" Because one only needs to look at Ontario(once the primary GDP producer of Canada) to see what high energy prices, and poor government decision making do."
Indeed, everyone should try that. Some of the best test scores on the planet, one of the highest percentages of post-secondary education, billions and billions in biomed research every year, and a long, healthy life span.
Beauty is in the eye of the beholder. Maybe if you took off the crap coloured glasses you might not thing everything stinks so much.
> And the performance degrades over time. These calculators don't seem to take that into account.
Depends on the calculator. But it's a small effect anyway, about 10% over 25 years. BTW, the panels are expected to last 50 years, I don't know where you got 20 from.
> Upstate NY has access to power from Hydro Quebec
You know who else does? Quebec. And a idled factory economy that looks much like NYs. And a 10% exchange rate in his favour.
There's a reason phinergy chose Quebec for their battery show-and-tell, and I'm surprised they've been so passive attracting similar endeavours. Simply put, anyone with a product where the energy input cost isn't a rounding error should be there. They generate at 1.1 cents/kWh.
"At $.25/W, that is a price of $50/m^2. This is in the range where it's sort-of-comparable with other roof claddings."
A solar panel is essentially a single-pane skylight, or screen door. Wholesale prices should on on par with them, and you shouldn't expect it to depress much below that. Shingles are unlikely to ever be on par.
> The cost of construction of PV panels is going up
No it's not. Raw material usage is going down continually. Intermediate steps and input chemicals is likewise decreasing, and being replaced by lower cost substitutes. The total material content and input stream in both materials and energy continues to decrease, and shows no sign of stopping.
> When that levels out, the cost of PVs will continue to increase.
Unless any one of the kerfless systems comes into production, at which point material use on the cell side goes down another 25 to 40%.
And yes, I've actually worked in a solar panel factory. I don't think you have done the same or I don't think you'd be making these statements.
"That said, I think the big manufacturers have really missed an opportunity in exactly the opposite direction of that you suggest - I don't give a damn about efficiency or how much space it takes up, I care about price per watt."
Which is all they concentrate on. Panel prices from the factory gate have fallen from $2 to 50 cents/W in the last four years. Efficiency has crept up from 14 to maybe 16 to 17%. They are doing precisely what you ask, and you're complaining?
"Sell me 10-20KW of 5% efficient panels for 25 cents per watt, and you'd have a very happy customer."
You are forgetting that panels aren't the only things in the system. If you care to run the numbers, I think you'll find that you're almost certainly wrong.
For instance, let's say you have enough room for 8 panels, like my new garage.. To be *able* to install the panels, I'll need to run DC wire from the roof to a point near the 240V pony panel (assuming you have one, if not...), put an inverter at that point, add a 30A breaker to the panel and connect the inverter to it, get a building permit, and then put the racking on the roof.
Racking is normally about 25 cents a watt, when measured against a typical 250 watt panel. So for a 2kW system we might expect to pay $500 for that kit. Inverters scale downward very poorly - a 2500W inverter is around 60 cents/W, while a 5000W one is around 40 cents/watt. That's because most of the parts are the same (the case, displays, controller, wiring, etc). An SMA 2500 is about $1500, while a 5k is 2200. The cabling and wiring needs to be done by an electrician and might take 1/2 a day, so let's say $750. The building permit, if you need engineering, is about $750 total. Total install time is about 2 man-days, so let's add $500 flat. Ok with that?
OK, so using 250W panels at 80 cents:
8 x 250 x 80 = $1600 + $1500 for inverter + $500 for racking + $750 for permitting etc + $750 for wiring + $750 for install = $5400
So that's $2.70 a watt. Ok, now let's do the same with your cheap panel:
8 x 75 x 25 = $150 + $1500 for inverter + $500 for racking + $750 for permitting etc + $750 for wiring + $750 for install = $2950
But now you only have 600W, so that's $6.60 a Watt. What a deal!
Yes, you can save some on the inverter, yes, you can DIY it and get rid of X and Y and Z. But I absolutely 100% assure you, the numbers end up in the same place every time, for small installs, higher wattage panels are almost *always* the way to go. If you don't believe me DO THE MATH YOURSELF.
> PV panels are also far less efficient than parabolic reflectors.
That is a non-sequitur.
Are you talking about using parabolic reflectors to heat a working fluid? If so, your statement is only true for very well lit, cloudless areas.
Further, in terms of *cost effectiveness* it's definitely *not* true. That's why PV is the fastest growing power source in the world and parabolics aren't.
Finally, you can't mount a parabolic system on your roof (easily anyway!).
Yes, they have their roles, but they are relatively limited.
> Right now a typical installation (complete, by a contractor, not DIY) is $7/watt for residential
I was doing residential installs three years ago for $5/W in Toronto. When I left the industry last year the going price for a 10k system was $26,000, fully installed and spinning the meter. It's even lower than that in Europe.
Look for the graph. It's based on 2011 numbers. Note the factory-gate price for panels at the bottom. They fell from 1.80 to 1.35 during a single year from 2010 to 2011. I know that they are down around 50 cents today, and you can easily buy a skid of panels at 80 to 90 cents. So if you just follow that red line three years into the future, you come to today's pricing at around $2/W.
> Most (if not all) solar panels depend on other rare earth materials that may be in short supply tho.
Not even remotely true. A very small subset of panels use some rare earths, and their percentage of the market is constantly falling. Those are, specifically, the CIGS and CdTe designs. The former was supposed to replace silicon by now, but got caught on the wrong side of the investment curve. Now it is a niche player, used where flexibility is required. The later was a major player for a very short period, and represented as much as 45% of the market in the late 2000's, but the main player, FirstSolar, is also a niche player today.
A normal solar panel consists of, approximately by weight:
glass (which includes the cells) aluminum (frame and back-side cell contacts) copper (wiring) plastic (wire insulation, junction box, backsheet) silver (solder and frontside wiring) sili*cone* (glue on the frame) more glue (special clear heat-spreading stuff)
> A bit like Tesla's batteries are depending on Lithium.
There's plenty of lithium for all our cars too. Supply is a problem, however, but that's political, not mechanical. In fact, almost all of what we need is sitting in a salt pan in Bolivia where you can just scoop it up with a bulldozer. But they won't let you.
Also, note that all of these materials, with the exception of the plastic, are HIGHLY recyclable. Panels and batteries can be something like 99% recycled with ease and at low cost.
There is essentially an infinite supply of all of these materials. That is, we could power EVERYTHING on the planet using these panels and still not make a dent in the existing markets for them.
"If the actual numbers work out when their quota sales guy arrives? Then you buy their SolarCity system, which you cannot modify or upgrade."
As someone that worked in the industry, not SC, I can't tell you that they do this for VERY good reasons. We found that offering any sort of option simply confused people, and led to drawn discussions that always ended up close to the original array anyway. Customers have all sorts of ideas about hanging panels over their windows or under the eaves and so forth. If someone wants to serve that market, go for it, but there's no money it in, least of all for the customer.
"There are better options, and cells with better efficiency"
Which is why he's buying this company. An average good panel today has cells in the 18 to 20% efficiency range, giving you total areal efficiency around 16% (wires, reflection from the glass, whitespace, etc). These guys make cells in the 22% range. For large arrays this has no real effect, but for small systems the costs are dominated by installation, so if you want the numbers to work out you have to get every watt you can. Given the small and fixed area of your roof, that means using the highest power panels you can find. In spite of any higher costs, this always wins in the end.
That said...
"Silevo claims that its panels have achieved a 22 percent efficiency and are well on their way to achieving 24 percent efficiency."
No, it claims their *cells* have done this. Their web page clearly states their panels are around 18%.
"It suggests that 10 cents per watt is saved for every point of efficiency gained."
That is a vague statement, 10 cents on the panel, or ten cents on the total system? You do get a bit of savings downstream because you're installing less panels, but racking is about20 to 25 cents/W, so improving the panel by 1% might get you a penny or two (maybe), not 10 cents. I am skeptical of this number.
Still, I wish I could buy them. I have a hole on my new garage that's just right for an eight panel array.
Hmmm, edit appears to have disappeared. Forgive me if this shows up twice.
> [[Citation needed]] - "all the time" is not a mathematical statement and therefore cannot be included in your (pseudo) mathematical reasoning.
https://en.wikipedia.org/wiki/Riggatron - was about 50% http://books.google.ca/books?id=KSA_AAAAQBAJ&pg=PA203 - calls for 80%, gives no reasons (maybe dup) aries.ucsd.edu/HAPL/MEETINGS/0511.../SheffieldApproachFusion.ppt - 20% first year, 50% after http://books.google.ca/books?id=5A51AgAAQBAJ&pg=PA139 - "70 to 80% [...] These values cannot be achieved today..." http://hifweb.lbl.gov/public/Sharp/HIF_documents/Perkins-future%20fusion.pdf - makes fun of IFE predicted cap factors, says ICF would be 80% in order to work but doesn't really argue for that number
> [[Citation needed]] - not to mention that since the system is not under significant pressure, > the containment building (if actually needed) will be far simpler and far cheaper than that needed by a nuclear power plant.
The magnets are under significant pressure, and represent a serious physical risk. A failure of the blanket releases tritium. To ensure that one doesn't lead to the other getting into the environment, you need a very strong containment vessel and building on the same level of size and strength as a fission version. You *have* seen the ITER containment building, right?
http://fire.pppl.gov/fusion_science_parkins_031006.pdf - breaks it down in detail http://dotearth.blogs.nytimes.com/2012/10/19/a-veteran-of-fusion-science-proposes-narrowing-the-field - Hirsch says "it is virtually certain that the regulators will demand a containment building for a commercial tokamak reactor that will likely resemble what is currently required for fission reactors" http://www.osti.gov/scitech/servlets/purl/7117740 - this is a very old study, but you can get a feel for the buildings as part of the overall system costs. They flatline it at 10% of the overall project cost, but they estimate that to be as low as 35 cents/kWt. http://books.google.ca/books?id=iuC3IFwk5ZsC&pg=PA342 - no dollar figures, but this shows you why you need a big expensive building
That last one is pretty good overall if you're interested in this stuff. It's based on the UWMAK-1, which was a study carried out by Bechtel and WISC. Using current figures you get CAPEX numbers like $1.8 for the blanket, which you can scale using the second-last ref to suggest building costs on the order of $6/W. If you want to understand why all of this starts adding up, go to the UWMAK home page and look at the image. See the *little person* at the bottom? Now scale that thing out to a complete torus:
http://fti.neep.wisc.edu/studies/UWMAK-I
For comparison, wind turbines are currently going in at $1.50 to $2.00, and NG combined-cycle plants around $1 to $1.50.
Which you could have shown me up by posting some numbers to show why I'm full of FUD. But you didn't. So first, pto, meet kettle. Secondly, now you have some reading to do.
> I won't fault any of your numbers, but failure to acknowledge the role of serendipity in the history of science and technology is just a statement of your own ignorance, not a convincing argument.
Right in the message you are replying to"
"'Now wait' you say... what if advancement X causes the price of fusion to fall? Well sure, but what if advancement Y causes the price of peanut turbines to fall."
This clearly acknowledges the role of serendipity. It is entirely possible that there will be a technical breakthrough and suddenly everything I said about fusion moot. But it is just as likely there will be a breakthrough in some other power source that renders fusion moot. That's the problem with serendipity, it's random.
I don't work in solar. I do have panels on my garage roof though.
> They agree that fusion is challenging
Of course; just challenging enough that a little more money will definitely, totally fix the problem.
> with childhood classmates, neighbours and professional colleagues now decades into their work in both next-gen > fission and current fusion reactor design
Excellent, I'd love to debate them. Feel free to set something up.
> the worst kind of bad science is advocacy that overreaches your expertise
No, the worst is when someone starts attacking the messenger while hiding behind an alias and making claims of great expertise.
> Anyone posting on/. ought to be well aware of the long, long history of technical prognostications > of exactly the kind you are posting here that turned out to be utterly, absolutely wrong.
Yes, *technically*. But this isn't about the *technical* side. This is about the *economics* side. Fusion reactors will be, forever, more expensive than a fission reactor. There is no way around this. And even today, fission reactors are too expensive to build. And that's that.
Look, someone might indeed invent a totally new way to build a fusion reactor that is actually simpler and cheaper than a fission one. If that happens, then we'll use fusion reactors.
But that's just playing dice. It's exactly as likely, or much more likely actually, that someone will come up with a way to 1/2 the price of PV. And if that happens, then even a 1/2 as expensive fusion reactor still won't get built.
> What was the usefulness of general relativity in the early 20th century ? Nothing before artificial satellite (i.e. GPS).
Bzzzt. GPS would work perfectly without GR. The only difference would be one less correction factor.
> What was the usefulness of galois theory in the 19th century ? Nothing. Now we have got major applications (coding theory...).
Galois theory was invented to solve a well known problem in mathematics of that era.
> What was the usefulness of Fast Fourier Transform (known since ~1800) ? Nada Now everywhere.
I think you are confusing the discrete form, not the FFT. The version of the FFT used today is from the 60s IIRC. Fourier analysis in general was invented in order to greatly simplify heat transfer equations.
Let me guess, this list came from a page supporting one or another wild-eyed idea and used arguments like these to suggest why the fact that the system didn't actually work should not be a reason to doubt it?
Take a Tesla S. Remove 2/3rds of the li-ion. Add one of these. Car loses 500 lbs. One-way range increases to ~1600 km. Refuelling for short trips is about 5 minutes. Longer ones takes a swap, just like now.
> Your argument appears to be "we haven't solve the technical and practical challenges yet, so we never will."
What?!? I said the *exact opposite* of that.
I said that even if they get it working, there's no reason to build it.
Here, let me put this in crayon for you. Right now I can go and buy a turbine from GE, hook that up to a food dryer system from some hippy store, and use it to dry out peanut butter and feed them into the turbine. I *guarantee* you this will actually work, and produce net energy. What, you don't believe me? Fine, read this:
Better yet, it's carbon neutral, because the CO2 you release by burning it is sucked back into the next tree. Now of course the power coming out would cost ten times what you'd get by burning bunker oil, and bunker oil produces power at ten times the rate of a wind turbine, but *it will work*, for sure. Fusion? Meh, maybe by 2050. Maybe not. And of course, fusion will likely cost even more.
So what problem does a fusion reactor solve that a peanut turbine doesn't? None. So why isn't anyone racing to built peanut turbines? Because they cost too much. And fusion costs more than that.
And THAT is my argument.
"Now wait" you say... what if advancement X causes the price of fusion to fall? Well sure, but what if advancement Y causes the price of peanut turbines to fall? And when you look at all the research in the world, there's a lot more going into making cheaper peanuts than fusion.
I am being a bit facetious here, but not that much. I've been looking at this problem for three decades now, and it's not getting any better. Quite the opposite, fusion is getting more and more expensive. Its just not going to happen. You need to spend your energy on something that will actually happen, even if it's not as good in theory.
> Because thorium isn't easier or cheaper to use in the US
Nor India, who imports all the fuel they need. And as the supply from Africa and Australia remains solid for the foreseeable future, the economic argument is unlikely to work for anyone, including India.
> You said you where aware of India working on this, but I guess not aware that they want it running in less than a decade, not decades.
I'm perfectly aware of this. I'm also aware that they said the same thing a decade ago. And the decade before that. And I'm also aware that they laid out this plan *in 1954*.
None of this inspires confidence, especially in a market where PV and wind will almost certainly (we're talking 99.9% here) cost less. You are aware, I'm sure, that India currently has plans to install more PV than nuclear, right?
> Your assumption of course is that all other factors stay the same
Exactly the opposite, I'm taking into account the changing market at every turn.
Right now commercial PV is around 8 cents and is expected to fall in 6 to 7 cents by the end of this year. Right now wind turbines are producing power for between 4.5 and 9 cents, and it is expected the price will collapse to the 5 cent mark over time. These numbers include factors for intermittency, transmission upgrades, and anything else you might think of.
So, thorium. In spite of multiple decades of ongoing research, we still have no working thorium reactor. In fact, that's true in spite of the fact that the reactor just down the road from me can run on it. So if we have reactors right now that can use it, and they're not, surely there is a reason for this, right?
And the reason is that the price of building the infrastructure needed to commercialize the fuel pipeline is enormous, and at current U2 prices, utterly pointless. As I'm sure you're no doubt aware, the price of the U2 fuel cycle development was paid for by WWII, which provided a large subsidy to plants in countries with military needs. That leaves only Germany as a country that had to develop a fuel cycle *without* an interest in bombs, and look how well that turned out for them.
So basically people that actually work in the power industry, especially the nuclear power industry, see that this is not a technical problem (well, it is) but a practical one. One that is *not* getting solved any time soon. Perhaps this is simply a chicken-egg problem, and anyone that cracks one side will produce a reason to attack the other. But to date that hasn't happened, and there you have it.
> was wondering the same thing... perhaps spiking up for major experiments/phases?
Correct. The assumption in that graph is that applying more money means you need to do less experiments. The super-funded option has an initial series of experiments, followed by a testing phase, followed by a second round of construction and testing. The dot represents commercialization.
The problem is that all predictions about the "amount of science" left have been wrong, every time. In 1953 Spitzer predicted commercial systems by 1970 and outlined a four-step path for stellerator development, culminating in the D model that would be a production prototype. In fact, the system plateaued in the B model, and the C model was never completed in its original form because it was clear that approach would never work and they should move to tokamaks instead.
Every device has gone through a similar evolution, or much worse. There were dozens upon dozens of designs in the 1950s and 60s, not in terms of machines, but entirely different approaches to building a reactor. Most of these have proven to have plateau points well below any sort of break-even, technical or practical. Mirrors, picket fences, astrons, bumpy toruses, electron-beam ICF, zeta-pinch, theta-pinch, etc etc etc etc etc. Today we are left with two approaches, laser ICF and tokamak.
So explore the tokamak for a second. With each generation of machine the cost of staying in the game increases another order of magnitude. Today, we can only afford to build one machine. Originally ITER was going to be the testbed for a design known as DEMO. DEMO would be the testbed for a commercial reactor. The end was in sight.
Oh well, not any more, because now DEMO is the testbed for PROTO. PROTO will be the testbed for a commercial reactor. Time frame for PROTO? 2050 at the least.
I will not be alive in 2050. Many of the people reading this won't be. However, long before that, other forms of power will commercialize. We'll keep sinking money into this pit throughout though, because, as this article notes, that's the way the pork works.
There's *all* truth to that. Let me put this simply; there is almost zero chance that fusion, in its current form, will *ever* be a practical power source.
Now when people read a statement like that they get their backs up about the future, and progress and science and all that. But that's not the issue. The issue is that *fusion isn't the only power source on the planet*. As long as one of these is "better" that fusion, then fusion won't happen. That's all there is to it.
So why do I state my conclusion so forcefully? Because math.
The Levelized Cost of Electricity is the key determinant in telling you whether or not a system will be built. The formula basically tells you what you have to charge for the power coming out of your system in order to break even. Anything above that number is gravy.
The formula, which you can read in depth here: http://matter2energy.wordpress.com/2012/01/24/your-own-grid-parity-pv-system/
basically boils down to five numbers. The first is the amount of money you pay for the plant, and more specifically, the amount of interest you pay on the loans you took out to build it. The second is the cost of fuel to produce a given amount of power. The next is the peak power that the plant can produce, and next is the percentage of time that the plant actually does produce that. Finally there's the lifetime of the plant, which feed into all of the others. It's something like this:
price of your power = (all the money you put into the plant over its lifetime) / (all the power that you exported to the grid)
We measure money in dollars and cents. We measure power in kWh. This is why your power bill lists a figure in cents/kWh, and why the grid operators measure in $/MWh.
Ok, so fusion. So the price of fuel for a fusion reactor is low, about the same as a fission plant. So we can eliminate that figure for a rule-of-thumb calculation, and leaves us with the lifetime cost of the plant, the CAPEX+OPEX. Now we look at the other side, and we see two figures, the peak power and the percentage of time it runs. We can simplify by listing our CAPEX/peak power as a single number, dollars per watt.
So basically the entire cost structure comes down to the cost of the reactor, and the amount of time it spends running. The rest we can scale out linearly against other power sources.
So what do we know about these two factors?
Well in terms of percentage power, or capacity factor as we call it, fusion reactors are not competitive. Because of neutron embrittlement, they need to be shut down all the time so the reactor core liner can be removed and replaced. Newer designs place lithium-infused blocks inside the containment vessel; this means the vessel itself lasts longer but you still need to open it up all the time to get at those blocks. Generally we might expect a fusion plant to have a capacity factor on the order of a good hydro plant, on the order of 60%. For comparison, a fission plant is around 90%, a wind turbine is 30%, a solar panel is about 15%.
Ok, now the CAPEX. Any fusion reactor of practical output is going to be one of the most fantastically complicated devices ever made. They are utterly crammed with high-end materials, poisons, huge electrical and magnetic systems, high-end vacuum pumps, etc. Depending on the design, it's also flammable, and the fire will cause radioactive rain, so you still need a complete containment building. Now on top of this all, the energy density of a fusion system is *tiny*, so you need to build *enormous* reactors.
And that's where it falls apart. There is simply no way, under any reasonable development line, that the cost of building the plant, and servicing its debt, can possibly be made up by the electricity coming out. PV, one of the worst power sources in terms of cents/kWh, is currently running at about 15 to 20 cents/kWh. A fusion reactor almost certainly cannot be built that will produce power at under ten times that cost. And that's assuming it ever "works
Slashdot Mirror
" Because one only needs to look at Ontario(once the primary GDP producer of Canada) to see what high energy prices, and poor government decision making do."
Indeed, everyone should try that. Some of the best test scores on the planet, one of the highest percentages of post-secondary education, billions and billions in biomed research every year, and a long, healthy life span.
Beauty is in the eye of the beholder. Maybe if you took off the crap coloured glasses you might not thing everything stinks so much.
Well, there is the winter...
> And the performance degrades over time. These calculators don't seem to take that into account.
Depends on the calculator. But it's a small effect anyway, about 10% over 25 years. BTW, the panels are expected to last 50 years, I don't know where you got 20 from.
http://matter2energy.wordpress.com/2012/05/21/green-apples/
> Upstate NY has access to power from Hydro Quebec
You know who else does? Quebec. And a idled factory economy that looks much like NYs. And a 10% exchange rate in his favour.
There's a reason phinergy chose Quebec for their battery show-and-tell, and I'm surprised they've been so passive attracting similar endeavours. Simply put, anyone with a product where the energy input cost isn't a rounding error should be there. They generate at 1.1 cents/kWh.
"At $.25/W, that is a price of $50/m^2.
This is in the range where it's sort-of-comparable with other roof claddings."
A solar panel is essentially a single-pane skylight, or screen door. Wholesale prices should on on par with them, and you shouldn't expect it to depress much below that. Shingles are unlikely to ever be on par.
> The cost of construction of PV panels is going up
No it's not. Raw material usage is going down continually. Intermediate steps and input chemicals is likewise decreasing, and being replaced by lower cost substitutes. The total material content and input stream in both materials and energy continues to decrease, and shows no sign of stopping.
> When that levels out, the cost of PVs will continue to increase.
Unless any one of the kerfless systems comes into production, at which point material use on the cell side goes down another 25 to 40%.
And yes, I've actually worked in a solar panel factory. I don't think you have done the same or I don't think you'd be making these statements.
"That said, I think the big manufacturers have really missed an opportunity in exactly the opposite direction of that you suggest - I don't give a damn about efficiency or how much space it takes up, I care about price per watt."
Which is all they concentrate on. Panel prices from the factory gate have fallen from $2 to 50 cents/W in the last four years. Efficiency has crept up from 14 to maybe 16 to 17%. They are doing precisely what you ask, and you're complaining?
"Sell me 10-20KW of 5% efficient panels for 25 cents per watt, and you'd have a very happy customer."
You are forgetting that panels aren't the only things in the system. If you care to run the numbers, I think you'll find that you're almost certainly wrong.
For instance, let's say you have enough room for 8 panels, like my new garage.. To be *able* to install the panels, I'll need to run DC wire from the roof to a point near the 240V pony panel (assuming you have one, if not...), put an inverter at that point, add a 30A breaker to the panel and connect the inverter to it, get a building permit, and then put the racking on the roof.
Racking is normally about 25 cents a watt, when measured against a typical 250 watt panel. So for a 2kW system we might expect to pay $500 for that kit. Inverters scale downward very poorly - a 2500W inverter is around 60 cents/W, while a 5000W one is around 40 cents/watt. That's because most of the parts are the same (the case, displays, controller, wiring, etc). An SMA 2500 is about $1500, while a 5k is 2200. The cabling and wiring needs to be done by an electrician and might take 1/2 a day, so let's say $750. The building permit, if you need engineering, is about $750 total. Total install time is about 2 man-days, so let's add $500 flat. Ok with that?
OK, so using 250W panels at 80 cents:
8 x 250 x 80 = $1600
+ $1500 for inverter
+ $500 for racking
+ $750 for permitting etc
+ $750 for wiring
+ $750 for install
= $5400
So that's $2.70 a watt. Ok, now let's do the same with your cheap panel:
8 x 75 x 25 = $150
+ $1500 for inverter
+ $500 for racking
+ $750 for permitting etc
+ $750 for wiring
+ $750 for install
= $2950
But now you only have 600W, so that's $6.60 a Watt. What a deal!
Yes, you can save some on the inverter, yes, you can DIY it and get rid of X and Y and Z. But I absolutely 100% assure you, the numbers end up in the same place every time, for small installs, higher wattage panels are almost *always* the way to go. If you don't believe me DO THE MATH YOURSELF.
> solar is not viable in terms of cost and this is unlikely to change anytime soon
You can do the math yourself:
http://wp.me/py8qF-4i
If you're paying more than 15 cents, PV is at least worth looking at.
> PV panels are also far less efficient than parabolic reflectors.
That is a non-sequitur.
Are you talking about using parabolic reflectors to heat a working fluid? If so, your statement is only true for very well lit, cloudless areas.
Further, in terms of *cost effectiveness* it's definitely *not* true. That's why PV is the fastest growing power source in the world and parabolics aren't.
Finally, you can't mount a parabolic system on your roof (easily anyway!).
Yes, they have their roles, but they are relatively limited.
> Right now a typical installation (complete, by a contractor, not DIY) is $7/watt for residential
I was doing residential installs three years ago for $5/W in Toronto. When I left the industry last year the going price for a 10k system was $26,000, fully installed and spinning the meter. It's even lower than that in Europe.
http://emp.lbl.gov/sites/all/files/german-us-pv-price-ppt.pdf
Look for the graph. It's based on 2011 numbers. Note the factory-gate price for panels at the bottom. They fell from 1.80 to 1.35 during a single year from 2010 to 2011. I know that they are down around 50 cents today, and you can easily buy a skid of panels at 80 to 90 cents. So if you just follow that red line three years into the future, you come to today's pricing at around $2/W.
> Most (if not all) solar panels depend on other rare earth materials that may be in short supply tho.
Not even remotely true. A very small subset of panels use some rare earths, and their percentage of the market is constantly falling. Those are, specifically, the CIGS and CdTe designs. The former was supposed to replace silicon by now, but got caught on the wrong side of the investment curve. Now it is a niche player, used where flexibility is required. The later was a major player for a very short period, and represented as much as 45% of the market in the late 2000's, but the main player, FirstSolar, is also a niche player today.
A normal solar panel consists of, approximately by weight:
glass (which includes the cells)
aluminum (frame and back-side cell contacts)
copper (wiring)
plastic (wire insulation, junction box, backsheet)
silver (solder and frontside wiring)
sili*cone* (glue on the frame)
more glue (special clear heat-spreading stuff)
> A bit like Tesla's batteries are depending on Lithium.
There's plenty of lithium for all our cars too. Supply is a problem, however, but that's political, not mechanical. In fact, almost all of what we need is sitting in a salt pan in Bolivia where you can just scoop it up with a bulldozer. But they won't let you.
Also, note that all of these materials, with the exception of the plastic, are HIGHLY recyclable. Panels and batteries can be something like 99% recycled with ease and at low cost.
There is essentially an infinite supply of all of these materials. That is, we could power EVERYTHING on the planet using these panels and still not make a dent in the existing markets for them.
Did you actually READ the article you linked to? It's a lithium battery. Look at the diagram. Sheesh,
There are, however, interesting side-solutions:
http://matter2energy.wordpress.com/2014/06/06/phinergys-battery-energy-storage-problem-solved/
"If the actual numbers work out when their quota sales guy arrives? Then you buy their SolarCity system, which you cannot modify or upgrade."
As someone that worked in the industry, not SC, I can't tell you that they do this for VERY good reasons. We found that offering any sort of option simply confused people, and led to drawn discussions that always ended up close to the original array anyway. Customers have all sorts of ideas about hanging panels over their windows or under the eaves and so forth. If someone wants to serve that market, go for it, but there's no money it in, least of all for the customer.
"There are better options, and cells with better efficiency"
Which is why he's buying this company. An average good panel today has cells in the 18 to 20% efficiency range, giving you total areal efficiency around 16% (wires, reflection from the glass, whitespace, etc). These guys make cells in the 22% range. For large arrays this has no real effect, but for small systems the costs are dominated by installation, so if you want the numbers to work out you have to get every watt you can. Given the small and fixed area of your roof, that means using the highest power panels you can find. In spite of any higher costs, this always wins in the end.
That said...
"Silevo claims that its panels have achieved a 22 percent efficiency and are well on their way to achieving 24 percent efficiency."
No, it claims their *cells* have done this. Their web page clearly states their panels are around 18%.
"It suggests that 10 cents per watt is saved for every point of efficiency gained."
That is a vague statement, 10 cents on the panel, or ten cents on the total system? You do get a bit of savings downstream because you're installing less panels, but racking is about20 to 25 cents/W, so improving the panel by 1% might get you a penny or two (maybe), not 10 cents. I am skeptical of this number.
Still, I wish I could buy them. I have a hole on my new garage that's just right for an eight panel array.
Hmmm, edit appears to have disappeared. Forgive me if this shows up twice.
> [[Citation needed]] - "all the time" is not a mathematical statement and therefore cannot be included in your (pseudo) mathematical reasoning.
https://en.wikipedia.org/wiki/Riggatron - was about 50%
http://books.google.ca/books?id=KSA_AAAAQBAJ&pg=PA203 - calls for 80%, gives no reasons (maybe dup)
aries.ucsd.edu/HAPL/MEETINGS/0511.../SheffieldApproachFusion.ppt - 20% first year, 50% after
http://books.google.ca/books?id=5A51AgAAQBAJ&pg=PA139 - "70 to 80% [...] These values cannot be achieved today..."
http://hifweb.lbl.gov/public/Sharp/HIF_documents/Perkins-future%20fusion.pdf - makes fun of IFE predicted cap factors, says ICF would be 80% in order to work but doesn't really argue for that number
> [[Citation needed]] - not to mention that since the system is not under significant pressure,
> the containment building (if actually needed) will be far simpler and far cheaper than that needed by a nuclear power plant.
The magnets are under significant pressure, and represent a serious physical risk. A failure of the blanket releases tritium. To ensure that one doesn't lead to the other getting into the environment, you need a very strong containment vessel and building on the same level of size and strength as a fission version. You *have* seen the ITER containment building, right?
http://fire.pppl.gov/fusion_science_parkins_031006.pdf - breaks it down in detail
http://dotearth.blogs.nytimes.com/2012/10/19/a-veteran-of-fusion-science-proposes-narrowing-the-field - Hirsch says "it is virtually certain that the regulators will demand a containment building for a commercial tokamak reactor that will likely resemble what is currently required for fission reactors"
http://www.osti.gov/scitech/servlets/purl/7117740 - this is a very old study, but you can get a feel for the buildings as part of the overall system costs. They flatline it at 10% of the overall project cost, but they estimate that to be as low as 35 cents/kWt.
http://books.google.ca/books?id=iuC3IFwk5ZsC&pg=PA342 - no dollar figures, but this shows you why you need a big expensive building
That last one is pretty good overall if you're interested in this stuff. It's based on the UWMAK-1, which was a study carried out by Bechtel and WISC. Using current figures you get CAPEX numbers like $1.8 for the blanket, which you can scale using the second-last ref to suggest building costs on the order of $6/W. If you want to understand why all of this starts adding up, go to the UWMAK home page and look at the image. See the *little person* at the bottom? Now scale that thing out to a complete torus:
http://fti.neep.wisc.edu/studies/UWMAK-I
For comparison, wind turbines are currently going in at $1.50 to $2.00, and NG combined-cycle plants around $1 to $1.50.
http://gallery.mailchimp.com/ce17780900c3d223633ecfa59/files/Lazard_Levelized_Cost_of_Energy_v7.0.1.pdf
> but then substitute FUD for actual numbers
Which you could have shown me up by posting some numbers to show why I'm full of FUD. But you didn't. So first, pto, meet kettle. Secondly, now you have some reading to do.
> I won't fault any of your numbers, but failure to acknowledge the role of serendipity in the history of science and technology is just a statement of your own ignorance, not a convincing argument.
Right in the message you are replying to"
"'Now wait' you say... what if advancement X causes the price of fusion to fall? Well sure, but what if advancement Y causes the price of peanut turbines to fall."
This clearly acknowledges the role of serendipity. It is entirely possible that there will be a technical breakthrough and suddenly everything I said about fusion moot. But it is just as likely there will be a breakthrough in some other power source that renders fusion moot. That's the problem with serendipity, it's random.
Found a better link to the one article:
http://fire.pppl.gov/fusion_science_parkins_031006.pdf
> and here's why I've bet my career on solar
I don't work in solar. I do have panels on my garage roof though.
> They agree that fusion is challenging
Of course; just challenging enough that a little more money will definitely, totally fix the problem.
> with childhood classmates, neighbours and professional colleagues now decades into their work in both next-gen
> fission and current fusion reactor design
Excellent, I'd love to debate them. Feel free to set something up.
> the worst kind of bad science is advocacy that overreaches your expertise
No, the worst is when someone starts attacking the messenger while hiding behind an alias and making claims of great expertise.
> Anyone posting on /. ought to be well aware of the long, long history of technical prognostications
> of exactly the kind you are posting here that turned out to be utterly, absolutely wrong.
Yes, *technically*. But this isn't about the *technical* side. This is about the *economics* side. Fusion reactors will be, forever, more expensive than a fission reactor. There is no way around this. And even today, fission reactors are too expensive to build. And that's that.
Look, someone might indeed invent a totally new way to build a fusion reactor that is actually simpler and cheaper than a fission one. If that happens, then we'll use fusion reactors.
But that's just playing dice. It's exactly as likely, or much more likely actually, that someone will come up with a way to 1/2 the price of PV. And if that happens, then even a 1/2 as expensive fusion reactor still won't get built.
> What was the usefulness of general relativity in the early 20th century ? Nothing before artificial satellite (i.e. GPS).
Bzzzt. GPS would work perfectly without GR. The only difference would be one less correction factor.
> What was the usefulness of galois theory in the 19th century ? Nothing. Now we have got major applications (coding theory...).
Galois theory was invented to solve a well known problem in mathematics of that era.
> What was the usefulness of Fast Fourier Transform (known since ~1800) ? Nada Now everywhere.
I think you are confusing the discrete form, not the FFT. The version of the FFT used today is from the 60s IIRC. Fourier analysis in general was invented in order to greatly simplify heat transfer equations.
Let me guess, this list came from a page supporting one or another wild-eyed idea and used arguments like these to suggest why the fact that the system didn't actually work should not be a reason to doubt it?
> To some degree, I still like the idea of plug-in hybrids for the time being
Especially when the hybrid is this:
http://www.cbc.ca/news/technology/electric-car-with-massive-range-in-demo-by-phinergy-alcoa-1.2664653
Take a Tesla S. Remove 2/3rds of the li-ion. Add one of these. Car loses 500 lbs. One-way range increases to ~1600 km. Refuelling for short trips is about 5 minutes. Longer ones takes a swap, just like now.
> Your argument appears to be "we haven't solve the technical and practical challenges yet, so we never will."
What?!? I said the *exact opposite* of that.
I said that even if they get it working, there's no reason to build it.
Here, let me put this in crayon for you. Right now I can go and buy a turbine from GE, hook that up to a food dryer system from some hippy store, and use it to dry out peanut butter and feed them into the turbine. I *guarantee* you this will actually work, and produce net energy. What, you don't believe me? Fine, read this:
https://en.wikipedia.org/wiki/Chrysler_Turbine_Car
Better yet, it's carbon neutral, because the CO2 you release by burning it is sucked back into the next tree. Now of course the power coming out would cost ten times what you'd get by burning bunker oil, and bunker oil produces power at ten times the rate of a wind turbine, but *it will work*, for sure. Fusion? Meh, maybe by 2050. Maybe not. And of course, fusion will likely cost even more.
So what problem does a fusion reactor solve that a peanut turbine doesn't? None. So why isn't anyone racing to built peanut turbines? Because they cost too much. And fusion costs more than that.
And THAT is my argument.
"Now wait" you say... what if advancement X causes the price of fusion to fall? Well sure, but what if advancement Y causes the price of peanut turbines to fall? And when you look at all the research in the world, there's a lot more going into making cheaper peanuts than fusion.
I am being a bit facetious here, but not that much. I've been looking at this problem for three decades now, and it's not getting any better. Quite the opposite, fusion is getting more and more expensive. Its just not going to happen. You need to spend your energy on something that will actually happen, even if it's not as good in theory.
> Because thorium isn't easier or cheaper to use in the US
Nor India, who imports all the fuel they need. And as the supply from Africa and Australia remains solid for the foreseeable future, the economic argument is unlikely to work for anyone, including India.
> You said you where aware of India working on this, but I guess not aware that they want it running in less than a decade, not decades.
I'm perfectly aware of this. I'm also aware that they said the same thing a decade ago. And the decade before that. And I'm also aware that they laid out this plan *in 1954*.
None of this inspires confidence, especially in a market where PV and wind will almost certainly (we're talking 99.9% here) cost less. You are aware, I'm sure, that India currently has plans to install more PV than nuclear, right?
> Your assumption of course is that all other factors stay the same
Exactly the opposite, I'm taking into account the changing market at every turn.
Right now commercial PV is around 8 cents and is expected to fall in 6 to 7 cents by the end of this year.
Right now wind turbines are producing power for between 4.5 and 9 cents, and it is expected the price will collapse to the 5 cent mark over time.
These numbers include factors for intermittency, transmission upgrades, and anything else you might think of.
So, thorium. In spite of multiple decades of ongoing research, we still have no working thorium reactor. In fact, that's true in spite of the fact that the reactor just down the road from me can run on it. So if we have reactors right now that can use it, and they're not, surely there is a reason for this, right?
And the reason is that the price of building the infrastructure needed to commercialize the fuel pipeline is enormous, and at current U2 prices, utterly pointless. As I'm sure you're no doubt aware, the price of the U2 fuel cycle development was paid for by WWII, which provided a large subsidy to plants in countries with military needs. That leaves only Germany as a country that had to develop a fuel cycle *without* an interest in bombs, and look how well that turned out for them.
So basically people that actually work in the power industry, especially the nuclear power industry, see that this is not a technical problem (well, it is) but a practical one. One that is *not* getting solved any time soon. Perhaps this is simply a chicken-egg problem, and anyone that cracks one side will produce a reason to attack the other. But to date that hasn't happened, and there you have it.
> was wondering the same thing... perhaps spiking up for major experiments/phases?
Correct. The assumption in that graph is that applying more money means you need to do less experiments. The super-funded option has an initial series of experiments, followed by a testing phase, followed by a second round of construction and testing. The dot represents commercialization.
The problem is that all predictions about the "amount of science" left have been wrong, every time. In 1953 Spitzer predicted commercial systems by 1970 and outlined a four-step path for stellerator development, culminating in the D model that would be a production prototype. In fact, the system plateaued in the B model, and the C model was never completed in its original form because it was clear that approach would never work and they should move to tokamaks instead.
Every device has gone through a similar evolution, or much worse. There were dozens upon dozens of designs in the 1950s and 60s, not in terms of machines, but entirely different approaches to building a reactor. Most of these have proven to have plateau points well below any sort of break-even, technical or practical. Mirrors, picket fences, astrons, bumpy toruses, electron-beam ICF, zeta-pinch, theta-pinch, etc etc etc etc etc. Today we are left with two approaches, laser ICF and tokamak.
So explore the tokamak for a second. With each generation of machine the cost of staying in the game increases another order of magnitude. Today, we can only afford to build one machine. Originally ITER was going to be the testbed for a design known as DEMO. DEMO would be the testbed for a commercial reactor. The end was in sight.
Oh well, not any more, because now DEMO is the testbed for PROTO. PROTO will be the testbed for a commercial reactor. Time frame for PROTO? 2050 at the least.
I will not be alive in 2050. Many of the people reading this won't be. However, long before that, other forms of power will commercialize. We'll keep sinking money into this pit throughout though, because, as this article notes, that's the way the pork works.
> there is some truth
There's *all* truth to that. Let me put this simply; there is almost zero chance that fusion, in its current form, will *ever* be a practical power source.
Now when people read a statement like that they get their backs up about the future, and progress and science and all that. But that's not the issue. The issue is that *fusion isn't the only power source on the planet*. As long as one of these is "better" that fusion, then fusion won't happen. That's all there is to it.
So why do I state my conclusion so forcefully? Because math.
The Levelized Cost of Electricity is the key determinant in telling you whether or not a system will be built. The formula basically tells you what you have to charge for the power coming out of your system in order to break even. Anything above that number is gravy.
The formula, which you can read in depth here:
http://matter2energy.wordpress.com/2012/01/24/your-own-grid-parity-pv-system/
basically boils down to five numbers. The first is the amount of money you pay for the plant, and more specifically, the amount of interest you pay on the loans you took out to build it. The second is the cost of fuel to produce a given amount of power. The next is the peak power that the plant can produce, and next is the percentage of time that the plant actually does produce that. Finally there's the lifetime of the plant, which feed into all of the others. It's something like this:
price of your power = (all the money you put into the plant over its lifetime) / (all the power that you exported to the grid)
We measure money in dollars and cents. We measure power in kWh. This is why your power bill lists a figure in cents/kWh, and why the grid operators measure in $/MWh.
Ok, so fusion. So the price of fuel for a fusion reactor is low, about the same as a fission plant. So we can eliminate that figure for a rule-of-thumb calculation, and leaves us with the lifetime cost of the plant, the CAPEX+OPEX. Now we look at the other side, and we see two figures, the peak power and the percentage of time it runs. We can simplify by listing our CAPEX/peak power as a single number, dollars per watt.
So basically the entire cost structure comes down to the cost of the reactor, and the amount of time it spends running. The rest we can scale out linearly against other power sources.
So what do we know about these two factors?
Well in terms of percentage power, or capacity factor as we call it, fusion reactors are not competitive. Because of neutron embrittlement, they need to be shut down all the time so the reactor core liner can be removed and replaced. Newer designs place lithium-infused blocks inside the containment vessel; this means the vessel itself lasts longer but you still need to open it up all the time to get at those blocks. Generally we might expect a fusion plant to have a capacity factor on the order of a good hydro plant, on the order of 60%. For comparison, a fission plant is around 90%, a wind turbine is 30%, a solar panel is about 15%.
Ok, now the CAPEX. Any fusion reactor of practical output is going to be one of the most fantastically complicated devices ever made. They are utterly crammed with high-end materials, poisons, huge electrical and magnetic systems, high-end vacuum pumps, etc. Depending on the design, it's also flammable, and the fire will cause radioactive rain, so you still need a complete containment building. Now on top of this all, the energy density of a fusion system is *tiny*, so you need to build *enormous* reactors.
And that's where it falls apart. There is simply no way, under any reasonable development line, that the cost of building the plant, and servicing its debt, can possibly be made up by the electricity coming out. PV, one of the worst power sources in terms of cents/kWh, is currently running at about 15 to 20 cents/kWh. A fusion reactor almost certainly cannot be built that will produce power at under ten times that cost. And that's assuming it ever "works
> Still, though, both fission and fusion are much better than the alternatives (fossil fuels).
Fallacy of the excluded middle. There is no way fusion will ever compete with this:
http://gallery.mailchimp.com/ce17780900c3d223633ecfa59/files/Lazard_Levelized_Cost_of_Energy_v7.0.1.pdf