the number of kilowatt hours produced:
.9 * 30 years)
.9 * (60 years) * (1 GW)
Right, so the way you prove that nuclear is cheaper than the alternatives is by ignoring all the costs except for the one that makes your argument look good.
the '12 cent/kWH' figure that you get factors in a lot of the original R&D cost
Now you're making stuff up again.
Ok, lets talk some figures here. I'm not 'making stuff up' - my argument is that the statement 'Nuclear is Expensive' is a inherently useless, dumb statement. Its like saying 'cars are expensive' or 'computers are slow'. And that the 12 cents/kwH figure that you cite lumps everything together into one big basket, rather than looking at things critically.
Is nuclear expensive? If you're talking about 1st generation nuclear power I'd agree with you (it being done primarily for military reasons). If you're talking about 2nd generation (ie: 40 year old technology) nuclear power, it really depends on what you include in your scope. And how long the plants last. If you've got a second generation nuclear power plant (like the ones at Vogel) that are producing electricity at a cap of 2 cents/kWH, and that cost about $5000/KW to build, the economics of running the plant become pretty straightforward:
5 billion dollars + (2 cents/(kilowatt hour)) *
where
(10.25 billion dollars) / (1 GW *
and you get about 4.33 cents / kilowatt hour (if you don't believe me, believe frink). If they last 60 years:
(5 billion dollars) + (2 cents/(kilowatt hour)) *
that's approximately 15.5 billion dollars for the running of the plant. And at a total cost of about 3.3 cents/kWH.
So - where does Diesendorf get his 12 cent/kwH number? Well, first of all, there is the insurance 'subsidy' that you talk about (which a large part is the 'Price Anderson act' which protects nuclear producers with a cap on disasters) and the R&D 'subsidy'. If you factor in that research used to develop those plants, plus failed reactors (ones that started and never completed), plus the past R&D going back to 1945, you get (I'm guessing) about 12 cents/(kilowatt hour). And that's the number that you are apparently attracted to.
But that's my point - these costs (failed reactors, botched administration) are unlikely to re-occur. The nuclear plants *themselves* are competitive. And the R&D isn't useless - like say, the cost of supporting a war to keep ourselves hooked on oil, or the cost to cleanup the environment and the damage to agriculture. Its not money wasted, its a one time charge spent for plants to come that are now being built in (ironically) india and china with our R&D. If you take the millions of man-years used in us perfecting nuclear power and put it into engineering practice, you drop the startup cost from $5000/KWH to approx $1000/KWH, and the running of the plant from 2 cents/kwH to about 1 cent, which is cheaper than your average coal station. Which is what is being built in india at this very moment (the advanced westinghouse reactor).
So, in *further* answer to your question, 'is nuclear expensive'? No, not *modern* nuclear power. I'd also argue that most of the 'subsidies' that are mentioned in these calculations aren't really subsidies; that they lead to bigger and better things, and to a point where the 'subsidies' end. Sort of like the subsidies for wind and solar energy, which you seem so fond of.
Ed
(ps - I didn't mention Lovins because he bolstered my argument; I mentioned Lovins to show that I had read more than
And when you turn them on, its like printing money - it costs on the order of 1.5 - 2 cents/kW to maintain and run them.
It costs 12c/kWh for nuclear power. I've no idea what hole you pulled your figure from, but I could take a wild guess.
First of all, your "original comment" was 'Nuclear is expensive and nobody wants a nuclear reactor in their backyard. End of discussion'
I'm glad you've finally noticed. I've no idea why you've put "First of all" in front of that sentence though, because you're the one who should be apologising for not paying attention to what I said.
I'm feeding a troll here, but here goes -
1) I said that the operating costs for nuclear energy are about 1.5-2 cents a kilowatt hour. As for the base cost, like I said its approximately $1000/kWH. And I got that figure from both an operational plant (the Vogel twin plants), and the profit figures from nations who rely on nuclear electricity (france, sweden).
2) the '12 cent/kWH' figure that you get factors in a lot of the original R&D cost which is not necessary to repeat, as well as a lot of the political costs that were associated with that R&D. It also has to do with the assumption that you are only going to be running the plant for 20 or so years, and it looks like most plants will be running for a lot longer than that (about 60 in fact). In practice, the true operating cost is about 1.5 cents/2 cents an hour. And - unlike wind - there is no subsidy for the operation of nuclear plants. Do you really thing the Vogel plants would be selling nuclear power at 5 cents per kwh, at a 7 cents loss? Do you really think that *france* would be selling 1GW-9GW nuclear power for 8 cents/hour to businesses (or to other countries like italy) for a 4 cent loss/kWh?
And that doesn't even count 3rd generation/4th generation nuclear reactors, which have orders of magnitude improvement over current, 1970s technologies.
Why isn't there more nuclear power? Again, its politics. Its ironic, because so much effort has gone into fighting nuclear power buildup that plants that to *real* damage, like coal, are sailing through and the greens are actually *retarding* sustainable development. Its no wonder that true greens (like James Lovelock) are amongst the strongest supporters of nuclear power.
3) what you *say* is worse than useless when you don't back it up with figures, references, reasoned argument and discussion. The fact of the matter is that I didn't see your original comment at first, just your second mouthed platitude (the more you look into it the more.. etc. etc.). So, you tell me to go to the first comment, which I dutifully do so, and see that its another useless mouthed platititude. Doesn't do me - or anyone else - any good.
Read some books. Please. Start with those I mentioned, and then read Vaclav Smil, Amory Lovins (although he's a bit 'pie in the sky', he's got some good ideas on efficiency), and maybe deffeys. I'm not 'nuclear zealot' - I am not in the nuclear industry, and have no vested interest (other than wanting to keep living with the comforts of civilization). And I think that wonders can be done with solar and wind - to shave load off of peak power. But *replacing* peak power is a totally different thing - Denmark is finding real barriers in doing this, even though they are under political mandates to try to totally rely on renewables - and their spot price for electricity is twice as high as the countries with supposedly 'too expensive' nuclear power.
Ed
Why would I do that? This is why I am not wasting my time discussing this with you. My primary comment - as demonstrated in the SUBJECT - is that nuclear power is expensive. That is an indisputable fact as even the nuclear power lobby admits that it is only profitable in the US when subsidised, and is best suited for countries that do not already have cheap plentiful supplies of fossil fuels (such as France).
First of all, your "original comment" was 'Nuclear is expensive and nobody wants a nuclear reactor in their backyard. End of discussion', which is just as fatuous (and empty headed) as the comment that I responded to.
Second of all, I could argue with that comment (as devoid of info as it is) in three ways. First, if you look at the countries that have large amounts of nuclear power, their energy cost per-kWh is on par with those who rely on coal, etc. Source here(look at France && Sweden && Germany for examples). And those countries which rely heavily on renewables tend to have *twice* as high an energy cost (look at the table for Denmark for example!)
Second, I could say that we already are heavily subsidizing coal, natural gas and oil in the form of military encounters, damaged productivity in agriculture, damaged health, loss of life, and so on - and that with large-scale solar power that this subsidy would continue - although not as great as it would be with coal. For it takes coal/oil/etc to *build* solar panels and solar towers, for an energy source which is highly disperse and needs a lot of infrastructure to put 'online'. That's where the figure of 200 g/kWh capacity is a killer. Its 25% that of coal - which means if we scaled up solar, and used 4 times as much electricity as we did now - which is not a stretch considering the amount of population in the world and how hungry they are for energy - we'd still be generating the same amount of greenhouse gases and associated emissions we are today. Which is not an acceptable solution.
And third - and finally - I'd say that you have a woeful misunderstanding of the economics of power plants. All the pain that we've caused ourselves on disposal and lots of the overruns on the building costs of current power plants are caused by politics, pure and simple. Its easy to make anything you want as expensive as you want if you throw hurdles, lawsuits and FUD at the problem. Take away the FUD and the things are relatively in-expensive to build - current LWR around $1000/kW to build in comparison with coal being $1100/kW. And when you turn them on, its like printing money - it costs on the order of 1.5 - 2 cents/kW to maintain and run them. The source of fuel - uranium and its preprocessing costs - is close to free because it is so energy dense. And the important thing - especially in comparison with coal/natural gas/solar/wind is that their capacity factor is *great*. Which means they run almost all the time - circa 90% around the year 2000 in the US, as compared to circa 60% for coal and circa 30% for wind. Which means they can be used easily for industrial base load.
And fourth - finally,finally, there are magnitudes of improvement left in new designs. As I said in my previous post, self-contained, automatic, passively safe breeder reactors such as the sstar are designed to slash the amount of source material by 70 times (by using U-238 instead of U-235) and to cut the capital costs in manufacture by about 2 to 3 times due to new materials manufacture.
So yes, the economics of nuclear power plants is quite cheap once you get the politics out of it. As for costs of decommission and storage, those are greatly reduced if you simply let the radioactive actinides disintegrate and then move the relatively inert radioactive waste - or do what the french do and simply reprocess the fuel, and embed the residue in volcanic glass. We produce about 2300 metric tons of waste a year to deliver about 1/5t
I used photovoltaic cells because that is the most commonly known technology.
In the commercial space, superheated water and thermal towers are the most commonly used technology.
In the commercial space, solar provides far far less than one percent of TPES, and when it does it is in niche areas. Why do you think that is?
Now, I would suggest stopping the mouthing of empty platitudes and do some research of your own.
I have done far more research on this topic than you could imagine.
Great.. then spill it. Show me a counter-example study showing g/CO2 figures from cradle to grave for different solar energy technologies, and I might believe you. Otherwise, what good is your 'I have done far more research on this topic than you could imagine' statement?
Its a discussion group, so you know.. discussion might be nice...
Ed
Just curious, I looked at your website, and you don't seem to mention 3rd/4th generation nuclear power.
In particular, I'm interested in the use of nuclear power *not* for electricity generation, but for heat generation, to replace natural gas for such activities as:
* ethanol production from cellulose
* hydrogen production from coal
* steam production for oil sands development
To me, the biggest story about nuclear is not going to be the electrical plants (although those are important too) but in the creation of portable, off-the-grid thermal generators. For example (as seen on slashdot), the sstar, which I *hope* they are planning on converting to use for thermal use. 100 MWe == 350 MWth == a *lot* of ethanol produced.
Ed
I used photovoltaic cells because that is the most commonly known technology. Thin-film PV cells (eg - copper indium diselenide and cadmium telluride) use less materials/square meter, but their conversion efficiencies are less and the infrastructure needed to support them is more; the amount of CO2 emissions per watt-hour (200 g/Wh) is pretty much a constant amongst the different technologies.
source: here
Now, I would suggest stopping the mouthing of empty platitudes and do some research of your own.
People protest nuclear because NuclearIsBad(tm). Education is the only way to combat this.
The more education you have regarding nuclear fission power, the more you protest.
No offense, but this is total and utter BS. Here are just a few very well reasoned out proposals/cases for nuclear power:
here
here
here
In short, the amount of C02 that even circa 1950s nuclear power plants emit is 5 times *less* than wind, 50 times *less* than solar (where do you think all that material to make solar cells comes from anyways?). As for cost, new nuclear power plants are about the same as natural gas turbines (which are as cheap as it gets) and have orders of magnitude left for improvement with new materials. As for safety, coal with its byproducts kill about 50,000 people yearly. Since its the extraction of materials blamed for premature deaths, if solar was scaled up to the same level as coal was, it would kill about 1000-2000/yr.
Face it; every source of energy has its risks. With nuclear, most of those risks can be mitigated because you get so much bang for your buck with the fuel that you can watch the risks carefully. Passively safe systems make things even more secure. Its one of the biggest irony in history that the Greens are fighting tooth-and-nail probably the most environmentally conscious enery source of all. (outside of conservation)
horos
I hope you are being ironic here; Saying "But those evil arrogant Romans got their comeuppance is sort of like saying the "Tzars sure got them comeuppance by the communists and their secret police".
Yeah sure.. but lots of people had to die in both cases, of course.
I really don't want a fractured internet. It would suck, and it *could* happen.
Its not as risky as people are thinking..
1) the sales of the DS and of touch-games are pretty impressive.
2) the controller itself has an 'expansion' slot, into which you can put a standard controller (for backwards compatibility to play existing games)
I've got to hand it to nintendo. What a coup. I'm definitely getting one of these.
Ed
(ps - to hear it from the horses' mouth.. go to here
Very informative presentation by the CEO of nintendo. )
> Your calculations are reasonable; you don't make
.2 == 6 watts of waste heat (the rest being translated into electricity, which eventually would end up as heat, but in a diffuse sense). And you did start by saying that you would need kilograms of the stuff to run a laptop, which isn't true.
;-)
> your mistake until the end when you come up with
> 15 grams (2200 Curies) of Sr-90 and figure
> that's just great for personal electronics with
> a small form factor!
> I don't know if you move in Sr-90 circles, but
> 15 grams is a lot of pure Sr-90 to have. It
> would emit 30 W of heat.
well, to be totally fair it would emit (30 W *
Anyways, 6W of heat is a bit, but not anything a decent cooling system couldn't handle (even air convection would do it). You would need to leave your laptop on all the time, however.
> Its radiation would kill you in a few seconds,
> if you were directly exposed to it. It would
> deposit in your bones like calcium if it ever
> escaped its lucite container. Its weight would
> be the least of your concerns.
well, the trick is not to get directly exposed then, isn't it..
Anyways, the curie is not the unit of measurement about health, but the sievert.
You're right though, I didn't do my homework about this and if the battery escapes its containment you are pretty much toast.
I guess its a question of how much faith you have in the ability of people to make an 'spill proof' container.. Given a system where the battery was bolted to the unit (and hence undetachable), and not requiring aqua chemistry simplifies the engineering problem. The working conditions would be difficult though (you'd probably need a completely automated factory to do it)
I did the math however, and it would be feasible to mass produce these - the nuclear industry has produced 40,000 metric tons of high level nuclear waste since its inception, and of those 40,000 tons, approx 500 out of every 10000 particles is sr-90.. So, we've got:
40000 * 500/10000
or 2000 tons of the stuff. Divide that by 15 grams and we could make 130 million of said batteries..
Increase the share of nuclear power by a factor of 10 and you get 1.3 billion batteries. Hell, at that level you could probably provide enough electric power for cars..
Ed
> You obviously haven't done the calcs. Efficiency
. htm
.17 mol
> is not even the major problem here. The problem is
> that ionizing radiation is an extremely deadly
> form of energy- a median lethal dose in humans is
> about several watt seconds per kg of body mass.
> And to get even a tiny little bit of heat, you
> need an enormous amount of radiation.
I'm not sure, but I don't think you know the difference between alpha decays, beta decays and gamma rays. As per:
http://www.ehs.ucsf.edu/Manuals/RSTM/RSTM%20Chap1
"1 cm of lucite is enough to shield the most active beta emitters (32 P, 90 St)".
Beta decay rapidly breaks up when it encounters matter.
As for tritium, its beta decay is so weak that it could be stopped by the single layer of dead skin cells in your body.
And your math is off. To power a 15 Watt laptop, using Strontium-90 as a source, and assuming 50% efficiency, you need:
15 Watts * 2 efficiency factor / 2.3 MeV
equals
81 x 10 ^12 decays / second.
Given Strontrium's half-life is 28 years, this means:
8.83 x 10 ^8 secs = ln 2 / lambda
lambda = 7.84991 * 10 ^ -10;
N0 - 8.1 x 10^13= N0 * e^-lambda * 1 sec
Solving for N0, we need
1.03695 x 10 ^23
atoms of the stuff, or
Given Strontrium's molecular weight is 90 g/mol, this equals approximately 15 grams. And given that batteries weigh in excess of a pound, this is highly doable, if it weren't for the cost of strotrium.
In fact you mentioned it yourself - Pu-238 (with a halflife of 87 years) gets a half-watt a gram, which means - if you had a 'nuclear battery' that was 80% efficient, you'd need:
15/.5/.8 = 38 grams
of plutonium to run a laptop.
However, of course you wouldn't use this, because plutonium is an alpha emittier, and hence is a lot more damaging than the betas. Hence, the use of betavoltaics in places like pacemakers, etc.
Ed
.. that's the old technology, which is at .1-.5 % efficient (that's right .001-.005)
Because of this inefficiency, there is lots and lots of waste heat, which causes the problems that you speak of.
If the nuclear battery mentioned on slashdot truly reaches its supposed 200 times efficiency, this is *80%* or thereabouts efficient, which means that there is a lot less material to radioactively decay.
Second of all, the batteries studied operate off of beta decay, which essentially means that they give off an electron. Electron == electricity, so the chances of needing lots of cooling equipment are probably not true.
So.. I haven't done the calcs yet, but nuclear batteries would probably be more feasible than you are thinking from an engineering standpoint.
On the other hand, tritium and strontium-90 are *damned expensive*, and would be so unless there was a circuit that connected supply (the nuclear power industry) to demand (the batteries) and even then, its doubtable that the nuclear power plants could produce enough strontium and tritium to keep everybody happy...
horos
no, plutonium is not 'highly radioactive and among the most lethal toxic substances known to man'. See:
http://www.fortfreedom.org/p22.htm
where bernhard cohen (sp?) offered to eat as much plutonium as nader did caffeine on a TV show, just to disprove this DAMN DESTRUCTIVE MYTH/PIECE OF MISINFO for all time.
Nader didn't take him up on it.
> It just occured to me you don't know how
> transformers work. They step up one direction, and
> step down the other. Even if I generate only 240
> volts AC with PV using a small residential
> inverter it goes through my transformer and steps
> up to medium voltage distribution lines (4-20kV),
> it can go from there through a distribution
> station transformer to 230kV+ for long distance
> transmission power lines. Transformers are TWO
> WAY. There is no low voltage transmission
> requirement here.
Wrong. Transformers *can* be two way and be converted to be two way, but by legacy they are primarily either 'step-up' transformers or 'step down' transformers. And by far, the transformers that you see on the top of telephone poles (in america at least) are 'step down only'.
It is up to the *utility* to provide the infrastructure that you need in order to feed back solar into the power grid and *only on demand*. And of course they are fighting this tooth and nail. Hence they hardly have done *any* of this work.
To this end, they usually convert the pole transformers to be dual and rig it so that if you generate excess power, your metering runs backwards. (since hardly anyone has solar, this power then usually makes it to the next house down the block before being stepped down)
Note that this is a far, far cry from using solar to satisfy base demand. You need to:
a) complete the upgrade of the power grid to
get the power from many small residential
systems to one or multiple
distribution station transformers
b) figure out how to dynamically route this
power to where it is needed.
amongst other things (like quality control, billing, etc.)
These steps are not easy. Like I said, right now #b is a heavily manual process, even when the load stations are right next to the consumers of that electricity - and even the slightest mismanagement can cause blackouts and other system failures.
The problem grows exponentially as the power is intermittent and spotty - you really need to have the end-points which need the power be able to talk to the transformers to see exactly how much current they have at one given time. And of course this hinders centralized control - if a power line needs to be worked on, how do you guarantee that no current is flowing in it? And how do you tell the network that that line is down?
The problem reminds me a lot of the internet - and notice that the internet was a trillion+ dollar investement. And, after 30 years, we *still* can't rely on the internet for crucial services.
Let me ask you a simple question - do you really believe that this is going to be a dead simple process, and that the solar market will sustain 40% a year growth for 30 years?
Not at all. You're still underinformed about PV
> technology. PV itself is low voltage, however
> its converted to AC using an inverter (same
> thing for fuel cell BTW) and transmitted over
> the grid just like normal electricity. Done all
> the time, every day, all over the world. 93-98%
> efficient (I know, we haven't accounted for that
> yet in our land calcs). PV very nicely
> integrates into the current electricity
> structure - no change needed The EROI numbers
> given already account for the inverter energy
> inputs, and its conversion efficiency. The EROI
> stands as it is.
no.. I'm quite aware of how the PV cells work, thank you very much. However, it sounds like you are uninformed of how 'the grid' works.
Like I said, what you state up above is perfectly fine when it comes to local generation and electricity use. In fact, that's what the calcs were tailor made for - they hold up for solar as a LOCAL energy source, not on a high-power grid. Its sort of assumed everywhere that solar's going to be used locally, not for base-level power. And that's what I was asking for EROEI studies on - *base level* solar power. Fine - you disagree with Odum. Show me a study that contradicts his findings, with the assumption that you are using solar power for base-level power.
For the 'grid' is a misnomer - its NOT a two-way street - its designed for centralized distribution, and has been so for approximately 120 years. And that's where the transformity of solar falls down.
"the grid" is designed for high voltages over long distances - both for efficiency reasons, and for facilitated control. Hence, you *can* send electricity 'back to the grid', but its not going to get very far unless you have the infrastructure to support it.
FYI, here's a small primer on how large scale generation works: first there is a generation source, producing low (5-20 kV) voltage electricity, which is then 'stepped up' by a transformer to above 100 kV over alternating current. This is then sent through power wires to a substation which then 'steps down' the current to residential levels (there may be multiple transformers up and down). That's where the 7% efficiency rating comes in.
So - how does this relate to solar? The dimensions of the wires are low, the voltages are low. Hence the efficiency is when transferred through the network is low.
This inefficiency (the one you quoted) is probably accurate for local area networks, but transferring across towns, cities, or even states is bound to be much less efficient unless a boatload of transformers - are used to both step up the excess power, and to step down the excess power on the remote locations that might need it. It just makes common sense - you can't expect to transfer 100 volts cross country without serious inefficiencies - unless of course, superconducting, exceedingly cheap wire comes about - which I'm not holding my breath on.
Then of course, there is the question of routing. How do you know where to send the energy? If say, Minnesota has a huge snowstorm (covering all the solar cells), how does new mexico know that its supposed to send Minnesotans power, rather than Minnesota getting it locally? And if a snowstorm does happen, how do the solar cells get maintained and cleaned?
Right now, its relatively easy - and we *still* can't get it right. Base power load stations are assigned certain areas, and each routing is done via physical switch (that fans out the power generated from a power station to a given set of transformers). Currently, the op centers are heavily manual. Changing to the solar, diffuse way would require:
a) a proliferation of op centers
b) an autonomous power routing network
Of course #b is the preferential way to go, but has that even been started? How does the accounting happen? How do people get paid for their excess electricity and from whom? How do you figure out who to bill?
Solar does *n
> What's the end result? EROEI calculations beyond
> first or second order become quite tricky and
> controversial.
of course they are tricky and controversial. But that doesn't mean that they don't have value. And - Cleveland's quibblings aside - it doesn't mean that Odum's emergy calculations are meaningless. Cleveland really is counting the angels on a head of a pin here - what's important about Odum is not that he gives absolute truth, but that he gives a good FIRST ORDER APPROXIMATION of truth.
In part, solar got such a low score when providing base power because its 'transformity' with our current power grid was low. Every single time you burn a gallon of gasoline, or oil, you are leveraging trillions of dollars in sunk costs, simply because our power system evolved from concentrated power sources. And so, when the power networks arose, they radiated outward from centralized sources. And hence they are designed to transfer large amounts of power in a centrally distributed fashion. Oil gets a high emergy score LARGELY BECAUSE IT FITS WELL INTO THIS INFRASTRUCTURE. That's why people are still bothering with oil 'even though' the EROEI is 3.
Like it or not, you have to take this into account when dealing with solar. Solar goes directly against our current energy infrastructure's grain. It is a) low power and b) intermittent. Not only are we going to have to build the solar panels and collectors, but:
research and design and build low impedance
transmission wire.
design, build low-voltage transformers,
design, build low voltage transmission
and load baring towers,
upgrade existing networks to
minimize transmission costs,
research ways on storing solar energy into
energy carriers
develop the infrastructure to handle these
energy carriers
Hence we'd need to reinvent ourselves in a big way in order to use solar energy in load-based systems.
Now, I don't have all the answers, but a) doing all of this sounds like a century long project, not something that happens overnight. b) if you prorate the energy costs of this research, development and infrastructure, and balance it against the amount of energy we get from solar cells, I'd bet you a boatload of money that it puts the overall EROEI of solar below 1 for some time to come no matter how you calculate it.
And if you are right about the EROEI of oil being 3 in the US, the time for doing this is DEFINITELY short, far shorter than what it will take.
What we need is a band-aid to keep going, one that works with our current infrastructure, one that provides base level power in a scalable way. We'd still need to find an easy way to make oil, and build the infrastructure to make that oil (thermal depolymerization springs to mind) and that's where nuclear power and especially breeder reactors come into play. Its really the only thing that we've got left (well... excepting coal. But I shudder to think of the environmental costs of an entire society built on coal).
For breeder reactors fit *very well* into our current paradigm. Right now, the de rigeur choice of load baring power systems are gas turbines, which are 60%+ efficient, can do co-generation, and are portable. They are from 10-150MW in power and go for about $300/kW.
There is no *technical* reason why these couldn't be swapped out for inherently safe breeder reactor turbines, manufactured for the same (or less) money.
They then could be shipped to any location that needs base-level power - especially mining locations and places 'outside the grid' - where they are especially attractive because they don't involve a huge amount of fuel shipments. Newer ceramics technologies could drive costs down, and ultimately lead to higher power densities and greater profit per unit.
Now, I've got to ask you - do you *really* think that solar is going to bridge the high power infrastructure gap and provide us with a viab
I think you misunderstand me...
I agree with you that simply putting up solar cells for gathering non-base load energy is a viable proposition, even perhaps a profitable one.
But you are making a far more bold claim: that not only can we use the distributed solar cell technology for non-base, localized power, but that we can concentrate it and use it to run cities and factories, refineries and ships and cars.
That is a *far more difficult* claim to make, and sorry, but I don't believe you.
In the LCS of Alsema's study that you gave me, he SPECIFICALLY STATED that he was only talking about production costs, and didn't go into collection costs, efficiencies of storage, or transmission. He basically says 'we're producing x amount of power, we assume its going to have 100% usability, hence we have an EROEI of X'. Which is what I don't want to hear.
What I want discussed is the whole package:
1) design installation
2) collector materials cost
3) collector cost
4) administrative cost
5) operation and maintenance
6) concrete and other building materials
7) structural aluminum and other supporting materials
8) rebar costs
9) wiring costs
10) operational facilities
11) storage costs and efficiencies
12) transmission and transformer costs
*These* are the numbers that I want to read, and I want to read them from an actual, operational site that provides *base load power*. Something like Odum did but updated (btw - the figures were for 1991-1994). If you have such a document, produce it (perhaps Kato has done one).
Until you do so, the figures you mention are all theoretical. Of *course* land use matters, and of *course* storage and transmission costs matter. You *do* know that transmission efficiencies go down by the voltage generated and by distance, don't you? What about the infrastructure for transformers and switches for power generation?
And if you run out of roof space, are you really going to put thin-polymer photovoltaic cells onto parking lots or on the sides of roads? How long do you think those will really last? Who maintains the facilities for power switching? What about breakage?
No - I don't have blinders on, I'm just a tad more skeptical than you. You seem to have this odd idea that its going to be a breeze to replace the centralized infrastructure of the last 100 years with a decentralized untested method of power generation, without any fallback just in case it doesn't work out as rosy as you think.
So - how about it. Do you have an updated, true, end-to-end document on efficiency for a solar installation providing base power that shows a decent end-to-end EROEI? *That's* where I've seen EROEIs of anywhere from 1 to 2...
Ed
(ps - yes, I did make a mistake about the 14% efficient cells, and thanks for catching it. But it still wasn't clear in the entech doc exactly how much area per module there was, so I'm still unsure on what the land efficiency is..)
(
pps - and yes, you still are comparing apples to pears with EROEI. I admit my mistakes, you should admit yours. And I'm still interested in whether or not you think its worthwhile to support breeder reactor tech..
)
First, of course it matters whether or not the 'overall (land) efficiency' is 10-40%. It matters by a factor of four. It turns your estimate of 7% of texas (which is IMO too low) to about 28% of texas, just for collection.
.41 (as in less than one) to solar voltaics. Cleveland gives 1.7 (improvable to 5.1) to 10.1 (improvable to 30.1).
.41.
.1% of our energy source, even after 130 years of research).
.41 - was on an installation that was intent on providing baseline power (which solar does horribly unwell), and installations that are directly on top of houses for local power supply will probably perform much better (even if there is still the question of storage penalty).
Your whole argument is based on the fact that 'we can use solar without the further use of land' - if we need to cover the equivalent of texas/alaska with solar cells, of course this simply isn't true.
Second, look at the entech doc that you gave me, specifically figure 4:
PVUSA long term performance data:
Array DC efficiency (ranges from 2% to 14%)
So, the '30% efficient cell' becomes 14% DC efficiency, before any storage costs are considered, and before DC => AC costs are considered, and hence you need to at least double your land estimates.
Third, look at the EROEI document that you gave me, and look under 'solar':
eroei
Note that Odum has given a value of
Now, why are the numbers different? It depends on *methodology*. Your paper - the one that cites the EROEIs of 30-75 - is fairly limited in scope when it comes to energy and environmental impact. It even says so in the LCA scope: that they are only going to be dealing with *production* energy costs (its not clear about whether mining costs are included).
Hence, the EROEIs that you give are tailor made to be more rosy than the overall numbers. Cleveland tends to be out in lala land, because he doesn't base his analysis on economic study rather than scientific study(?) which inherently makes upcoming new technologies more rosy than they actually are (ie: all that hype surrounding them). It also didn't help that the study you cited was done right in the middle of a economic bubble, further distorting it.
Which is why I favor Odum's numbers (from 'Environmental Accounting'), which include:
1) design installation
2) collector materials cost
3) collector cost
4) administrative cost
5) operation and maintenance
6) concrete and other building materials
7) structural aluminum and other supporting
materials
8) rebar costs
9) wiring costs
10) operational facilities
And is where he gets the overall EROEI of
Furthermore, they are based on a real, voltaic power installation in Texas, which operated throughout the 1990s, and are based on an audit performed by the operational manager. (he's done a couple more audits like these, and they've all come out about the same - things are slowly improving, but institutional efficiency always lags)
*Thats* where the EROEI comes down - when you take in account all the supporting infrastructure that surrounds the collection source. After all, if you only look at the EROEI of the wellheads, whilst forgetting the refineries, pipelines, and so forth, oil's 'EROEI' comes out to over 200, too.
So when you say that the EROEI of Solar is 75 versus the EROEI of oil being 10-30 and light-water nuclear being 4, that's exceedingly misleading. You really are comparing apples to pears here.
What is more fair is if you do the same analysis for each source of power in the same way, like Odum has done. And in this case, solar comes out fairly poorly (which of course is *why* its
Now - that's not to say I completely agree with Odum. The EROEI he gives -
But all this really points to is that solar has its nic
Tell you what. You find me an example of real-world solar cell efficiency that approaches 40%, as well as pointers as to producing them with an EROEI greater than 2, and I'll concede the point.
Not theoretical efficiency, not 'calculated based on baseline', but real and large-scale usage.
My bet is that can't find any. For what its worth - Smil (who has 30 years of energy research under his belt) knows about the non-polymer solar arrays that you talk about, and talks about them in Energies at the Crossroads.
He too mentions the 40-50% efficiencies you mention, but then goes on to mention that when that gets translated into real world use the efficiency goes down to 10-20%, and mentions the upkeep costs.
He then goes on to say that - reasonably, and before any storage costs are considered, that you can get 30 W/m^2 with current technology (as of 2003).
So, anyways, the proof is in the pudding. So perhaps we could come to some agreement if you had some real-world pointers and real-world examples, as well as EROEI studies.
> Hmmm. I take that post as agreement. Obviously
> Smil is massaging the truth (and you are starting
> to look silly defending this guy, who in 7 posts
> is proved wrong again and again - time to read a
> broader group of authors?). I'm looking forward to
> getting past your FUD, and discussing your other
> questions, but first we must get pass the
> nonsense.
Why don't you read the guy instead of denigrating him? Using one of your data sources and an off-the-cuff calc, I get 170 W/m^2 - the one you just gave, I get 200. I looked back at the ERBS and it gives me about the same. Its not a show-stopping difference.
What a lot of people don't like (and don't understand) about Smil is that he *doesn't* spin the facts, and hence steps on a lot of people's toes (including probably yours).
He also looks at the big picture, rather than the upfront figures like you are doing. Which is why I like him, because it gives me context when judging an energy source.
First of all - YES I'm aware that W/m^2 is a power measurement per unit area. If we amortized usage of all energy sources across the world, and averaged them out, it would come out to about 12 TW continuous usage.
And 87 PW is the amount of continuous power that the sun outputs that we can use. I like using it better because its simpler and you can get a good generalistic feel about an energy source using it - kwH tends to lend itself towards electricity itself.
Now - what I mind is that you are still double dipping. More than I at first realized. You mention 40% efficient solar collectors, but they are *multijunction* collectors, '500 sun' collectors, that are getting the energy. These are 40% efficient because solar collection is inherently more efficient at greater concentrations of energy.
These both are concentrators, *and* have to swing out a larger surface area to do so at the same time.
Lets use your document to see how much area they need to swing out:
"SOLAR RADIATION FOR FLAT-PLATE COLLECTORS FACING
SOUTH AT A FIXED-TILT (kWh/m2/day) Average: 4.3.
"SOLAR RADIATION FOR 2-AXIS TRACKING FLAT-PLATE
COLLECTORS (kWh/m2/day) Average: 6.6.
Now, there is only so much radiation per m^2, so it takes approximately 150% as much area to do tracking collectors. So the average *real* efficiency for these things is going to be 40% * 4.3 / 6.6 = 26%. At least before we consider the next part.
For I look later on in the doc and get:
DIRECT BEAM SOLAR RADIATION FOR CONCENTRATING COLLECTORS (kWh/m2/day) Average: 3.4
Notice how the average, 3.4, is less than 4.3? That's the efficiency penalty for the concentration - and for going through the lenses (second law of thermodynamics gets its share). It doesn't tell me whether or not the concentrating collectors were tracking or not (my guess is they were because you need to follow the sun in order to concentrate the light effectively) but we'll assume that it is not (and hence they aren't halving the energy available) So in the best case that's a further penalty. 26% * 3.4 / 4.3 == 20%.
So we are back at 20% efficiency overall. Even if your numbers are correct, 200 W/ m^2 (although that's still questionable) and 200 W/m^2 * 20% = 40 W/m^2.
And this doesn't address moving parts storage, etc, which of course are further drains on the efficiency.
So as I said, you are optimistic by at least a factor of two, most likely more than 2.5 here.
Ok, now as to your second point. You have a point about electricity (storage, transmission, and EROEI issues excluded), but transportation and heating is a different matter.
Given the infrastructure costs that we have invested in the ICE, its not very likely that these are going to go away any time soon (40+ yrs). So we are going to need to produce either natural gas or oil in large amounts to feed the current fleet of cars and various industries (including the photovoltaics industry which would need the plast
*sigh*
,and multiplied by 365.
I got the number from your source, over a thirty year period, at:
text rredc
I avoided anything that was tracking, took all the fixed rates, added them up together, averaged them
Unlike you, however, I'm trying to take into account the various inefficiencies and penalties that you get from using solar power - from penalties for storage, penalties for conversion into other energy carriers, penalties for the infrastructure involved in this conversion, penalties for transmission, and penalties for handling peak load.
And of course penalties for all the energy used to build and maintain the solar cells and the architecture behind them in the first place. I've seen EROEI's from anywhere less of one to about 3, taking the whole infrastructure of solar cells into account.
So yes, we disagree pretty much from the get go. Trackers are a red herring - there is only *so* much power available to gather from any given spot, and trackers just save on material costs.
I could calculate again, but we are essentially calculating different things. Your experiment is quaint at best, and misleading and false at worse. Money is a problem, yes, but so is transmission, storage, inefficiencies in capture, inefficiencies in maintenance, etc. etc. etc.
*That's* where I get an infrastructure the size of texas. By your naive calculations you are already up to 7% - double that, and that's more realistic IMO.
For example, the solar concentrators which you talk about have to have moving parts to get that 40% efficiency because they need to track the light source to concentrate the energy. These parts, whilst efficient, take an energy penalty of their own, and due to the need to directly track the sun point source in the sky, require a larger amount of area per device than you imply.
Average this out, and you're back to getting about 20-60 W/m^2 of energy for solar - which is the reference figure that I've seen everywhere except for you.
This of course is *before* any of the efficiency penalties that I've talked about.
You also are naively using 3 * 10 ^ 13 kW as our total energy source that would need to be replaced - the difference is that 90+% of that energy is in a form that we can directly use - natural gas for heating and gasoline for burning, coal for making steel, etc. You therefore take a penalty if you want to convert the solar into these forms of energy.
Lets take a different tack - overall, the earth intercepts at its surface about 87 PW of energy average. We use 12 TW (source: 'energies' by smil). That means that we use 1/7200 of the total solar flux of the planet. At 100% efficiency, therefore we would need 1/7200 of the earth's surface covered in solar cells - and since over 70% of the earth's surface is covered by water, approximately 1/2200 of the continents.
Since most of the solar needs to be converted to usable forms, I've been using a 60% penalty, which means that about 1/1360 th of the earth's continents would be needed to be covered by solar cells if everything was 100% efficient.
This equals 225,000 km^2. At 40% efficiency, this equals 562,000 km^2.
Since the current infrastructure is approx 290,000 km^2 - and of that, around 2% of that is energy production - the costs for the infrastructure in just collecting the solar power are bound to be more expensive than the current costs.
Just collecting is - even at your rather optimistic calcs - over 100 times more costly in terms of real estate than our current scheme!
And that is not even considering the *other* infrastructure involved, or the inefficiencies that I've talked about (and you haven't responded to). Or the low EROEI.
Anyways, look, I have NO problem with solar power. I hope that it goes gang-busters, I really do. You have yet to convince me. You just happen to be in solar technology before scalability issues become a factor.
But go ahead, lets stop with this area issue. You said that you'd like to respond to the other issues, so please do.
horos
pps -
I think I see where things are off, and where our numbers differ. They do in fact reconcile. You CANNOT USE A TRACKING MECHANISM TO DETERMINE SOLAR POWER DENSITY.
For if you do, you are basically double-dipping your calculations.
Figure: If a plate is tracking the sun, it is swinging out an area greater than the plate's area itself (for as it tracks, it gathers energy that would otherwise miss). In the process IT BLOCKS OTHER PLATES THAT WOULD HAVE OTHERWISE GOTTEN THE SAME ENERGY.
If you take, say the numbers for Birmingham, AL over the last 30 years,
your reference from rredc.nrel.gov
You'll notice that the flat collectors get an average of 3.5 kwH/m^2/day, which turns into about 165 W/m^2.
Its the dual tracking collectors that get the large amounts that you are talking about. However, the tracking collectors can't be placed directly next to each other, because they cast a SHADOW on each other.
My guess is that if you take this shadow into account, the benefits you get from tracking are greatly reduced (not eliminated because of better conversion efficiency) and the total that you CAN get is approx, on average 170W/m^2 (which goes along with the satellite data)
This is the only explanation that makes sense. My satellite numbers aren't lying, neither are your numbers - they are just double dipping when you make the assumption that they can be applied to large areas in 100% coverage.
Hence, the 100*100 sq mile solar concentration is way too small, by an order of magnitude or so, and we are still talking about an infrastructure the size of texas.
horos
1) your first link doesn't work (its broken) In fact, I couldn't even reach medc.nrel.gov.
2) second, the number I gave (140 W/m^2) is from the Earth Radiation balance satellite. I'm not sure what's going on here, but I'd hardly expect that group to lie.
3) the storage, transmission, and maintenance costs (especially from EROEI) makes it *much* more inefficent than just a 'third'.
Here's another example of efficiency loss - the sun is highly variable in the year, you hardly capture any power at all during the winter months. Hence, the need to store the electricity captured from the summer for months on end. Hence, high inefficiency.
You want to prove that solar is a decent replacement to me, address the efficiency concerns AFTER you have collected the solar energy, and the EROEI that you get from making solar cells.
All it takes is an overall efficiency of 10% (which is not at all hard given the cost of storage, maintenance, and manufacturing, battery and or energy carrier construction, and solar cell replacement ) and you are back to an infrastructure the size of texas.
And that is even with your numbers, which I think are *highly* optimistic.
Elsewise you really are looking at the world through rose-tinted glasses.
Anyways, I'll look for a third party that can corroborate your numbers, since I can't check them for myself. I did look at the NREL and sharp links, but they didn't go into *any* of these issues. They seem stuck on the 'look at us catching 40% of the sun's energy, isn't that cool!!!' or 'huge potential market opening up!!!' stage of things. You have to consider the entire energy life cycle, or your analysis is meaningless.
horos