San Francisco — After a meteoric rise in the 1960s and '70s, the star of the United States nuclear power industry has gone into eclipse. If it is not to deteriorate into a mere caretaker of existing reactors within a decade or so, radical changes are needed. But there is little unanimity about how to make those changes, or, indeed, whether such revitalization is possible.
The biggest unknown in nuclear power's future is demand for electricity through the end of the century. Projections of annual growth vary from 0 to 4 percent. If demand grows less than 1 percent a year, the industry will need very few big new power plants of any type until well into the next century. But if demand grows at a more robust 3 to 4 percent, a substantial number of new plants will be needed, beginning in the late 1980s.
The reason for the uncertainty is that no one is sure how much of the recent decline in demand for electricity has resulted from recession and how much from conservation. (Between 1976 and 1982 annual growth in electricity sales dropped from 6 percent to - 1.5 percent.) The answer to this key question will become apparent within a few years if the nation's economic recovery continues. But currently it remains a matter of sharp debate.
On one side, ''Electricity consumption has continued to track the GNP (gross national product) pretty well,'' argues Carl Walske, president of the Atomic Industrial Forum, an industry trade group.
Last year, Mr. Walske compiled a list of things he feels must happen before the nuclear industry can expect any new orders. Increasing electricity demand between now and 1986 heads the list.
''This year electricity demand turned up for the first time in several years. So far so good,'' Walske says.
Others are less sanguine. Charles Komanoff, an independent industry analyst, attributes the 2.5 percent increase in demand last year primarily to an abnormally cold winter and hot summer.
''Compensating for this, you come out much closer to half a percent,'' Mr. Komanoff says. At most, he predicts, electrical sales will grow a half percent a year through 1989. He cites several reasons for this: When newer, more costly nuclear power plants come on line, rapidly rising electric bills will galvanize new conservation efforts. Cogeneration, technology that produces electricity and industrial process heat at the same time, is coming along rapidly and will reduce industrial demand. And natural gas prices are stabilizing, so there should be fewer factories converting from natural gas to electricity in the near future.
Another prerequisite for a nuclear power revival is that reactors be an economically competitive source of electricity. This is reflected by the next two items on Walske's list: lower inflation rates and lower interest rates.
Because nuclear power plants are such large construction projects and inflation in the construction industry tends to run much higher than at the consumer level, they have been particularly ravaged by inflation. Because a nuclear reactor is the most costly type of power plant and counts on its extremely low fuel cost to compete with other generators, it is most affected by high interest rates.
On these counts, the industry spokesman says he is heartened by the decline in the rate of inflation. But, while the cost of money to utilities has also dropped, it is not low enough to favor nuclear projects.
The economics of nuclear power are as controversial as any other aspect of the industry. For instance, a recent paper by Worldwatch Institute, an environmental policy research group, argues that it has failed the market test: ''Enough data exist to show conclusively that new nuclear power plants are not cost effective in the United States compared with other new plants. Even if all the unique safety and health dangers of nuclear power were removed, a US utility planner choosing between a coal and nuclear power plant solely on the basis of economics would have to select coal.''
While this may be true on average - a conclusion vigorously disputed by the industry - it does not appear to hold for the most effectively managed nuclear plants. It has become clear that the economics of nuclear power vary radically with a utility's management ability. Thus, construction costs vary by threefold and the amount of electricity a completed plant produces by twofold.
The Palo Verde reactors operated by Public Service of Arizona are an example of a well-managed nuclear program. The utility estimates that the cost of electricity produced by these plants in 1986 will be 10.4 cents per kilowatt-hour compared with 11.4 cents from coal.
Standardization alone can make a vast difference in nuclear costs. It is clear that if any new plants are ordered, they will be of a standard design, rather than the custom plants that have been built up to now. The Palo Verde reactors are a case in point. The three units are identical.
''I couldn't feel more strongly about standardization,'' says Tom Woods, vice-president of the Arizona Nuclear Power Project. ''Construction costs, excluding labor, will be about 15 percent less for Unit 2 and 25 percent less for Unit 3. The savings amazed even us!''
So nuclear power can be cost-competitive with coal, considered its major challenger. But increased electricity sales and a substantial cost advantage will not be enough to sell new nuclear power plants. This is reflected by the final items on Mr. Walske's wish list: regulatory stability, and new methods to decrease the financial risks of nuclear power.
Regulatory stability has yet to be achieved, Walske maintains. But Manning Muntzing, a lawyer who headed federal reactor regulation when it was still under the old Atomic Energy Commission (AEC) and more recently served as president of the American Nuclear Society, sees positive signs on this front.
''When I was with the AEC regulatory division, we had a very strong emphasis on quality assurance. I think the NRC (Nuclear Regulatory Commission) got away from that in the late 1970s. Now I see them coming back to our old position. As a result, they have been finding things that are wrong. This reads badly but, quite frankly, it is good,'' Mr. Muntzing argues.
While the nuclear industry has a formal tendency to blame federal regulators for its problems, much of its difficulties stem from a basic mismatch between the capability of the utilities and a very demanding technology. The measures proposed to deal with this apsect of the situation fall into two basic categories. One involves upgrading the capability of institutions charged with operating and policing nuclear power. The second involves major changes in the technology to make it safer and more manageable.
One likely institutional change is the construction of ''turnkey'' nuclear power plants. In this approach, a nuclear reactor manufacturer agrees to build a power plant for at a set price. Because of the size and complexity of nuclear power plants, ''there are a number of utility executives who doubt their ability to manage these plants. We want to help share the risk. But, if we do, we want overall project management,'' says James Moore, vice-president and general manager of the water reactor division of Westinghouse.
But the congressional Office of Technology Assessment (OTA) points to a potential hitch in this approach: Once a utility takes over a plant, it may know much less about it than it currently does.
This could be overcome by a second innovation that has support within the industry: the formation of special regional power companies to build and operate nuclear plants. This is a logical extension of the current trend toward multiple utility ownership and single utility management of reactors.
There are several advantages to such a reorganization. It would place the construction and operation of plants in the hands of elite cadres with the greatest experience and capability. It would spread the risk over a larger base. It would allow the industry to continue to benefit from the economies of scale of the large 1,300-megawatt powerplants: Even at a 2 to 3 percent growth rate, these large plants represent larger chunks of generating capacity than most utilities will be able to swallow. It is also possibile that such a company might be exempt from state rate regulation. It might be difficult, however, for a company whose sole assets were nuclear powerplants to obtain financing.
An alternative to this is the formation of new federal agencies similar to the Bonneville Power or the Tennessee Valley Authorities, either to own or to finance nuclear plants in various regions of the country. According to the OTA, this ''may be the only way to maintain nuclear power as a national energy option.'' But there is no guarantee that the federal government would be any better at managing nuclear plants, OTA researchers admit.
A fundamentally different way to attack the industry's problems is to improve the basic technology, and there are two different schools of thought on how to do this: improve current reactor designs or shift to entirely new designs with greater intrinsic safety.
The Japanese Ministry of International Trade and Industry has taken the lead in the gradual approach. They are coordinating a national effort involving Japanese utilities and reactor vendors Westinghouse and General Electric to improve current designs.
For example, Mitsubishi and Westinghouse have teamed up to produce an advanced pressurized water reactor. ''Rather than stick with designs more than a decade old, we wanted to start with a semiclean sheet of paper, to come up with a new design with reduced cost and construction time and improved safety,'' Mr. Moore of Westinghouse explains.
They intend to build the first such reactor in Japan beginning in a year or two. They have also filed for a preliminary design approval with the NRC.
But there are those who argue that a radical, technological break with the past will be necessary to restore public and utility confidence in the nuclear option. A prominent nuclear scientist, Alvin Weinberg, has authored an influential report that comes to this conclusion.
Today's light-water reactors (LWRs) are direct descendants of designs developed for military purposes. Since then, a number of designs have been proposed that appear to have some safety and operational advantages over LWRs. The one with the most support is called the high temperature gas reactor (HTGR).
HTGRs have a long history. In 1958, a group of utilities chose it as the one most suited for commercial use. While their ideas gained some support in the AEC , most of the impetus has come from industry. Through last year, about $1.5 billion has been spent in the United States on this concept: $500 million by the government and $1 billion by the private sector. Philadelphia Electric Company operated a small test reactor of this type from 1967 to 1974. And Public Service of Colorado has been running a prototype 330-megawatt HTGR since 1974.
In today's LWRs, uranium fuel is held in long tubes called fuel rods. These are bundled and placed in a reactor vessel through which water is pumped. The water is a moderator, slowing subatomic particles to the speed necessary to sustain the chain reaction that converts minute amounts of uranium into large amounts of energy. The water also carries heat out of the reactor to generate electricity.
The HTGR's uranium fuel is interspersed with graphite, which acts as the moderator. The heat is removed by pressurized helium gas, which is pumped through the reactor.
The gas-cooled design is safer than the water-cooled for several reasons:
* The helium does not pick up nearly as much radioactivity as does water, thus workers in HTGR plants are exposed to much lower amounts of radiation. Also , helium is noncorrosive, so it causes little damage to core components.
* The power density in the core is 1/10 that of an LWR and the graphite can absorb large amounts of heat before becoming damaged. This means that operators have hours instead of minutes to correct malfunctions before the core is damaged.
The HTGR also has some potential economic advantages:
* Because the helium can be heated to higher temperatures than water, HTGRs are more efficient. They can also be used for cogeneration.
* In some newer designs, they can be built so that they are fueled while in operation. Current reactors must shut down frequently for refueling, which reduces their availability.
What has stirred the most interest recently, however, is the possibility of building modular HTGRs of about 100 megawatts in size. This type of ''inherently safe'' design would keep the core from overheating by natural convection, even if all the plant machinery failed.
Not only would this be added protection for the public, but it would also safeguard the utilities from the risk of destroying millions of dollars of property.
The modular HTGR was the subject of a recent study at the Massachusetts Institute of Technology (MIT). Its conclusion was that this type of reactor ''can, in principle, help remedy the current woes of US utilities,'' says Lawrence M. Lidsky, a professor of nuclear engineering at MIT.
But this concept has yet to be clearly defined, says Merwin L. Brown, director of program development at the Gas Cooled Reactor Association (GCRA), a group of 30 utilities formed to promote this technology. Because of the surge of interest in smaller reactors in general, and inherently safe designs in particular, they are seriously evaluating the concept, he says. The basic idea is to make these reactors small enough so they could be fabricated in a factory and shipped to the site on special trucks or barges. There, the reactors would probably be placed in underground silos. A utility might buy four of these and use the steam from them all to power a turbine. These would likely cost more per kilowatt than current reactors, although mass production might offset some of the economies of scale that would be lost. But their smaller size and lower risk might still make them attractive.
This year for the first time a president's administration jumped aboard the HTGR bandwagon. The Department of Energy has $35 million in its 1985 budget earmarked for this purpose. In the past, it has been supported solely by Congress. And several pronuclear congressmen are calling for major research on inherently safe designs.
The cards now seem stacked squarely against the nuclear industry. But new institutional arrangements could upgrade its management and new technology transform nuclear power from the most demanding to one of the most forgiving ways to generate electricity.