According to the California Energy Commission, PG&E currently sells about 5.4 billion kilowatt-<\h>hours of electricity to customers in San Francisco. (This figure doesn't include energy used by the city government, since government agencies use power from the city's Hetch Hetchy dam.) Residential, commercial, and industrial customers all pay different rates. If a MUD sold power at current PG&E rates (as provided to us by PG&E spokesperson Ron Low), it would bring in $562 million in revenue (enough to create a big annual surplus - roughly $36 million.) But a MUD or power agency almost certainly wouldn't sell power at PG&E's high rates - one major attraction of public power is that it offers cheaper electricity. So in both of our scenarios, we assumed that the rates would be at least 10 percent below PG&E's rates. In fact, as our study shows, rates could drop as much as 20 percent without harming the MUD or WPA's viability. What's it cost? There are three basic categories of costs that the agency would have to cover. The first is payments on the bonds, the second is generating or buying power, and the third is basic operations and maintenance (paying the staff to keep the system up and running, to send out bills, to read meters, as well as operating the repair trucks, etc.). Electricity can't just be delivered to the doorsteps of customers like canned ham in a UPS box. It has to be distributed through a network of transformers, substations, wires, and poles and measured with individual meters. And until the public power agency owns that distribution network, it can't sell a single kilowatt. Unfortunately, the system that's now in place in San Francisco is owned by PG&E - and almost everyone involved agrees that it would be cheaper, easier, and quicker for the city to take over that system than to build a new one from scratch. That's what SMUD did and what most other public agencies that have gotten into the power business in the past half century have done. A MUD or a city power agency would have the right to seize PG&E's property by eminent domain. But PG&E would be entitled under law to fair compensation for the taking of its property, and one of the most complex, bitter - and crucial - issues involved in establishing public power will be the price tag. "This is not an easy case at all," Richard Epstein, a professor of law at the University of Chicago and a national expert on eminent domain, told us. "I can guarantee you that nobody, but nobody, has any idea right now what fair compensation would be." The issue will almost certainly be settled in court. PG&E insists that its San Francisco property is worth a small fortune - as much as $1.4 billion. In a 1996 study the Economic and Technical Analysis Group suggested that the price could be anywhere from $315 million to $1.2 billion. The ETAG study, which was highly favorable to PG&E, suggested that the most likely figure was around $795 million. The reason those figures are so widely divergent is that there are numerous ways of evaluating what a utility's property is worth. The simplest is to establish what PG&E originally paid for the property, then factor in depreciation. That's how insurance companies decide what they have to pay you if your car is stolen. The process generally leads to a low figure favorable to the city. But courts have recently been somewhat more friendly to an analysis that recognizes that utility property is more valuable than, say, a private car, because the utility property produces income. One way to address that is by valuing the property at its replacement cost and factoring in the value of a "going concern" - which, of course, leads to a much higher price.