AS THE LAST MILITANT neutral on the subject of nuclear power, and as a managing director of the nuclear industry's best customer -- the Tennessee Valley Authority -- I want to tell the industry's advocates that their basic argument these days is wrong.

Both the industry and its critics agree that the nuclear option is in serious trouble. Where the industry goes awry is in the cause of its demise. What's killing it is not the Nuclear Regulatory Commission or the media or Ralph Nader or Jane Fonda. It is their product itself. The nuclear industry is stonewalling the hard evidence that they need to change the design of their reactors.

The litany of nuclear reactor cancellations throughout the United States is a long one. Between October 1980 and June 1982, plans for 21 were abandoned -- more than one a month. When I joined the TVA in 1977, 14 large ones were under construction, the nations' most ambitious nuclear power program. Over the past four years, we completed two reactors, but halted work on eight others because the cost of making them safe was simply going to be too much.

The nuclear industry will lose credibility altogether if it persists in trying to blame its problems on an emotional public and a regulatory process overreacting to ill-founded fears.

It's time to confess that we went too far too fast in deploying large-scale designs of a reactor type we knew too little about. Right now, we are in the midst of a de facto moratorium on new reactor orders induced by energy conservation, a stagnant economy and the skyrocketing and unpredictable cost of nuclear power.

This period should not be used simply to apply more Band-Aids to existing designs that seem incapable of meeting safety concerns at a price electric consumers are willing to pay.

The last three nuclear units TVA halted were going to produce electricity estimated to cost about 13 cents per kilowatt-hour, more than triple our present average cost of about 4 cents per kilowatt-hour. Furthermore, the number of changes required appears open- ended, since many safety issues are still unresolved. Indeed, in a sense TVA has not really "completed" any of its nuclear plants. At our Browns Ferry nuclear plant in northern Alabama, which has been commercial for years, we still have more than 600 people at work making backfitting changes.

In short, the cost of a new nuclear plant isn't just high, it's unpredictable. No sane capitalist is going to build something for which he can't derive a cost/benefit ratio because the cost is unknowable. That's why the nuclear industry is in the doldrums.

Therefore, what we should be doing is fundamentally reexamining the nuclear option. The central question is: Is there an inherently better technical option than the apparently "unforgiving" reactor design we have today? I believe the answer may very well be yes.

A lot of people, of course, would say why bother? Conservation can substitute for nuclear power and solar can be expanded in the future.

We at the TVA are second to none in our commitment to conservation and alternative energy sources. Since I joined the TVA in 1977, we have conducted energy conservation surveys of more than 600,000 living units. We've issued low-interest loans for insulation, weatherization, heat pumps, solar, and wood heaters for more than 300,000 living units. We estimate that we've recently passed the 1 billion-kilowatt mark in energy saved each year as a result.

But despite -- or maybe because of -- our extensive experience with these technologies, I am convinced that they are not the whole answer to our energy problems.

For the next few decades, I don't see any realistic way that less nuclear power will mean anything other than more oil and coal. And the marginal dangers of more nuclear plants of a safer design seem to me less of a threat than the added risk and economic ruin from greater reliance on imported oil, not to mention the problems of acid rain and the global threat of carbon monoxide build-up in the atmosphere associated with coal.

So let's acknowledge a few things about our present nuclear reactor designs and move on.

No matter how much we rebuild and retrofit these light-water reactors, the Three Mile Island accident has revealed that the then-existing nuclear technology was capable of self-destructing. The problems that NRC regulations are attempting to correct were real. The doubling and tripling of construction lead-times only reflect the chaotic state of nuclear plant design as the industry scrambles to retrofit safeguards that experience has shown to be necessary.

An improved standardized version of these current light-water reactors would still be overly dependent on the skill of operators. The exposed piping would still be subject to leaks and, consequently, to loss-of-coolant accidents.

While risks to the public can be and are being reduced as a result of all these retrofits and design changes, we cannot be sure that accidents wouldn't result in a melt-down and a destructive release of radioactivity.

Of course, the reactor itself could be crippled, causing the kind of disaster that has, in fact, already cost over $1 billion in immobilized equipment and necessary cleanup. Obviously, that kind of risk is just unacceptably high -- not only for the public, but for the utilities and the financial institutions that provide the capital to build the things.

A recent Swedish modification of the basic light-water reactor design addresses this problem in an imaginative way. The Swedes, paying homage to the American penchant for acronyms, call the concept PIUS (Process Inherent Ultimate Safety). This design puts all of the major components, along with the piping that connects them, inside a single, large, pre-stressed concrete vessel. The long-term cooling for the nuclear fuel is provided in a fashion that is not dependent on switches and pumps. Instead, the cooling comes from natural circulation of a large pool of water contained inside the concrete vessel itself.

This "passive" approach eliminates the need for conventional electrical and mechanical "active" safety systems and the reliance on operators to prevent an accident. It can provide cooling for about one week without external emergency cooling systems, without the use of electricity, and without operator action.

Based on its design concept, the plant should be able to tolerate operator error and multiple failures of almost all the active systems. The American experts who have examined this concept can find no basis for challenging the designers' claim that it is incapable of having a melt-down.

No technology can be absolutely safe. But clearly, a design that has the potential of being melt-down-proof deserves the necessary funding to be pursued in a thorough research and development effort.

However, neither the nuclear industry nor its critics seem interested. The industry has too much invested in the current technology, and the nuclear critics don't want to face up to the risks of expanded use of coal and imported oil in the pre-solar era.

Another alternate reactor concept is called the High Temperature Gas-cooled Reactor (HTGR). It also holds promise of greater safety and efficiency. It is far less likely to melt down because it has far greater ability to withstand and hold heat. This is made possible by the use of helium as the coolant. Unlike water, helium is already a gas. Therefore, it cannot turn into explosive steam.

This reactor is also aided by the use of graphite in the manufacturing of the reactor itself. Graphite has excellent characteristics for withstanding high heat, even if an accident were to cause the heat to head way above normal.

Helium also has fewer impurities than water, which means it doesn't corrode the pipes. And it has a lower tendency to become radioactive. As a result, an HTGR would have lower in-plant radiation levels, making maintenance easier, less hazardous and less expensive.

Finally, this reactor design uses uranium more efficiently than our present reactors do, and it could be designed to use thorium, a very abundant fuel.

The problem with HGTRs is that 10 years ago, when everybody thought light-water reactors would be commercially successful, their development was halted. The Ft. St. Vrain plant in Colorado is the only operating, high-temperature gas-cooled reactor in the country. Granted, like all other pioneering efforts, it encountered problems.

But the biggest problem is that while a number of improved designs exist on paper, there is no program, either from the federal government or from industry, for research, development, and a demonstration plant that could determine the real worth of this obviously promising approach.

The issue is whether we need a much better nuclear product or whether the current designs are adequate.

Is the problem with nuclear power an hysterical public and a weak-kneed Nuclear Regulatory Commission? Or is the problem that we should be looking for alternatives to the very product itself, when we are not?

My own opinion is that the fears of the public -- now that we have seen a nuclear reactor self-destruct -- are very real. They are not going to go away. We have a stalement. The industry is trying in vain to prove that the fears of the public are unfounded. As we say in East Tennessee, "That dog won't hunt."

We ought to realize that with nuclear power, we are still experimenting. We are still developing a very complex technology. We stopped the research and development effort much too soon.

But we can't allow ourselves to be locked into the errors of the past. The nuclear critics -- who say that no nuclear plants can work -- and vested business interests -- who refuse to consider changing a fundamentally flawed product -- may both be wrong. Obviously, therefore, we should not and cannot let either of them totally control our future nuclear policy for us.

The existing technology is just not good enough. But the national interest will be harmed if, in the pre-solar era, we become inextricably tied to imported oil or polluting coal.

Therefore, the nuclear option must not be allowed to die. If the administration is indeed as pro-nuclear as it claims to be, it must make the necessary investments in exploring differents systems and new technical advances.

And if the nuclear industry does not support such an effort, then it -- and not the nuclear critics -- must bear the ultimate responsibility for the death of nuclear power.