Though esoteric control engineering theory might be outside some reader’s interest, a brief overview of Nyquist’s [131] and Rosenbrock’s [123,134] theorems is necessary in order to appreciate the practical problems described in Chapters 2 and 3. Here the stabilities of neutron reactor kinetics, flow in boiling channels and a national electricity Grid are analyzed. At first sight it appears surprising that a national Grid stability criterion [80,84] can be formulated as a single input-single output problem when many plant controls and a myriad of ac generators and motors are intimately involved. This example particularly illus­trates the important engineering skill of identifying the reduced set of variables that dominate a complex physical system in order to effect a successful solution.

Subsequent chapters concern some research activities and opera­tional legislation that aim to underwrite the safety of nuclear power plant following the Three Mile Island incident [66] in 1979. In this respect international collaboration has been sponsored by individual governments, the OECD, the European Union[23] and the PWR Oper­ators Club. The success of these initiatives is confirmed by the absence of any later major technological or operational faults[24] with BWRs and PWRs.

Chapter 4 cites some European statutory probabilities [59,65] for the occurrence of Design Base and Severe Accidents (fuel melting) along with operational requirements [59,108] relating to the progres­sion of plant damage and the need for an operator command structure based on professional skills and training. Hazards and Risks from some pertinent fission products after an unlikely environmental release are described, together with the legislated exposure limits (Sv) for on-site operators and the neighboring public. Though Event Trees and

Risk analyses are increasingly deployed for safety assessments of many industrial processes, weaknesses in their application are identified as failures to address benefits and humanity’s greater acceptance of one manner of death from another. Nevertheless with no real alternative they are still used for assessing nuclear power plant, whose principal Risk to people is the development of thyroid cancers due to the natural concentration of absorbed radioiodides [163] after an unlikely environ­mental release of fission products. In this context, it should be noted that 80 to 90% of naturally presenting cases are successfully treated by surgery [164]. The Farmer-Beattie analysis, which is repeated here with the more appropriate Poissonian rather than a Normal distribution, demonstrates that the expected annual incidence of thyroid cancers from the spectrum of Severe Accidents in an AGR station is some two orders of magnitude less than the number of natural presentations. Granted rationality, nuclear power with its attendant benefits should therefore have gained full public approval! Accident precursors and safety systems for fast reactors and PWRs are briefly outlined. How­ever, because the latter appears to be the most widely adopted type of future civilian plant, the robustness and diversity of their safety systems is illustrated for a large loss of coolant accident (LLOCA).

There are rare circumstances when a coherent explosive rate of heat transfer occurs as a result of a high-temperature liquid mixing with a readily vaporized one. Though unimaginably powerful natural explo­sions took place at Krakatau and Santorini, the situation of Severe Accidents in PWRs with molten core debris (corium) is radically different. Specifically, the natural events involved a rapid violent mixing of Gigatonne quantities, whereas the spatial neutronics and hydraulics of water reactors allow only the progressive melting of their 100 tonne cores over hours [59,65]: and then only if all the diverse safety systems were to fail concomitantly. In nuclear safety assessments such an explosive heat transfer is called a molten fuel-coolant interac­tion (MFCI). The Three Mile Island incident in 1979 invigorated global research into the progression of core damage [59,65,213,269,270] and the physical phenomena in Severe Accidents. In this respect, the bounding Hicks-Menzies analysis of 1965 had raised concern that not unrealistic masses of corium in an MFCI could challenge the integrity of a fast or water reactor’s containing vessel. As well as releasing radioactive fission products, its rupture could create rapidly accelerating metal fragments (missiles) which could potentially breach the surrounding reinforced concrete containment building, and thereby allow an environmental release of radioactivity. However, a simulation developed by the author during 1989-92 confirmed the 4 to 5% conversion efficiency of heat into mechanical work that had been observed in many independent kilogram-sized MFCI experiments. This efficiency is some six times smaller than an isentropic Hicks — Menzies value which disregards the highly effective anisentropic heat transfer from an MFCI bubble into the bulk coolant by interfacial condensation. Identification of this physical process allows a valid extrapolation of the 4 to 5% experimental-scale value to tonne-sized reactor quantities thereby materially benefiting reactor safety assess­ments. Novel finite element models for the impact of plant missiles or aircraft on reinforced concrete structures or major pipe work were also validated during this same period. Following the foreclosure of the EFR project, European research on MFCI and impacts was discontinued. The pertinent UK reports were then archived leaving a much smaller staff complement to pursue successful reactor decommissioning, waste glassification and passive safety systems. Chapters 5 and 6 respectively outline these MFCI and impact research archives to assist engineers and scientists newly entering the resurgent nuclear industries.

After the Three Mile Island incident the nuclear power industry sought unremittingly to improve in-depth plant safety [298,302]. For example, more robust fuel cladding [300] and steel [276,277] or concrete containments [286] have been developed. In addition cooling circuits deploying natural circulation have been proposed [108] as potentially more reliable and cost-effective by eliminating the need for active power supplies and pumps. However, due to intrinsically smaller heat transfer rates than with forced convection, natural circulation cooling systems are relatively much larger. As about 60 to 70% of capital costs reside in civil engineering works, they do not appear economically viable for the main on-load cooling of water reactors of order 1 GWe. Nevertheless, because the radioactive decay of fission products peaks at around 10% of pre-trip (scram) power, passive safety circuits exploiting heat removal by natural circulation are cost-effective [108] and their development has become an ongoing activity [109] associated with proposed Generation IV plants [109,298,302]. Chapter 7 concludes the book with a discussion of further advantages and disadvantages of heat removal by natural convection and a passivity classification for reactor safety systems. All proposed passive safety systems address the problems of decay heat removal to ensure core debris-bed cooling [65], pressure relief inside a concrete containment [101] and the blocking of emergency core cooling systems (ECCS) by debris [100]. Because the driving forces from differential densities and gravity are relatively much weaker than with forced convection, careful design and validated analyses are necessary to be sure that these passive systems function as intended. Indeed the IAEA falls short of recom­mending them as direct replacements for the active safety systems in presently operational plants [10].

A final thought to ponder is that the human and environmental consequences of the Three Mile Island, Chernobyl and Fukushima nuclear incidents are all together dwarfed by the 171,000 deaths caused by the failure of the Banqiao hydro-dam [11]. It is hoped that Chapter 1 will contribute to an objective quantified debate on future electrical energy generation which encompasses national and global issues of Grid demand patterns and land resources [324].