Certainly this criterion is best met by reactor concepts using conventional components and systems operating at coolant temperatures and pressures within the envelope of significant operating experience. Water as a coolant for SMRs has been selected explicitly because of the satisfaction of these conditions. Experience with water reactors using the essential design features selected for water-cooled SMRs goes back to the beginnings of the nuclear electricity generation and propulsion age. The major caveat regarding the achievable reliability of water-cooled SMRs relates to those having selected the integral configuration, the placement of all NSSS components and piping within a single pressure vessel. While the Otto Hahn merchant vessel successfully used this reactor configuration and operated commercially for nine years, the potential reduction in operational reliability of this configuration due to its limited accessibility for primary system component monitoring, maintenance and repair can be confidently assessed only through many more years of operating reactor experience.

Sodium-cooled reactors have generally had a mixed, albeit limited, record of operating experience. The US Experimental Breeder Reactor II (EBR-II) and British Dounreay Fast Reactor (DFR) records were exemplary, the Russian BOR-60 and BN-600 and the French Phenix reactor experience was on balance satisfactory, while the Japanese Monju experience has been very troubled, principally due to a sodium leakage event as was the Superphenix experience. Similarly the lead-bismuth — cooled Russian submarine reactors operated reliably but with the need for careful attention to coolant chemistry control and freeze prevention after the major accident in 1968 before adequate understanding existed of the need for rigorous control of coolant oxygen concentration to prevent lead oxide slag formation (Toshinsky and Petrochenko, 2012). Helium-cooled reactors, e. g., the experimental reactors AVR and THTR in Germany and the commercial Fort St. Vrain unit in the US, also have had a mixed operating record.

Hence it can be concluded that, based on operating experience, the water-cooled SMR class has a significant advantage over the other coolant types with regard to its promise of operational reliability. The operational reliability of non-water-cooled reactors will be uncertain until sufficient demonstration plant operational experience is accumulated.

The principal coolant characteristics influencing this operational experience — e. g., coolant toxicity, corrosion effect on bounding surfaces, and coolant freezing and boiling temperatures — are shown in Table 1.4. Coolant toxicity has been expressed in terms of radiological, biological and chemical factors.

Biological consequences arise from decay of 210Bi which yields 210Po. The polonium then chemically combines with lead as PbPo(s). Should water enter the primary system due to a failure of the ingress penetration barrier coincident with a steam generator tube leak, it would react with the PbPo(s) to produce H2Po(g), a volatile alpha-emitting aerosol of biological inhalation concern. The designers of the lead-bismuth-cooled SVBR-100 reactor (see Table 1.2), who are well versed in Russian submarine experience, cite that operating experience has resulted in the development of measures for providing adequate radiation safety. For water-cooled reactors, water chemistry measures typically include introduction of boron and lithium in the form of boric acid and lithium hydroxide for corrosion control, although some SMRs, e. g., the B&W mPower design, have eliminated the use of soluble boron for reactivity control. Neutron activation of 6Li and 10B produces tritium, 3T, albeit in small quantities, which nevertheless is a biological hazard if ingested.

Occupational contact hazards of a chemical nature exist for lead through high levels of exposure due to inhalation and occasionally skin contact. Similarly asphyxiation due to accidental immersion in helium (or in nitrogen typically used to inert BWR containments) is a potential hazard. The more significant, well-recognized chemical oxidation reactions of zirconium cladding and sodium are covered as a safety concern under potential energy release in Section 1.4.1.

Of all the coolants, helium, because it is an inert gas, poses the least corrosion potential, and its activation is minimal as demonstrated by the Fort St. Vrain experience that showed very low activity in the coolant compared to light-water reactors. The aggressive attack of lead and lead-bismuth on metal cladding (e. g., in HT-9 and the Russian equivalents EP 823 and EP 450) has forced the limitation of coolant velocity in lead — and lead-bismuth-cooled core designs to 3 m/s. This in turn has necessitated the provision of a large coolant flow area to bound core coolant temperature rise. Hence lead and lead-bismuth cores have fuel pins spaced with a large pitch/diameter square lattice array. However, recent development (Short and Ballinger, 2012) of a composite material for cladding and structural application may mitigate such limitations.

Finally, the operability of liquid metal coolant systems requires trace heaters around piping and components of sodium, lead and lead-bismuth reactors to prevent coolant freezing when insufficient heat is available from power operation or decay heat. The high freezing temperature of lead, 327 °C, compared to the modest values for sodium, 98 °C, and lead-bismuth, 125 °C, renders lead disadvantageous as a reactor coolant in this regard. However, with these high freezing temperatures both

Table 1.4 Inherent coolant characteristics affecting operational reliability Water1 Helium Sodium2 Lead2 Lead-bismuth2 Radiological 16O(n, p)16N 16N^16O + 5 to 7 MeVr (T1/2 = 7.1 s) None but erosion created dust liftoff from sudden depressurization can cause mechanical clogging 23Na(n, r)24Na (Tm = 15 h) 1.38, 2.76 MeVr s 23Na(n, 2n)22Na (T1/2 = 2.6 yr) 1.28 MeVr 204Pb(n, r)205Pb (T1/2 = 51.5 days) 1.28 MeVr Same as lead plus 209Bi(n, r)210Bi(e) 210Po 210Po (a, r low prob.) 206Pb (T1/2 = 138 days) 5.3 MeV a; 805 keV r Toxicity Biological 6Li(n, a)3T 10B(n, 2a)3T 10B(n, a)7Li(n, na)3T (T1/2 = 12.3 yrs) None None Trace amounts of Po from 205Pb to 210Po by neutron capture and b- decay PbPo(s)+H2O=PbO+H2Po(g) (volatile alpha-emitting aerosol) Chemical None Asphyxiation hazard None Exposure to high levels of lead through inhalation, ingestion or occasionally skin contact can lead to the medical condition known as lead poisoning Same as for lead Corrosion Prevention of stress corrosion cracking of stainless steel requires significant attention. Also significant corrosion- induced crud formation potential None Sodium is practically non­corrosive with respect to stainless steel. Corrosion is lower than for lead or water Aggressive corrosion by: • direct dissolution by a surface reaction • intergranular attack. Oxide film formation tends to inhibit the corrosion rates. Need to limit velocity to about 3 m/s to avoid cladding corrosion. Same as for lead Melting (freezing)/ Boiling points (°C) 0/100 NA 98/883 327/1737 High freezing temp — need trace heating 125/1670 Lower freezing temperature advantageous vs lead 1Lin (1996). 2Todreas et al. (2008).

lead and lead-bismuth eutectic will solidify in ambient air, providing a means for sealing small leaks in the primary coolant boundary. On the other hand, the high boiling points with the attendant low vapor pressures of these liquid metal coolants allow reactor operation at atmospheric pressure without the source of stored energy associated with a high-pressure coolant. Operation at low pressure allows reduction of the required thickness of the pressure vessel and other primary pressure boundary components. Nevertheless, for the heavy lead coolant the dimensioning of these vessels must be carefully evaluated to satisfy seismic design criteria.