Inverter info for high power systems [Archive]

Lion Kuntz

14th January 2007, 04:44

I bit off more than I can chew, so I'll appeal for input.

I am trying to conceptualize the H2-PV world. So far I have been able to see how the polycrystal PV part can be created and cost justified, but the going gets sticky with the with the high power required in key places.

One acre of PV cells is 523 kilowatts, which one can blithely say. It's easy to say that an EMC casting furnace uses 12 kWhs/kilogram and can process 30 kilograms Si throughput per hour. It's another thing entirely to try to get a handle on 360 kVA radio frequency power.

Going from PV DC to RF AC is one thing. That's just in the silicon production end of it. Once the PV is created and installed, a 2,000 square foot blue PV roof of 13% efficiency PV is 24 kilowatts, considerably more than any home systems anticipate.

The EMC casting system is so cheap that the PV would be far less expensive than the inverter systems on the market today.

Basically, inverters are a serious bottleneck in going to an H2-PV world.

There are already 3 critical bottlenecks in making the PV itself, and power supply (initially grid electricity ultimately replaced by PV made by the grid power as bootstrap) as mentioned above. I believe two of the three are mainly solved by now, but the third bottleneck, power supplies is related to distributed blue rooftops too. Whether it's an acre of PV cell surfaces or a 2,000 sq.ft rooftop, the inverters are way out of line in costs from everything else.

Considering that H2-PV furnace operations will be casting 720 kilograms of silicon to solar grade ingots every day, which are basically big blocks of diode material, I am wondering if anybody can help me see the pathway how this can be useful in solving this critical bottleneck issue.

Lion Kuntz

14th January 2007, 19:57

Hi Lion,

I'm not sure what the objective is of what you're trying to do. Is this to show that a "silicon/solar energy economy" can be self-sustaining?

There have been studies to calculate how long it would take for a solar panel to produce enough energy to offset the energy that went into making it (the term used is "embodied energy"). An NREL study (http://www.nrel.gov/ncpv/pdfs/24596.pdf) showed this to be 1 - 4 years, with the thin-film types being fastest, because it uses the least materials and thus energy to produce (Evergreen's ribbon technology would be a type of thin-film).

-Rob-

I'm doing my own study because I don't like the political-corporate influence around the DOE/NREL. My own studies show 13 months energy payback in Arizona, 18 months in Florida. You have to move way up North before it takes 2 years for an acre of 13% efficient polycrystal PV to breed one replicant of itself. That's only true if calculated that the energy used is valued at California retail rate of 12 cents/kWh. Using PV energy created by prior PV instead of grid retail energy the energy payback is under 45 days in the Southwest. One acre of PV makes 3.1 megawatt-hours per day in the sunbelt: anybody who can't make an acre of PV with 140 megawatt hours of power should get out of the business and leave it for the professionals.

In the 1970s there were oil shocks as the Iranians and their arab OPEC buddies flexed their muscles, followed by a surge of interest in all thing alternative energy. Jimmy Carter put PV on the whitehouse, Reagan took them off again. The interest faded in the 1980s as gunboat diplomacy assured cheap oil supplies through Clinton's first seven years. It's a little hard to remember gasoline below $1 now, but I last paid that in 1999, just seven years ago.

The Gold Rush to cash in on solar in the late 70s and early 80s created a legacy of patent technologies, fully detailed and explained, which have now expired. For example, the NREL patented the EMC furnace technique to cast silicon in 1986, the method now used by 64% of all installed PV in the world. Thin-film is an insignificant blip on that radar screen.

While making 14%-16% efficiency PV seems low efficiency by 2007 standards, it is such a cheap process that it makes more wattage per dollar than anything else. Space is not such a premium as dollars are -- you can trade space for fewer dollars, as in $50/acre cactus-lands around Abilene, than you can trade the exact same dollars for high density PV efficiency. How many watts of 20% efficiency PV can you buy for $50?

Polycrystal silicon is my choice, and I am long past the argumentation stages about that. Now I intend to do the penny-by-penny analysis how Silicon blue rooftops will cover every single home in America.

These three patents: # 4588571 (http://h2-pv.us/wiki/tiki-index.php?page=4588571), # 4457903 (http://h2-pv.us/wiki/tiki-index.php?page=4457903), and # 4572812 (http://h2-pv.us/wiki/tiki-index.php?page=4572812), go from raw mined sand to cast solar-grade PV silicon ingots for less than 6 cents per square foot in energy costs (http://h2-pv.us/wiki/tiki-index.php?page=Bulk_Sand) if you use my high retail rate 12 cents kWh to do the math. The energy price goes down drastically when you have produced enough PV to pull the plug from the grid and power the furnaces off PV power you made yourself.

The Silicon is not the bottleneck: the collecting the streams of power from PV panels into rivers of high power is now the bottleneck. The inverters now are the bottlenecks.

Unlike silicon purification to PV, the 70s-80s Gold Rush did not contemplate grid ties, but was based on stand-alone battery-backup systems. There isn't the wealth of scores or hundreds of expired open-source patents to choose from for grid-tie expired patent ideas.

Most of the cost of inverters is in know-how, not materials, not manufacturing complexity. There's heavy metal in inverters and transformers, but not so much as in a car engine I can buy reconditioned at any auto parts store for $700. There's complicated precision electronics in inverters, but nothing as complicated as the 18x DVD burner I just paid $35 for retail price one-off. The combination of heavy metal and electronics does not justify prices of $5k or higher for a PV rooftop inverter system.

There are 105 million homes in America, 75 million of them detached single-family types, average size is 5 rooms. The inventor ought to be happy with $1/family and take his $75,000,000 knowhow and retire, else he risks others from reinventing the wheel to bypass his throttle point bottleneck.

PV costs the same as beer bottles and beer cans whenever it is made in the same quantities as beer bottles and beer cans because it's made out of the exact same stuff with trivial differences. The cost of PV should be 30 cents a square foot in ten years after the next imminent Gold Rush finishes.

The cost of pure enough silicon was the bottleneck, now it's not. The cost of waferstock cast crystals was the bottleneck, now it's not. As each bottleneck gets solved the choke-point moves somewhere else. We need to keep moving the choke points while they get progressively easier and smaller until there are none left.

For now, the bottlenecks are in the high-power apparatus, both supplying the grid and supplying the casting furnaces. These are the places where up to 9/10ths of the costs of complete systems can be shaved off, and will be.

I have descended all the way back to Nikola Tesla's 1880s patents as the grid was being invented in the first instance. I'm going to sift through the 1,000s of patents until I see a pathway that bypasses the bottlenecks that doesn't have the inventor raking off 50% to 90% of the inverter/transformer costs just for "knowhow".

As I said up top, I have bitten off more than I can chew. If I don't get some other people inputting their knowhow, WE are not going to meet a target of 30 cents a foot PV in ten years. Since 13% net efficiency is 12 watts/sq.ft., that 30 cents a foot translates to 2.5 cents/watt 13% silicon polycrystal PV.

You can think of a lot of ways that America would be greatly improved by abundant wholely clean renewable energy in ten years. What valid excuses can you offer for delaying that date?

FACTS:

* 13% efficient PV is 523kWh/peak sunny hour per acre of PV cell surfaces.
* 9 southwest sunny states have big parts or all that get a guaranteed daily average of 6 peak hours of sun round the year. That's daily 3.1 megawatt-hours per acre of PV cell surfaces, 1,145.4 megawatt-hours production per year per acre of 13% efficiency PV cell surfaces. That's "good enough".
* One acre of 13% efficiency PV cell surfaces mines equivalent energy per acre as one tone of coal ever 11.7 hours of peak sunshine, 183 tons of clean carbonless pollutionless coal mined every acre every year. In a 20 year warranty period one acre of PV mines 3,650 tons of clean carbonless pollutionless coal energy. The PV is BOTH the fuel and the powerplant, so comparing PV to coal must include the cost of 20 years construction and operation of the carbon-coal plant. Honestly, you should include the negatives of burning carbon-coal as costs too.
* One acre of H2-PV can electrolyze water to hydrogen and oxygen at a rate of 12 kg H2 per hour, 72 kg/day, 26,280 kg/year. That's 26,280 clean carbonless pollutionless gallons of gasoline equivalent.
* H2 can power your existing car of truck with a minor conversion rather than scrapping the entire fleet for fuel cells or PHEVs.
* 2.4 gallons of water makes 1 kg of H2 and 8 kg O2 per 43 kWh. About $6 of retail priced electricity will make 1 kg H2 plus 17 kg H2O2, both rocket fuels, with a current retail value of around $26 in the bulk sales wholesale volume market.
* 340,000 acres of prime sunny desert Arizona lands were planted in cotton in 2006, which takes 3 acre-feet of federally subsidized irrigation water (1,000,000 gallons) to produce an annual crop that nets the farmer $300 per acre per year. An acre of H2-PV produces 26,280 kg/H2 per year from 63,000 gallons of water, with a gasoline equivalent price value of $65,700 per acre.
* Zeolite-ZSM-5 sores 3/5ths the density of H2 as liquid H2, but it stores it at room temperatures and at normal pressure, without leak hazards, without fire or explosion hazards. Too heavy and bulky for transportation, but fine for stationary depot storage. Chabasites and activate charcoals store even higher densities. Then there are hydrides...