Back in 2003, wind farms cost between $1,000 and $1,300 per KW to install. A lowered dollar, increased embedded energy costs, and increased demand has driven this capital cost up to between $1,300 and $1,700 per KW. [Note: since this post requires a lot of documentation, I've put it all in a Google Spreadsheet, which you should be able to open in any browser.]

Given that we would be buying in volume, and could use the opportunity to develop our home wind turbine industry again (not quite as much at the mercy of the value of the dollar), it seems reasonable that we could keep our price at the $1,300 low end.

If we look at the actual capacity use of wind generators (around 30 percent, though new ones are a bit higher), the capital costs to replace electric consumption in 2005 would be under 2 trillion dollars. (Note that as demand increases, these numbers scale.) To keep storage needs to a minimum we would want to connect wind generators up to 3,000 kilometers apart by HVDC transmission lines. Basically we would want to connect the different regional and local grids, so that all our various DC islands were connected by DC bridges, producing a more reliable super-grid. (DC lines are much less vulnerable to outages that AC.)

Studies have shown that while a single wind generator would require about three days of storage to provide firm capacity, when you put those generators in large scale wind farms and connect many wind farms great distances apart, they can provide firm capacity with 80 percent reliability, even without storage. Twelve hours storage would bring this up to 96 percent reliability. (As a side issue, 22 hours would bring reliability to 99 percent +. Also, twelve hours pumped storage would consume less than 50 square miles over the entire U.S.; 22 hours would increase this to under 83 square miles.)

Electricity losses due to transmission and storage above that required by our current grid would bring the total cost to 2.35 trillion dollars. The cost of twelve hours worth of pumped storage, and 100,000 kilometers of HVDC, would increase that to under 2.5 trillion dollars.

This seems like an overwhelming amount until you consider the following:

We currently pay slightly under 300 billion dollars every year for electricity.

Around 66 percent of this (around $200 billion) reflects generation costs.

Wind farms last twenty years or more. At the interest rate the U.S. government currently pays for taxable twenty-year bonds, quarterly payments on those bonds would be less than the generating costs saved. You would not want to borrow $2.5 trillion in one go, and you could not build that much wind capacity in a year. Instead, phase out grid emissions over the course of a decade. You could borrow the money 250 billion at a time, or you could end the war in Iraq, subsidies to oil, gas, and coal, and repeal a fraction of the tax cuts we've suffered, and pay for it out revenues without borrowing a dime.

You would reduce greenhouse emissions from electricity production to near zero and massively reduce air pollution and the number of people who die from it. As a bonus, we would save slightly on our electricity bills, and build a wind turbine industry competitive with Denmark's.

There are all sorts of reasons this would be a fundamentally stupid thing to do. It does not include efficiency measure that could probably save more than half our current electricity use; it replaces low-carbon existing electricity sources; it produces an electricity monoculture relying entirely on wind, hydro, and pumped storage, without including any other renewable technologies like solar, wave, or geothermal. You can probably come up with other reasons it would be stupid yourself.

What you can't come up with is any reason it would be stupider than what we are doing now.

A last reminder: I've put the details in a Google Spreadsheet, -- "showing my work," with sources for facts and statistics documented. As a writer I like this better than linking every four words, or including a bunch of end notes in the document. It flows better. (And it is less work than an inline HTML table.) Let me know if it works for you as readers.