Canadian Coalition
for Nuclear

Regroupement pour
la surveillance
du nucléaire

Deterioration of CANDU Feeder Pipes

by Dr. Gordon Edwards
January 1997

[ pour la version française ]

  • The Nature of the Problem

  • Facts About Meltdowns

  • The Importance of Small Pipe Breaks

  • Back to the Feeder Pipes

  • The Nature of the Problem

    In early January, 1997, it was learned that CANDU reactors now operating in Quebec, New Brunswick, Korea, and Argentina -- and presumably those in Ontario, India and Pakistan -- are suffering from an alarming deterioration in the strength and integrity of certain small pipes, called feeder pipes. Each CANDU reactor has hundreds of such feeder pipes.

    The face of a CANDU reactor (under construction).
    Some feeder pipes can be seen to the right of the reactor face.
    ~ photo by Robert Del Tredici ~

    The feeder pipes perform a vital function -- they carry the primary coolant required to cool the nuclear fuel, thereby preventing the core of the reactor from overheating, melting, and releasing a significant fraction of the enormous inventory of radioactive poisons contained therein.

    Due to corrosion, the walls of these small pipes are becoming perilously thin -- in some cases, the thickness has been reduced by as much as 30 percent. Unless the situation is corrected, the plants most affected -- Point Lepreau (New Brunswick), Gentilly-2 (Quebec) and Wolsung (Korea) -- will be forced to shut down for safety reasons within five years.

    If the core of the reactor is the heart of the beast, then the feeder pipes are its arteries, and the coolant is its life blood. In the human body, the rupture of an artery is a major crisis -- for if too much blood is lost, the heart will stop beating. Similarly, in a CANDU reactor, the sudden rupture of one or more feeder pipes constitutes a serious emergency, necessitating the complete shutdown of the reactor and raising the spectre of a possible meltdown.

    Facts About Meltdowns

    If the core of a large nuclear reactor is not adequately cooled, for whatever reason, it will spontaneously melt at a temperature of more than 2700o C   ( or 5000o F ). When this happens, hundreds of radioactive poisons -- the byproducts of the fission reaction -- will escape into the reactor building, and a fraction of these will escape into the environment surrounding the plant. The amount that escapes depends on the integrity of the building, the state of the equipment, the actions of the workers, and the decisions of the managers at the plant. At Chernobyl in the Ukraine, it is estimated that between 3 and 5 percent of the entire radioactive inventory of the crippled reactor has so far escaped into the global environment.

    People unfamiliar with the principles of nuclear fission often presume that once the nuclear reaction is stopped, the danger of fuel melting is past. This, however, is incorrect. Even if the reactor is completely shut down, the danger of fuel melting continues for days and even weeks afterwards. The danger results from the heat generated by the intensely radioactive "fission products" in the used nuclear fuel. This heat -- the so-called "radioactive decay heat"-- cannot be turned off. It is quite sufficient to melt the core of a large power reactor, and it will surely do so unless the decay heat is removed from the core by some kind of cooling fluid.

    Realizing the potential danger to human health and the environment, engineers have provided each reactor with an Emergency Core Cooling System (ECCS) having a large reservoir of ordinary water, designed to cool the fuel in the event of a "Loss of Coolant Accident" (LOCA) caused -- for example -- by a pipe break in the primary cooling system.

    But there's a catch. The ECCS is not always available, even when the plant is operating at full power. Due to mechnical failure or operator error, the ECCS may be partially or totally disabled, and hence unavailable, unbeknownst to the plant operators. It happens year after year; during certain stretches of time, the ECCS is not fully available. The same is true of the other safety systems, such as the containment system, and the emergency shut-down systems. The Atomic Energy Control Board regularly records, measures, and publishes statistics on the unavailability of all safety systems at Canadian nuclear reactors. Consequently, there is a slight but very real chance that one or more of these systems will not be available when needed. In such circumstances, of course, the probability of a core meltdown is much greater than would normally be anticipated.

    The Importance of Small Pipe Breaks

    In 1974, the U.S. Nuclear Regulatory Agency published the Reactor Safety Study, commonly known as the Rasmussen Report (after its principal author, Norman Rasmussen, who was then head of the Nuclear Engineering Department at MIT). This 12-volume report -- the most ambitious study of reactor accidents done to date -- concluded that small pipe breaks in the primary cooling system are the most important single contributor to the overall probability of a complete core meltdown.

    In 1977, I and my colleague Ralph Torrie argued before the Ontario Royal Commission on Electric Power Planning that, because a CANDU reactor has far more small piping (e.g. in the form of pressure tubes and feeder pipes) than any U.S. nuclear reactor, the overall probability of a complete core meltdown is at least as great and probably even greater in a CANDU than in an American "light water" reactor.

    The Royal Commission's 1978 Report entitled "A Race Against Time" concluded:

    "The probability of a core meltdown at Pickering is said to be in the order of one in a million years. . . . However, two well-informed nuclear critics who participated in the hearings, Dr. Gordon Edwards and Ralph Torrie, have argued that the probability . . . could be about 100 times higher than the theoretical levels. This estimate is based on failure rates in the high pressure piping of the primary heat transport system being 10 times higher than has been assumed, and also on the fact that the availability of the Pickering ECCS has been demonstrated to be 10 times lower than postulated by the designers.

    "We believe that the Edwards/Torrie estimate is more realistic than the theoretical probability, not least because the Rasmussen Report has concluded that the probability of an uncontained meltdown in a light water (US) reactor is one in 20,000 years. (It has been suggested, moreover, that this figure could be out by a factor of 5 "either way".) Assuming for the sake of argument, that within the next 40 years Canada will have 100 operating reactors, the probability of a core meltdown might be in the order of one in forty years, if the most pessimistic estimate of probability is assumed."

    [A Race Against Time, 1978, pp.78-79]

    Then, in 1979, the Three Mile Island accident happened, which came close to being a complete core meltdown. President Carter appointed mathematician John Kemeney to head up a special Commission to investigate the causes and implications of the TMI accident. The Kemeney Commission concluded that the TMI accident was essentially a "small pipe-break" type of accident. Even though the accident was caused by a stuck valve rather than a broken pipe, the Commission found that the course of the accident and the consequences were much the same as they would have been had a small pipe in the primary cooling circuit actually broken.

    The Kemeney Commission also concluded that a major contributor to the TMI accident was a certain attitude on the part of the owners and operators of the plant, who had assumed that nuclear power is an inherently safe technology rather than an inherently dangerous one. The Commission warned that such an attitude, if not corrected, will inevitably lead to more reactor accidents.

    Back to the Feeder Pipes

    It is in this light that one should weigh the significance of the recent findings about the thinning of the walls of the feeder pipes in operating CANDU reactors. As the walls get thinner, and the pipes get older, the probability of a small pipe break in an operating CANDU reactor increases dramatically. Although such a pipe break need not result in any fuel melting, there is always the possibility that a core meltdown could result. If the Ontario Royal Commission is right in its probability estimates, the chance of a core meltdown in a single CANDU during its 30-year lifetime could be about 1 in 66 -- even without the thinning of the walls of the feeder pipes. That's a whole lot larger than the probability of winning anything worthwhile in a lottery.

    Of course, the feeder pipes can be replaced -- but at what cost? In 1983, a pressure tube suddenly burst inside one of the Pickering nuclear reactors. In fact, it was a small pipe break accident, which -- luckily -- led to no serious radioactive releases. However, the subsequent repair work -- replacing all the pressure tubes in units 1 and 2 -- required a four year shutdown of two nuclear reactors and a repair effort costing $700 million. By the 1990's, everyone began to learn of Ontario Hydro's staggering financial problems, caused almost entirely by its nuclear reactor program. It turns out to be very expensive to keep CANDU reactors operating safely; as a result, four of them have already been retired, and the others will need repairs costing more than a billion dollars.

    In view of all this, it is disheartening to realize that our federal government, which has given the Canadian nuclear industry over $13 billion in federal subsidies to date, has never subjected its nuclear policies to parliamentary debate or public inquiry. Jean Chrétien and his cabinet continue to subsidize and to promote the Canadian nuclear industry without a clear political mandate to do so, and without providing any transparent democratic process by which critics can challenge the industry to give a full accounting of the benefits, costs and risks of this perilous technology.

    And, from an ethical perspective, should we not ask why we are selling these reactors to other countries, when we ourselves cannot afford to maintain them in a safe operating condition?

    [ Accident Possibilities at Gentilly-2 ]

    [ Findings on CANDU Safety ]

    [ Reactor Accidents Sub-Directory ] [ COMPLETE DIRECTORY ]

    Since March 27th 1996, there have been over
    100,000 outside visitors to the CCNR web site, plus

    (counter reset July 2nd 1998 at midnight)