The Electric Car
in [Design]
|
From: Eeyore on 5 Oct 2007 09:36 Willie.Mookie(a)gmail.com wrote: > Eeyore wrote: > > Willie.Moo...(a)gmail.com wrote: > > > To intertie to the grid would increase costs to 6 cents per kWh at > > > best > > > > Based on a silly estimate of the cost of grid tied inverters which you have > > overestiamted thge cost of by at least 3 times and probably more. > > > > With a sensible supply of inverters that would be 2 cents / kWh or less. > > Graham, you might be right that $2 per peak watt is high. But it is > at the high end of the envelope. Fact is, no one is building DC > intertie at the 50 MW scale and above so no one knows what all the > costs will be. You need to talk to someone like ABB. > The first one won't be $0.40 per peak watt - which is > what you're suggesting I think. Though with a dedicated effort you > might get down this level. A real first system - for planning > purposes - I think $2 per peak watt is a conservative number (that is > its high - but likely the first go round) Very high if I can readily find grid-tied inverters on sale today at the 3kW level for < $1 a watt. You still have the problem of storing energy outside of the 'solar hours' of course. Graham
From: John Larkin on 5 Oct 2007 10:03 On Fri, 05 Oct 2007 11:16:57 -0000, Willie.Mookie(a)gmail.com wrote: >On Oct 5, 2:53 am, Eeyore <rabbitsfriendsandrelati...(a)hotmail.com> >wrote: >> Willie.Moo...(a)gmail.com wrote: >> > To intertie to the grid would increase costs to 6 cents per kWh at >> > best >> >> Based on a silly estimate of the cost of grid tied inverters which you have >> overestiamted thge cost of by at least 3 times and probably more. >> >> With a sensible supply of inverters that would be 2 cents / kWh or less. >> >> Graham > >Graham, you might be right that $2 per peak watt is high. But it is >at the high end of the envelope. Fact is, no one is building DC >intertie at the 50 MW scale and above so no one knows what all the >costs will be. The first one won't be $0.40 per peak watt - which is >what you're suggesting I think. Though with a dedicated effort you >might get down this level. A real first system - for planning >purposes - I think $2 per peak watt is a conservative number (that is >its high - but likely the first go round) > >So, ask yourself, at what scale? > >It is true economies of scale and leaerning curve effects do reduce >the cost of smaller units, and may reduce initial costs well below the >$2 estimate my detailed engineering analysis came up with. > >You by comparison have merely indicated a suspicion of where intertie >might be after a decade or more of large-scale development. Hogwash. In 90 seconds of research, you can discover that UPS systems - charger, batteries, inverter, enclosure, line cord, LED dislays - are going for 10 to 15 cents per watt RETAIL. Utilities have been shipping interstate power using DC for decades. It's worth it to them to convert from AC to DC, transport the power, and convert back to AC on the other end. Worth it on thr basis of efficiency, at acceptable capital expense. You're cooking the numbers to push the hydrogen thing. If you actually has solar power at 1/5 cents per kwh, there would be no reason to not sell it to the grid. > >The big question anyone who wants to have solar be an important source >has to ask, is where to start? After 60 years of development far less >than 1/1000th of our total energy supply comes from direct solar. >That means we're doing something wrong. It means that sunlight is a very low-density power source. And always will be. John > >So, I decided to do something different! And I've described my >reasoning and my approach and my success. To which a blindered and >shuttered community of enthusiasts, stuck in a world where the solar >industry is dominsted by the major oil companies, cannot seem to see >the forest for the trees. Tell us about your success. Show us links to working systems. > >There is adequate reason to believe that when you start engineering >large DC powered interties (much larger than that found on homes) you >will pay more than the $0.70 per peak watt (10 x more than my panel >costs) first time out. More likely $2.00 per watt (3x a mature >market0. Absurd. Utilities do gigawatt DC links all the time, and they don't pay multiple gigabucks for them. This is totally bogus. John
From: Willie.Mookie on 5 Oct 2007 10:05 On Oct 5, 9:36 am, Eeyore <rabbitsfriendsandrelati...(a)hotmail.com> wrote: > Willie.Moo...(a)gmail.com wrote: > > Eeyore wrote: > > > Willie.Moo...(a)gmail.com wrote: > > > > To intertie to the grid would increase costs to 6 cents per kWh at > > > > best > > > > Based on a silly estimate of the cost of grid tied inverters which you have > > > overestiamted thge cost of by at least 3 times and probably more. > > > > With a sensible supply of inverters that would be 2 cents / kWh or less. > > > Graham, you might be right that $2 per peak watt is high. But it is > > at the high end of the envelope. Fact is, no one is building DC > > intertie at the 50 MW scale and above so no one knows what all the > > costs will be. > > You need to talk to someone like ABB. I have! And I must say their transformers are one of the engineering points we used in our estimates of costs of building large scale solar panel arrays tied dirextly to the power grid. Lets make a distinction between AC/DC intertie used by Asea Boveri and Brown in their HVDC lines, which operate at farily continuous power transmission with slight variations over many days, varsus tying together huge arrays of solar panels that vary from 0 to 100% output in less than an hour. These have similarities certainly, but also substantial differences. When you sit down and actually start to build that system, you find costs, at least the first time through, escalating. I stand by my $2 per peak watt for the first 5 GW of installed systems. I will say that by the time the first 1,000 GW are installed, prices might be down to $0.70 - Note, I'm not arguing that direct intertie isn't efficient. It is hugely efficient. I'm arguing its not the path toward initial development of large-scale solar power. > > > The first one won't be $0.40 per peak watt - which is > > what you're suggesting I think. Though with a dedicated effort you > > might get down this level. A real first system - for planning > > purposes - I think $2 per peak watt is a conservative number (that is > > its high - but likely the first go round) > > Very high if I can readily find grid-tied inverters on sale today at the 3kW > level for < $1 a watt. Sit down and read a book or two on building power grids. Even if you get say 4% of your ENERGY from sunlight, you'll be getting at peak output 30% of your power. That's a hole different ball of wax to supply 30% of an entire grid - stably reliably - with highly variable solar power. Put differently, a designer of a 3 kW system attached to a grid supplied by a 3,300 MW power plant, doesn't have to worry about 1/10th of the things as a designer of a 1,000 MW solar panel array attached to the same grid. All those costs of the larger system are properly assigned to the intertie. > > You still have the problem of storing energy outside of the 'solar hours' of > course. Well, to be generous I didn't even get into that! haha.. But yes, this does add substantial cost. However, since energy storage systems can double as sources or sinks, a clever designer can use an inexpensive storage system to lower some of the costs - which is the direction I'm going in with my approach long-term > Graham
From: Willie.Mookie on 5 Oct 2007 10:34 On Oct 5, 2:51 am, Eeyore <rabbitsfriendsandrelati...(a)hotmail.com> wrote: > Willie.Moo...(a)gmail.com wrote: > > John Larkin wrote: > > > Willie.Moo...(a)gmail.com wrote: > > > > >(7) Adding an inverter and peak power matching hardware to solar > > > >panels cost $2 per peak watt. > > > > That sounds exhorbitant. I can buy small PF-corrected switching > > > supplies for 25 cents a watt, > > > Their efficiencies are far less than unity, and this is but one step > > in the inversion peak power matching process and then synchronizing > > with the AC grid in such a way as to contribute that power > > efficiently. > > Grid tied inverters here at retail for under $1 a Watt. Around 95% efficiency > too.http://www.grid-tie.com/Xantrex-GT-Inverters.html Those systems are kilowatt level or less. Imagine a local grid supported by 3,300 MW of conventional power plants. Add a 1 kW solar panel to it. The person designing the solar panel and attaching it to the power grid, doesn't really have much to consider Now, lets assume this is a place like Pennsylvania where due to weather you can't get more than about 4% of your energy from sunlight if its tied to the grid. That means that about 40% of your power will be coming from sunlight on the sunniest days. (we're not assuming storage, just direct intertie which is highly efficient) The intertie designer has a lot more to consider in maintaining grid stability - especially if his peaking plants take minutes or more to respond, and only account for less than 40% of your power. And this is only one diseconomy of scale. There are others. > I'd expect larger capacity units to be less expensive and it seems pretty clear > to me that as a very large customer you would be paying no more perhaps than > 30-40c a watt Sure, if all you were doing is installing the same intertie a million times to 1 kW generators, but at some point you start actually producing enough to affect more than your power meter. At some point you start actually changing the load of the entire grid to such an extent that you have to account for changing lighting conditions throughout the day due to weather and time of day. Since these conditions can change in seconds and affect huge amounts of power generation, they have to be accounted for in the intertie system - and paid for as intertie costs. My approach is to focus on cost and make hydrogen the ultimate intertie, and as long as I sell heating value at less cost than it costs to buy it from carbon sources, I've beat the carbon sources at their own game, I'm making a profit, and I'm well situated to dominate the market, not merely make tiny inroads to it. > > You're not really getting it. Or rather, you are letting your > > prejudices dictate to your mind what's possible. Which is why you are > > making foolish mistakes. > > It seems YOU are the one with prejudices here actually. That might hold water if I hadn't spent something like $400,000 studying direct intertie and had a team spend months looking at it and pulling their hari out. Efficiency was only ONE of MANY criterion - a more important one being profitability, and market dominance. When you start looking at doing everything with solar, you quickly come upon some diseconomies of scale that you need to address. Then once you've addressed them, you then choose, what's the quickest, least risky, path to dominance. And the answer is, hydrogen. There are other paths - in fact there is a path I call Genesis planet. You put an array of solar collectors in the sunnieset areas, and you build HVDC lines to connect them all to load centers - cities, towns and so forth. Each person has a demand curve for energy that they go through each day, and that is modulated throughout the year with season. This is a function of time, longitude and latitude. So, you can take population distribution and plug in this demand function with time for that population and get demand for the planet consuming power as an average US citizen say. Then, you take all the insolation data, combine that with your panel costs, land costs (which goes down the further you get from population centers! haha), and transmission costs - and you can calculate an optimal array of networked panels and intertie that exactly meets the needs of the planet. Excepting to meet the needs without any brownouts, you need to have oversupply at certain times. What do you do with the oversupply? Short it out, or shut it down. You can make fertilizer with it. You can make hydrogen with it. Then the question becomes, what do you do with the hydrogen? Remember this is Genesis planet, everything is solar - tied to panels. Even the cars tied to the solar grid - like slot cars, or electric trains, or trams or magnetic coupling. So, even Genesis planet isn't as efficient as the physics suggest, because when you install the panels and grid and switches onto the real Earth with the real population here - you end up shorting out about 30% of your produciton. Which doesn't effect cost that much,but is sure as hell is a lot of waste! Then, if you look at actual losses in long distance HVDC lines - the idea of using hydrogen as the medium (the famous (in our offices) solar proton and solar electron debate! haha) So, yeah, we looked at a lot of things - very seriously - and hired the best and brightest to do it. The detailed results are proprietary. But our game plan is clear. Especially when you go beyond just the distribution of people and sunlight and start to factor in the availability of other energy systems, the possiblity of growth, and so forth - and you find that there is a least risky, least costly path - which unfortunately relegates home panels to second or third teir status in the developed economies, and first or second tier status in developing countries, depending on where they are. But in terms of market dominance, developing countries isn't where the money is, or where the energy is used. Its needed, but not used. So, the game plan is clear. Dominate the hydrocarbon fuel production with solar assisted hydrogenation of low rank low value fuels. With those profits create an integrated oil company from a collection of remarketers and appropriately located coal mines. Then use the hydrogen distribution system that supports this operation to fuel stationary power plants, taking stranded fuel in trade. Then, use the hydrogen distirubtion system along with your retailing operation to sell hydrogen directly to mobile users. Meanwhile, with 60 GW a year production,rising to 1,400 GW per year in a few decades, 2 GW per year rising to 160 GW per year of solar panels will be sold for remote applications, mostly in developing regions. Also, an increasing number of panels at central locations will be shipping electrons instead of protons into a national and then international HVDC power grid - supplying up to 20% of all our energy needs in the best case - in the most developed nations.and another 4% of our energy in these same places will be in homes offices and factories that are off grid altogether. > Graham
From: Martin Brown on 5 Oct 2007 10:52
On Oct 5, 3:03 pm, John Larkin <jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote: > On Fri, 05 Oct 2007 11:16:57 -0000, Willie.Moo...(a)gmail.com wrote: > >On Oct 5, 2:53 am, Eeyore <rabbitsfriendsandrelati...(a)hotmail.com> > >wrote: > >> Willie.Moo...(a)gmail.com wrote: > >> > To intertie to the grid would increase costs to 6 cents per kWh at > >> > best > > >> Based on a silly estimate of the cost of grid tied inverters which you have > >> overestiamted thge cost of by at least 3 times and probably more. > > >> With a sensible supply of inverters that would be 2 cents / kWh or less. > > >> Graham > > >Graham, you might be right that $2 per peak watt is high. But it is > >at the high end of the envelope. Fact is, no one is building DC > >intertie at the 50 MW scale and above so no one knows what all the > >costs will be. The first one won't be $0.40 per peak watt - which is > >what you're suggesting I think. Though with a dedicated effort you > >might get down this level. A real first system - for planning > >purposes - I think $2 per peak watt is a conservative number (that is > >its high - but likely the first go round) > > >So, ask yourself, at what scale? > > >It is true economies of scale and leaerning curve effects do reduce > >the cost of smaller units, and may reduce initial costs well below the > >$2 estimate my detailed engineering analysis came up with. > > >You by comparison have merely indicated a suspicion of where intertie > >might be after a decade or more of large-scale development. > > Hogwash. In 90 seconds of research, you can discover that UPS systems > - charger, batteries, inverter, enclosure, line cord, LED dislays - > are going for 10 to 15 cents per watt RETAIL. > > Utilities have been shipping interstate power using DC for decades. > It's worth it to them to convert from AC to DC, transport the power, > and convert back to AC on the other end. Worth it on thr basis of > efficiency, at acceptable capital expense. How so? I thought the fundamental limitation of DC was that it was harder to transport over long distances (even at 500kV which is what I think the big DC interconnects use). I know they do HV DC mains grid interlink in Japan, but that is because the country has 50Hz and 60Hz power grids in different parts of the country. Why do US utilities do DC interlink on a single frequency? The size of the USA compared to the wavelength of 60Hz is not negligible, but it doesn't seem excessive either. > You're cooking the numbers to push the hydrogen thing. If you actually > has solar power at 1/5 cents per kwh, there would be no reason to not > sell it to the grid. Indeed. And there is an obvious conclusion. > >The big question anyone who wants to have solar be an important source > >has to ask, is where to start? After 60 years of development far less > >than 1/1000th of our total energy supply comes from direct solar. > >That means we're doing something wrong. > > It means that sunlight is a very low-density power source. And always > will be. There have been a few impressive solar concentrators. Largely for smelting exotic materials research and the odd one can do superheated steam generation too. But given the expense of refined silicon for high efficiency solar panels it seems worthwhile to at least double the incident solar flux onto them by using mirrors. It makes the wind loading worse of course. But mirrors are cheap and solar panels expensive (even more so when you have to worry about keeping them weatherproof). If someone can make plastic semiconductor based photovoltaic cells that are chemically stable under sunlight then the economics could shift, but for the moment solar power is best at doing hot water supply for suitably low latitudes (and total waste of time at 55N where I live). Regards, Martin Brown |