The general features that make a particular fluid (gas or liquid) attractive as a re­actor coolant are as follows.

1. High specific heat. Suppose we have a nuclear reactor that is generating heat at a rate of Q watts. Coolant at a flow rate W (kilograms per second) is passed to the reactor, entering the core at temperature 7 and leaving the core at temperature Tut — From the first law of thermodynamics (see Section 1.1.1), these quantities are related by the equation Q = WCp (Tout — !Tn),where Cp is the specific heat or specific heat capacity of the fluid. The specific heat is the amount of heat required to heat 1 kg of a substance by 1 K (1°C) and thus has the units joules per kilogram per kelvin. In designing reactors it is important to prevent excessive temperatures within the core, in order to avoid damaging the fuel and the core construction materials. The above equation indicates that this can be accomplished in two ways for a given inlet temperature of the coolant. First, the flow rate Wcan be so high that the outlet temperature is not too much higher than the inlet temperature, irre­spective of the value of Cp. Second, a fluid can be chosen that has a high value of Cp, which will also limit the outlet temperature. Of course, the out­let temperatures cannot be too low, or the reactor will not be able to gener­ate steam efficiently, as explained in Chapter 1. Also, with high flow rates

significant amounts of power are needed to pump the coolant, and this is power that is not available as electricity to the customer.

A special case is that in which the coolant is in the form of a boiling liquid. Here, heat can be absorbed by the coolant at its boiling point with no change in temperature and can be used to convert the liquid into vapor. The amount of heat required to convert one unit mass of liquid to vapor is called the latent beat of vaporization (joules per kilogram). The boiling-fluid coolant is often also used as the working fluid in the turbine (e. g., steam generated from a boiling-water coolant in a reactor is used in a steam tur­bine). For the reasons discussed in Chapter 1, the higher the boiling point of the fluid the higher the thermodynamic efficiency. Since boiling point in­creases with pressure, the boiling-coolant system should be operated at the highest practicable pressure. However, the higher the pressure, the more ex­pensive the system, and there is a trade-off between increased capital cost and increased thermodynamic efficiency.

2. High rates of heat transfer. The rate at which heat can be transferred from the fuel elements to the coolant is determined by a number of factors, which are discussed in more detail in Section 33. One of the parameters is the thermal conductivity of the fluid, which is the constant of proportionality between the rate at which heat is transferred through a static volume of fluid and the temperature gradient, i. e., the rate at which temperature is changing per unit length. Liquid metal coolants have high thermal conductivity, whereas gaseous coolants have relatively low thermal conductivity.

3. Good nuclear properties. For all reactors, it is important that the coolants should have low neutron absorption. As explained in Chapter 2, any neutron absorption by the coolant and structure reduces the number of neutrons available for the fission reaction. The neutrons should not react appreciably with the coolant to form radioactive isotopes. Excess radioactivity in the cir­culating system increases operational difficulties, as mentioned in Chapter 2. If the coolant is also acting as the moderator, good moderation properties are required (the processes of moderation were explained in Chapter 1). In fast reactors, of course, it is important that the coolant not moderate the neu­trons, since unmoderated (fast) neutrons are required in the reaction.

4. Well-defined phase state. It is preferable for the coolant to have the same phase state (i. e., liquids remain as liquids and gases remain as gases) during both normal and accident conditions. To achieve this in the case of liquids, a high boiling point is desirable to avoid changes of phase if the liquid is over­heated. A high boiling point also has the advantage of minimizing the pres­sure required to operate at a certain temperature level and of achieving high thermodynamic efficiency.

5. Cost and availability. Since the inventory of coolant in typical reactor sys­tems is quite high (hundreds of tons), it is important that the cost be mini­mized. Also, coolants may leak from reactor circuits, and this can be a significant cost in some cases. The ideal coolant should also be freely avail­able in a sufficiently pure form for use in the reactor circuit.

6. Compatibility. It is obviously axiomatic that the coolant should be compati­ble with the reactor circuit and not corrode it, even under the conditions of high radiation flux that occur in the core.

7. Ease of pumping. Fluids of low viscosity require much less pumping power to circulate them around the reactor circuit than do fluids of high viscosity. The viscosity of a fluid is related to its temperature, that of liquids decreasing with increasing temperature and that of gases increasing with increasing temperature. The viscosity of a fluid is indicated by the symbol Jl.

No practical fluid meets all of these requirements. All known coolants have one or more disadvantages. The thermodynamic and heat transfer characteris­tics of a coolant can be compared conveniently by using a parameter called the figure of merit, which derives from the heat transfer processes and the associ­ated pumping power required. The figure of merit F is defined as

where Cp is the specific heat, Q the fluid density (kilograms per cubic meter), and Jl the viscosity. The rather peculiar-looking powers appearing in this equa­tion result from the empirical correlations used to predict the pumping power and the heat transfer rates.

There are relatively few practical choices for reactor coolants. The ones mainly used are listed in Table 3.1, which shows their density, viscosity, specific heat, thermal conductivity, and figure of merit value. In terms of figure of merit, ordinary water is outstanding. However, it has three main disadvantages: its low boiling point, which requires operation at high pressure in order to reach even moderate thermodynamic efficiencies; its neutron absorption; and its corrosion properties. The latter two disadvantages require enrichment of the fuel and spe­cial containment materials, respectively.

Some Examples

5.1 INTRODUCTION

Incidents at nuclear power stations create a great deal of public interest and sometimes concern and alarm.

Many incidents occurred in the 50-year period up to 1995, though very few resulted in injury or death to plant operators or the general public. Mosey 0990) lists 60 or so separate events, and even this list is probably not compre­hensive. The three most serious events are the accidents at Windscale 0957), Three Mile Island 0979), and Chernobyl (1986). Brief details of each are in­cluded in this chapter.

Designers of nuclear power stations do assess the risks and consequences beyond the basis adopted for design. However, prior to Chernobyl, the actual release of radioactive fission products from nuclear accidents had been very much less than predicted from such analyses, indicating their general conser­vatism. Accidents can be examined by looking at which of the three basic safety principles, the Three Cs—control the reaction, cool the fuel, and contain the ra­dioactivity, has been breached and to what extent the overall defense in depth has been challenged.

If the world is to benefit from nuclear energy in the longer term despite the potential dangers involved, it is essential that the lessons learned from each ac­cident or incident are incorporated into future designs and into operator train­ing and safety management to make existing stations safer. It is beyond the scope of this book to examine all these incidents. Rather, we will select those examples which illustrate specific points we have highlighted in previous chap-

ters. The examples are chosen to illustrate both type of fault and type of reac­tor, as follows:

Light water-cooled reactors SL-1

Millstone 1

Browns Ferry 1 and 2

1bree Mile Island-2

Ginna

Mihama-2

Chernobyl

Heary water-cooled reactors NRX Lucens

Early gas-cooled reactors (the air-cooled piles such as those at Windscale in the United Kingdom) had horizontal channels, and the fuel elements were simply pushed in at one end and spent fuel was removed at the other. With the intro­duction of Magnox reactors, which had vertical channels and used a pressurized carbon dioxide coolant, this simple system was no longer adequate. The refuel­ing arrangements used for a Magnox reactor are illustrated in Figure 7.1. An array of vertical pipe comes from the top of the reactor vessel as illustrated (these are called standpipes). The refueling machine may be connected to any of the stand­pipes. This machine is shown in Figure 7.1 and is basically a pressure vessel that can be moved across the top face of the reactor. It is provided with adequate ra­diation shielding and therefore tends to be heavy. When the refueling machine is connected to one of the standpipes, a plug is removed from the top of the stand­pipe, allowing the high-pressure carbon dioxide coolant to enter the refueling machine vessel; thus, the vessel becomes an extension of the primary circuit of the reactor. Each standpipe serves a group of fuel channels. The fuel elements are lifted out of the channels using a grab, which is aligned above the particular chan­nel using a special mechanism called a pantograph (Figure 7.1).

SERVICE STORAGE TUBE LATCHES

SERVICE STORAGE TUBE FOR: STANDPIPE PLUG ASSEMBLIES CHUTIS CONTROL RODS

IRRADIATED FUEL STORAGE TUBE

NEW FUEL STORAGE TUBE GANTRY

IRRADIATED FUEL ELEMENT

REACTOR CORE

Figure 7.1: Magnox refueling machine.

Alternatively, in some cases, an aligning chute is used. In a typical refueling operation, all of the individual fuel elements in a fuel channel are removed and stored temporarily in magazines in the refueling machine vessel. New fuel ele­ments, already present in the vessel, are then inserted using the same mecha­nism. In the Magnox reactor, no special cooling is provided for the spent fuel elements within the refueling machine since natural convection of gas around the elements keeps them cool enough.

In the AGR the arrangement is similar to that used in the Magnox reactor, ex­cept that there is a standpipe for every fuel channel as illustrated in Figure 7.2. Thus, the complete fuel from the channel can be drawn up into the refueling machine as a single entity, and the complex pantograph or chute mechanism is avoided. In the AGR the fuel rating is much higher, and thus the decay heat re­lease rate is such that natural convection cooling of the spent fuel within the re­fueling machine may be insufficient. During the refueling operation, therefore, carbon dioxide from the reactor circuit is passed through the refueling machine and over the spent fuel. The fuel damage incident at Hinkley Point B, described in Section 5.4.4, led to increased attention to cooling during the refueling oper­ation and to the installation of backup emergency cooling systems.

In both Magnox and AGR reactors, the refueling machine containing the spent fuel is trundled over to a discharge point where the magazines are emp­tied into an irradiated fuel buffer store that is gas-cooled. Subsequently, they may be transferred (also using the refueling machine) to a more permanent storage at the reactor (normally a deep pool of water) before being finally trans­ported from the site. The sequence for an AGR is illustrated in Figure 7.3.

A gas-cooled reactor that we have not previously mentioned and that has a novel method of on-load refueling is the pebble-bed reactor developed in the Federal Republic of Germany. In this reactor the fuel is incorporated into graphite spheres that are charged into the top of the reactor, the spent fuel being discharged at the bottom. A small prototype of this form of reactor was operated for a considerable time.

This Second Edition has been several years in the making. My life-long friend and colleague John Collier died from pancreatic cancer on November 18, 1995. This Second Edition must sadly but proudly serve as a memorial to John and to his intense and firm conviction of the need for nuclear power for the future well-being of the human race on this planet. John Collier’s transparent honesty and humanity provided the best possible witness to the sincerity of this convic­tion. I, too, strongly believe in the ultimate necessity for nuclear power; there will be temporary situations where this need is not so obvious (for instance, the current availability of an excess of natural gas in the United Kingdom), but the long-term situation is clear. It is thus vital to continue research and development in the area and to maintain an adequate technology base. Everything possible must be done to develop public confidence in nuclear power, and the industry should not be averse to considering new concepts which spring from the lessons regarding inherent safety learned in the chemical industry. The main public concern is with the possibility of severe accidents, and the accidents at Three-Mile Island and Chernobyl have naturally served to fuel this fear. The nu­clear industty must recognize this problem of public acceptability and face up to it. Once the long term need for nuclear power is recognised and accepted, solutions can and indeed must be found. However, it is worth pointing out that of all modern industrial plant, even the present generation of nuclear power stations is among the safest. In a properly regulated environment, the present operating nuclear power stations provide a safe and economic means of energy production. However, the nuclear industry needs to give a lot more thought to the sources and consequences of major accidents if, as it seems inevitable to me, nuclear power generation will need to be expanded to meet the growing

energy demands. It is with this as a background that a large amount of the ma­terial in this book is concerned with nuclear accidents and their consequences.

For this Second Edition, the material has been extensively updated and re­vised. In the months before his death, John Collier carried out much of the work in preparation for this, and I would like to place on record my apprecia­tion of his contribution. Perhaps the most important new material is that associ­ated with the Chernobyl accident. This accident happened on April 28, 1986, at a time when the proofs of the First Edition had been produced. A short section was written in the First Edition about the accident but, of course, a full realiza­tion of the sources and consequences of the event was not at that stage possi­ble. We have attempted to rectify this in the current volume. We have also updated the section on the Three Mile Island accident to reflect the continuing developments in understanding and analysis of that event.

Other major modifications in the current volume, with respect to the First Edition, include an updating of the material on Earth’s internal heat generation in Chapter 1, major updating and revision of the general material on severe ac­cidents, and an updating of the material relating to fusion power generation.

I hope that this new edition will be a helpful update for those who pur­chased and used the First Edition and that it will serve to introduce a new gen­eration of readers to nuclear power and its enormous future potential.

G. F. Hewitt, 1996

There are a number of advantages in cooling a reactor core with a coolant that

vaporizes (boils) in the core itself.

1. The vapor produced can be fed directly to a turbine, and power can be gen­erated without an intermediate heat exchanger and/or vapor generator.

2. Boiling coolants are very efficient in heat transfer (see Section 3.3).

3. The evaporation process in the reactor core produces a mixture of vapor and liquid, which has a much lower neutron absorption than a liquid and at the same time maintains a very high heat transfer efficiency. As the proportion by volume of vapor in the coolant (commonly called the void fraction) in­creases, the neutron absorption decreases and there is an increase in the re­actor neutron population, or the reacitivity. If the coolant also acts as a moderator, the neutron population will decrease. Thus, reactors with boiling coolants that also serve as the moderator commonly have a decrease in neu­tron population with increasing void fraction, or a negative void coefficient. If the demand for steam from the reactor increases, therefore, the natural ten­dency of the reactor is to start to shut itself down, and the control system must be designed to accommodate this effect. In reactors of the pressure — tube type with separate moderators (e. g., graphite), there can be a positive void coefficient and the reactivity increases unless action is taken to offset the effect. It is noteworthy that when sodium boils in a fast reactor, where there is no moderator, an increase in reactivity is observed since there is a positive void coefficient in this case also.

The main disadvantages of boiling coolants are as follows:

1. The highly efficient boiling process can degenerate into an inefficient, essen­tially vapor-cooling process rather abruptly due to the phenomenon of dry­out or burnout, as described in Section 3.3.

2. Using vapor generated directly in the reactor core in the power generation system means that the latter system is somewhat radioactive, requires special design, and has increased maintenance and operating costs.

3. The rather complex behavior associated with the void coefficients, as de­scribed above, can also be a disadvantage.

Liquid-cooled reactors can inadvertently become boiling-liquid-cooled reac­tors in the event of a power excursion or a loss-of-coolant accident. We shall discuss this in detail in Chapter 4.

This incident occurred soon after the initial commissioning of the advanced gas — cooled reactors at Hunterston in Scotland. On October 2, 1977, the B2 reactor was shut down for modifications to the plant. On October 11, the carbon dio:x — ide gas pressure was being reduced when alarms, instruments reading, and gas samples began to show excessive moisture in the reactor coolant gas. Subse­quently, it was discovered that about 8000 liters of seawater had entered the re­actor vessel. Damage to the insulation in the annulus below the boilers was extensive. It had to be completely replaced and the reactor was out of service for about 28 months. The repair work cost £13 million (Gray et al., 1981).

At first it seems incredible that a large amount of seawater could enter the pressure vessel of a gas-cooled reactor. The circumstances were these. Figure 5.24 shows the gas circulator cooling system. During initial commissioning of the re­actor in April 1977, the demineralized water in the cooling circuit for the seals on one of the circulators was found to be acidic due to the presence of carbon diox­ide. Carbon dioxide was entering the cooling water through a crack in a seal weld. In order to allow the reactor to run until its planned shutdown in October, it was decided to continue the commissioning phase of the operation and run the acidic water to waste via a temporary connection to the reactor seawater cooling system, thereby avoiding corrosion of the circulator cooling system.

When the gas pressure was reduced below the seawater cooling system pres­sure, a flow path for the seawater was established. This would not have hap­pened if the isolating valves in the temporary drain connection, which had earlier been logged as shut, had in fact been shut. Actually, they were partly open.

This incident points to the dangers of temporary modifications made without

full analysis of all the implications and to the importance of positive indication of valve positions.

Salt deposits are attractive sites for long-term disposal of radioactive waste. The fact that salt is present in the solid form in a geological stratum indicates that it has been free from circulating groundwater since its formation several hundred million years ago. Thus, fuel placed in such a deposit would be free from the leaching action of the groundwater. Salt deposits of this type are quite common, particularly in the United States, and Figure 8.6 shows a conceptual scheme for ultimate disposal of radioactive waste in a salt stratum. Typically, a PWR fuel el­ement may be generating 500 watts of decay heat after 10 years, and this heat generation declines with a half-life of about 30 years since the heat release is dominated by the strontium and caesium decays mentioned above (see Figure 8.3). Thus, after 30 years, the heat release would be down to about 250 watts, and after 60 years it would be reduced to about 120 watts. At these levels, con­duction to the surrounding salt strata is sufficient to remove the heat while maintaining the outside surface of the containment canister to a temperature no higher than 1 00-150°C.

Figure 2.3 illustrates schematically the principal components of a typical nuclear fission reactor.

In this typical reactor the coolant (high-pressure C02 in the AGR case chosen for illustration) at high pressure is driven by the coolant circulator over the fuel element. In many reactors (including the AGR case illustrated) this consists of pellets of uranium in oxide form sealed in a can made of stainless steel. The can (or cladding) ensures retention of the fission products so they cannot enter the coolant stream. It also prevents the coolant from attacking the fuel, which would be possible with some combinations.

The fuel elements are embodied in a structure (the reactor core) that allows them to be surrounded by the moderator. In the AGR case, the fuel assemblies are stacked in vertical holes (channels) in the massive strncture of the graphite moderator. The whole is contained in a prestressed concrete pressure vessel re­taining the high-pressure carbon dioxide gas.

The coolant extracts the heat from the fuel elements. In many reactors, this heat is then used in a boiler or steam generator to convert water to steam. In the boiling-water reactor (B^WR the steam is generated directly in the reactor core. The steam is then passed through the turbine that drives the electrical genera­tor. The very low pressure exhaust steam from the turbine is passed to a con­denser where it is converted back into water and recirculated to the steam generator (or to the reactor in the case of the B^WR

As we saw in Chapter 1, the moderator may be a solid (e. g., graphite) or a liquid (e. g., heavy water). In light-water reactors, the coolant and moderator are both ordinary water. If the moderator is different from the coolant, it must either not react with the coolant or be separated from the coolant by a suitably inter­vening structure. In the heavy-water reactor, this structure is known as the ca — landria; it consists of a tank containing the heavy water penetrated by a series of tubes in which the fuel is mounted and through which the coolant passes.

The remaining main feature of the nuclear reactor core is the means of con­trolling neutron population, namely, the control rods. These consist of neutron­absorbing material such as boron or cadmium.

The number of neutrons produced per neutron absorbed is often referred to as the multiplication factor k. If k is the greater than unity, the neutron population increases; if k is exactly unity, the neutron population remains the same; and if k is less than unity, the population decreases. The rate of growth of the neutron

population depends on the neutron lifetime, i. e., that time between the creation of a neutron and its interaction with the fissile material to create further neutrons.

Most of the neutrons present in the reactor are the so-called prompt neu­trons. In thermal reactors they have a lifetime of typically 0.0001 to 0.001 s; in fast reactors their lifetime is even shorter. If the neutron population consisted of only these neutrons, it would grow very rapidly as soon as k slightly exceeded unity, and the reactor would be very difficult to control. This is because the time between successive generations is very short, and very rapid multiplication of the neutrons would be inevitable. For instance, for a neutron lifetime of only 0.005 s, the neutron population would increase (for k = 1.005) by over 20 times in 1/3 s, and this growth clearly could not be controlled easily.

Fortunately, at the steady state not all of the neutrons are of the prompt type; a small fraction (-0.7%) are of the delayed type, whose lifetime (as defined above) is typically 0.6 to 80 s. These delayed neutrons arise from the decay of fission products rather than directly from the fission process itself. Thus, at steady state only 993°% of the neutrons are of the prompt type and the popula­tion is “topped up” by delayed neutrons, whose number is just sufficient to maintain the steady state, i. e., k = 1.000.

The control system operates essentially on these delayed neutrons, and the response of the system is such that control rod movements over a time scale of 10-20 s can give adequate control over the chain reaction.

The system is designed so that the k value cannot exceed a critical value (1.007 for the example cited above) above which the k value for the prompt neutrons alone is greater than unity. If k were allowed to exceed this value, rapid growth of the prompt neutron population would occur and the system would be in what is known as the prompt critical condition. However, the de­sign of nuclear reactors is such that this condition is avoided.

The nuclear fission process results in intense radiation. The fission products also contribute substantially to the radiation field, and they continue to emit ra­diation after the fission reaction is closed down. Thus, it is very important to provide proper shielding around the reactor core. This shielding takes the form of a thick concrete biological shield. In the AGR plant illustrated in Figure 2.3 the prestressed concrete pressure vessel doubles as the biological shield.

Where necessary—as it is for water-cooled reactors—further protection is provided by housing the whole system inside a leak-tight containment building. We shall discuss the role of this containment building in possible nuclear reac­tor accidents in Chapters 5 and 6.

Figure 2.3 gives a generalized view of the components of one type of nu­clear reactor, and it should be realized that there are many possible permuta­tions of fuel type, coolant type, cladding, moderator, and steam generator. It would be tedious to describe every nuclear reactor type that has been built and practically impossible in any book of reasonable size to describe all those that have been conceived. Many of the early concepts for nuclear reactors departed from the format shown in Figure 2.3 in that they proposed to use the fuel in a fluid form, circulate it through the core, and pass it through heat exchangers ex­ternally before returning it to the core. The concepts included systems in which solutions of uranium salts were circulated through the core, slurries of fuel were made and circulated, or the fuel was circulated in fused-salt form or in solution in liquid metals. There was a tradition at Harwell[1] that in the early days it was possible to invent a reactor system in the bath in the morning and have a pro­ject by lunchtime. It took some years to realize that reactors that you have just thought of are simple, cheap, and reliable, whereas those you are actually working on are always complicated, expensive, and troublesome.

In the remainder of this chapter we shall concentrate on describing some of the main systems that have been implemented in practice and that form the basis of the development of nuclear power. These are the British Magnox and AGR (advanced gas-cooled reactors), the U. S. light-water reactors (BWR and PWR), the Canadian CANDU reactor, the Russian boiling-water graphite-moder­ated RBMK-type reactor, and the liquid-metal fast reactor.

A typical P’^TC. generating about 1100 MW(e) of electrical power would have a decay heat generation of about 200 MW(t) immediately after shutdown. This compares with 3400 ^MW(t) thermal energy generation immediately before shut­down. Removal of this decay heat is well within the capability of the low-pressure cooling system illustrated in Figure 4.4 provided the reactor can be depressurized rapidly enough to bring these into operation. Alternatively, if the steam generators can be operated effectively with the auxiliary feedwater system that is automati­cally switched on when the reactor trips, the decay heat can be removed via the steam generators, even at high pressure. A major difficulty arises when neither of these systems can be brought into play for reasons that will be described in Chap­ter 5. This is what happened in the Three Mile Island accident.

If the low-pressure cooling system and the steam generators are unavailable as a cooling mechanism, the only recourse is to feed water into the system via the high-pressure injection systems and the charging pumps (the pumps used to maintain the inventory of the system under normal operating conditions), the injected water bleeding out through the break. It is interesting to consider how the exiting fluid carries energy with it. The system is illustrated schematically in

Figure 4.11. If the water fed to the reactor is evaporated and exits as steam, this represents the maximum rate of energy release possible. If, on the other hand, the fluid leaves the reactor in the form of liquid water, not only is the discharge rate high (reducing the coolant inventory in the system) but the energy con­tained in the discharge is low relative to that in steam at the same mass flow rate. For these reasons, it is preferable to discharge steam rather than water. Dis­charges in the upper part of the circuit usually contain more energy than those in the lower part, where the existence of liquid water is more likely under tran­sient accident conditions.

Taking the case of steam ejection from the reactor circuit, one can estimate that the maximum rate of ejection corresponds to the release of 17,000 of energy per square meter. To eject the steam that could be generated by the decay heat just after shutdown, a hole of area 0.011 m2 would be required, corresponding to a hole diameter of 12 cm. The hole size required to reject the decay heat as a function of time from reactor shutdown (taking into account the decrease in decay heat rate as a function of time; see Table 2.2) is shown in Figure 4.12. One hour after shutdown the required hole size has dropped to 3.8 em.

If the actual break size is bigger than that required to release the energy in the form of steam, the energy lost will be greater than that being generated and this will result in depressurization of the primary circuit. Such a depressurization may quickly lead to actuation of the low-pressure emergency heat removal sys­tems. However, if the break size is smaller than that required to remove the en­ergy, then energy will be stored within the reactor system, leading to

Bleed and feed

Mass lost as steam
GOOD

low mass flow rate
high enthalpy change

Mass lost as water
BAD

high mass flow rate
low enthalpy change

pressurization of the primary coolant. The system may be partly controllable if the power-operated relief valves (PORVs) can be opened to increase the escape of steam and facilitate energy release. The PORVs are located on top of the pressurizer, and a typical PWR would have two such valves with a total flow area of about 0.002 m2, giving an energy release capacity as steam of about 34 ^W. This is clearly much lower than the 200 ^W of energy release corre­sponding to the decay heat immediately after reactor shutdown. In fact, it might be advisable to consider increasing the size and/or number of PORVs in future reactor designs to allow a higher rate of energy release.

If a break occurs and the available PORV area is insufficient to allow the en­ergy release, the reactor system will continue to pressurize, ultimately actuating the spring-loaded safety valves, whose total area is likely to be sufficient to allow the energy release. However, the latter form of release is somewhat un­controlled. The valves actuate and reseat at a specific pressure.

In the previous section reference is made to a number of specific phenom­ena that can directly influence the course of a severe accident. Rather than in­terrupt the discussion in that section of the effectiveness of the three containment barriers to cover these phenomena in detail, it was more conve­nient to deal with these topics in a separate section. Thus, this section covers

• fuel-coolant interactions—“steam explosions”

• debris beds and their cooling

• hydrogen formation—burning and explosions

• containment basemat melt-through and failure