In magnox reactors the fuel element temperatures are particularly important. Based on the results of fault studies, Station Operating Rules are formulated. These prescribe that a temperature assessment is made at regular intervals, typically once per shift, and after significant changes in reactor conditions.

Fuel elements which consist of a natural uranium rod encased in a magnesium alloy (magnox) sheath (can) have strict limits put on their operating tem­perature. These limits, together with the methods to be used to calculate the highest possible can tempera­ture are defined in the Operating Rules. In the case of the Berkeley magnox reactors this calculated value is then used to define the maximum operating tem­perature of six specially selected No 10 fuel element cans, which are displayed on a reactor desk recorder. A selection of fuel element No 10 (FElOs) which are suitably positioned in the reactor core are used for reactor symmetric fault protection as part of the safe­ty lines system. The thermocouples operate trip units which are designed to trip at preset values and also have ‘low margin1 to trip alarms. The units are fail­safe by design and would trip on excessive rate of change of temperature as would be apparent on open — circuit or supply failure. A description of the unit is given in Chapter 2. Protection using these trip units is on a double ‘2 out of 3’ system, so loss of any individual unit would not cause any reduction in the level of protection. Fuel temperature measurement is in the main only monitored by selected FElOs, but to assess overall reactor fuel temperatures a number of channels have been chosen to receive fuel elements with thermocouples at other positions. These tempera­tures are obtainable via the low speed temperature scanner (LSS) data logger as are some other FE10 temperatures; the remaining FElOs being available on the high speed temperature scanner (HSS) data logger.

The results of the temperature assessment are also used to determine the settings of the temperature trip units that feed into the reactor protection system.

At Hinkley Point A there are six thermocouples per reactor measuring FE5 temperatures: three are connected into the safety lines and the other three act as spares.

At Hinkley Point A, the temperature assessment is a calculation of maximum assessed steady state can temperature and maximum assessed transient can temperature based on measured CGO temperatures, control rod positions, etc, Following the temperature assessment, eight CGO thermocouples are displayed on an 8-point recorder.

Other stations have different arrangements and methods of assessment. At some stations the trans­ducers providing the basic information are connected to data loggers and computers which perform the necessary calculations automatically.

Channel gas outlet temperatures In the case of the Berkeley magnox reactors, each re­actor core consists of a graphite moderator with 3265 fuel channels through which coolant gas can pass. As the gas passes along a channel it picks up heat from the fuel elements reaching a maximum temperature at the channel gas outlet (CGO). A total of 312 chan­nels have been allocated thermocouples to measure CGO temperatures across the core. A number of CGO thermocouples are utilised in conjunction with tem­perature trip units (TTUs) for reactor asymmetric re­activity fault protection. As with the FE10 trip units there are predetermined trip levels and margin to trip alarms. The majority are connected to the high and low speed temperature scanners. A number of CGO thermocouples (T/C) are utilised for reactor tempera­ture control as part of the ‘sector’ auto-control system. Each individual sector uses eight T/Cs for control purposes and four T/Cs for temperature indications. The controlling T/Cs are arranged in an auctioneering circuit with any inbalance shown on a panel indicator. The remaining four T/Cs are used purely for informa­tion purposes and can be displayed either individually or as an average temperature on the panel recorder. In addition, for the benefit of the operator in deter­mining trends, there is a 12-point recorder situated in the east annexe of the CCR complex. As there are a large number of CGO thermocouples, the loss of any individual system would not cause the operator any major problem.

The operator may be faced with fault conditions that require him to shut the reactor down very quickly but which do not warrant a reactor trip. These conditions may not give him very much time to prepare and to a great extent he is following the same route as the — reactor trip condition. The following are some events which may cause him to adopt the emergency shut­down procedure:

• Failure of a standpipe assembly or of a control rod actuator assembly to seal after insertion.

• Incorrect insertion of a standpipe assembly indi­cating a fault on the assembly or the possibility of misplaced fuel.

• A dropped fuel element located either in a channel or on the charge plates in the reactor giving con­cern that fuel is damaged.

• Rising BCD signals which give concern as to the integrity of fuel cladding.

• Loss of BCD signals due to failure in the BCD equipment, i. e., failure of the BCD system.

• Gross moisture ingress to the coolant circuit from the failure of a boiler tube.

• Indications that there may be a gross CO; coolant leakage.

The basic intent of an emergency shutdown or a fast controlled shutdown is to ensure that the unit is taken off as quickly as possible in a safe and efficient manner, without damage to fuel cladding or plant and equipment, and to ensure that boiler feed is controlled such that an unacceptable boiler regime is prevented or water carry-over does not occur.

The shutdown sequence to be carried out is similar to that of the reactor trip except that control rods are not tripped-in to commence the sequence of events. The fault still has to be investigated and it is essential that the true cause is found as soon as possible, since this may introduce a problem into the shutdown sequence.

Uranium is the only naturally occurring fissionable element. It is a dense, shiny, metal which melts at 1130°C. There are three crystal forms:

• Below 66l°C it is orthorhombic (a-U).

• Between 661 °С and 769°C it is tetragonal (/3-U).

• Between 769°C and the melting point it is body- centred cubic (7-U)°C.

The а-/3 phase transmission in uranium is accompanied by a volume change of 1% and, since the metal is

non isotropic, thermal cycling through the transition temperature can produce internal cavitation of the metai and, at the very least, surface wrinkling. If this

s to be avoided, it is necessary to operate the ura­nium fuel below the tt-3 phase boundary so that, for pure uranium, 661 °С is the maximum feasible fuel temperature. It should be remembered however that, because of uranium’s high thermal conductivity, this upper temperature limit is not such a severe [imitation it would be in a ceramic fuel.

1 Een within the а-phase, however, thermal cycling can still cause problems because the thermal expan­sion coefficients of a-uranium are markedly aniso­tropic; the metal expands in the а-direction whilst contracting in the c-direction. In a bar with completely randomly oriented grains this would be of no con­sequence apart from the high levels of internal stress but, eiven a small degree of preferred orientation, most bars will produce some shape change, This ef­fect becomes more pronounced as the cycling tem­peratures are raised closer to the ck-j8 transition. Since any reactor component is bound to be subjected to considerable thermal cycling over its life, this is a potentially serious problem. The anisotropy of a — uranium is also responsible for the phenomenon known as irradiation growth. Here, the vacancies and inter­stitials which arise from the neutron damage tend to redistribute non-uniformly, producing an elongation in the ^-direction and an equal contraction in the а-direction. Again, this produces high internal stresses and, given some preferred orientation, changes in shape or ‘growth’.

If these problems are to be avoided, it is essential to use a type of uranium fuel with a very low degree of preferred crystallographic orientation. In the UK thermal reactors, ‘adjusted’ uranium has been used (Eldred, Harris, Heal, Hines and Stuttard, 1973 [3]). This is an alloy of uranium with 1000 to 1500 ppm aluminium and 200 to 500 ppm iron, both of which occur in commercial purity material. The aluminium forms a very fine dispersion of UAh precipitates which, on quenching from the j3 phase, ensure a uni­form grain size (typically 0.25 mm), with very little preferred orientation. Such adjusted uranium has per­formed well in the UK magnox reactors (Eldred et al,

1 ’13]) and, though some shape changes are ob-

■ici’.ed (Fig 1.34), these are sufficiently small to be accommodated by the ductility of the magnox can.

The avoidance of growth by the elimination of preferred orientation, does not prevent the generation 0! high internal stresses caused by the non-isotropic condensation of point defects (or by thermal cycling, though this is not so significant in-reactor). Such internal stresses make the uranium susceptible to de­formation by the yielding creep mechanism of Roberts and Cotterdl, 1956 [4], whereby additional applied heviatoric stresses can produce much more rapid creep than would occur without the internal stresses. In uranium this mechanism is important in the same

GROWTH RATE NOT alloy SENSITIVE — — CAVITATION PURE uramum

———- cavitation AOJUS-ED uranium

———- SOLID ftSStON PRODUCT SWEll’NG

———- STRESS-TREE GAS BUBBLE SWELLING

Fig. 1.34 Modes of dimensional instability in alpha
uranium

which arise from the differential strains between grains and which are brought about by irradiation growth and thermal cycling. At burn-ups greater than about

5.4 GWd/t and at fuel temperatures between 300 and 500°C, swelling becomes more rapid as the fission gas bubbles formed in the early part of the irradiation become larger. At this stage, as shown in Fig 1.35, the swelling rate can increase by more than an order of magnitude producing the so called breakaway swell­ing. The corresponding increase in the internal surface area of the fuel and the release of fission products into these pores make the fuel very chemically reac­tive. Since the consequences of a can failure could then be severe, fuel is generally removed from the reactor before this process becomes advanced. It is, perhaps, a little surprising that extensive bubble growth can occur in this fairly low temperature range (0.4-0.55 Tm where Tm is the melting point in K) and it seems that the constant generation of dislocations by the growth process may well play a part in sweeping small gas bubbles into the grain boundaries.

It is quite possible, of course, that a solution to this problem could be found but this would require extensive research and development and there is little incentive to do this since, at around 6 GWd/t(U) the fuel also begins to run out of nuclear reactivity. Nonetheless, further opportunities for increasing the channel average irradiation, as opposed to the peak value, do still exist through the use of stratagems such as axial flux shape flattening and axial and radial shuffling of elements.

Current PWR practice is to control the pH of the re­actor coolant (in the presence of boric acid), with the enriched lithium 7 form of lithium hydroxide (LiOH). This material has the advantage that it is produced in-core in the coolant during operation from the boron — 10 isotope by the reaction B-IO (n, a) Li-7. In addi­tion lithium 7 does not produce radioactive species when irradiated and has a low neutron capture cross- section. The latter means that it does not play a sig­nificant role in the core reactivity.

Control of the reactor coolant pH is implemented by the addition of lithium hydroxide (LiOH), typi­cally over the range 0.7-2.2 weight parts per million (wppm) as lithium (equivalent to 2.4 to 7.5 wppm

specification and the boric as 7LiOH) according £ Thjs produces a pH over

acid concentration^ ^ ^ 25°c for jow and high ht-

the range 6. у°оП5 respectively, which corresponds to fdu/n concentr 6.8 to 7.3. This, in turn, will

i-i гзя£^

a eeneral materials corrosion and limit the

mechanism of corrosion product release and transport between in-core and out-of-core RCS surfaces, A modification of this approach is to a lower lithium limit of 0.2 ppm, as discussed later.

Naturally-occurring lithium consists of two isotopes, Li-6 (7.42%) and Li-7 (92.58%) and since lithium 6 leads to the undesirable formation of tritium (H-3) by the reaction,

Li-6 (n, 7) H-3 (tj_ 12.26 y)

lithium hydroxide enriched to 99.9% lithium 7 is used. Enriched lithium hydroxide used in this way is ex­pensive, due to the required isotopic enrichment pro­cess, and can be subject to constraints on availability.

10.7,3 Alternative alkalising agents

Three alternative alkalising agents have been consi­dered in a PWR context, the incentive — being the high cost of enriched lithium (7) hydroxide:

• Ammonia (NH3). Concentrations of ammonia in water (present as ammonium hydroxide NH4OH) over a typical range of 10 wppm to 4 wppm could produce appropriate pH control. However, there is a major disadvantage in its radiolytic decomposi­tion; requiring continuous addition and leading to the production of N2 and H2.

• Sodium hydroxide (NaOH), There are no solubi­lity limitations on the use of sodium hydroxide, and both sodium hydroxide and sodium metaborate (Na BO2) have an increasing water solubility with temperature. However, there would be concern in the area of stress-corrosion cracking of austenitic steels, and in addition the naturally-occurring so­dium isotope Na-23 (100%) would produce sodium 24 by the reaction: Na-23 (n, 7) Na-24 (tx 15h). This would lead to a rapid and significant contri­bution to primary circuit activity.

• Potassium hydroxide (KOH). Appropriate pH con­trol could be achieved with potassium hydroxide concentrations between 1 and 12 wppm. The solu­bility of potassium metaborate (KBO2) would be expected to increase with temperature (i. e., exhi­bit a positive temperature coefficient) by analogy with sodium metaborate, and contrary to lithium metaborate.

As with sodium, a problem arises with activation of natural potassium which consists of two isotopes,

K-39 (93%) and K-41 (6.9%), by the following re­actions:

K-39 (n, 7) K-40

K-41 (n, 7) K-42

The most significant use of potassium hydroxide is at Loviisa (Finland) which is a PWR of Russian design. This is based upon the view that potassium is less corrosive than lithium towards zirconium alloys in the situation where local concentrations occur. The potas­sium hydroxide requirement is derived as between 4 and 12 wppm depending upon the boric acid concen­tration, and after allowing for the presence of lithium (from B-10), ammonia (added to produce a dissolved hydrogen concentration in core by radiolysis) and so­dium impurity. This chemical control is used to main­tain a pH at 300°C of around 7 throughout the fuel cycle.

The bellows units (Fig 2.10), consist of a series of convolutions in either low alloy or stainless steel forming the pressure boundary and a restraint system absorbing the axial pressure loads which act on the duct system. The restraint units have either flexible tongues or bars connected to the units on either side of the convolutions, or rigid structures with hinge pins. The restraint system has also to protect the convolutions from shear and torsional loads and be flexible enough to limit the loads in the ducts and at the duct terminal ends (pressure vessel, boilers, and gas circulators) to acceptable values.

Main gas valves

The main gas valves are required to isolate a boiler circuit from the reactor and to permit access to the duct and boiler system with a CO2 or air atmosphere in the reactor. The design requirement for the valves

is that they should be capable of rapid closure in the postulated event of a fracture of a duct or pipe attached to the duct system and be leak-tight when shut. The valves have electrically operated opening and closing mechanisms, but can also be manually operated.

Two types of butterfly valves have been developed for use at magnox stations. In one type the seal be­tween the body and the valve disc is obtained by jacking the disc onto a seat in the closed position. In the other type, the seal is effected by inflating a flexible metal sealing ring, welded to the valve body, onto the disc when it is in the closed position.

Boilers

The boiler shells, which contain the tubes through which heat is transferred from the coolant gas to the water/steam, are large pressure vessels of some 6 m diameter and up to 30 m in height. The shells are designed and built to the same standards as the re­actor pressure vessel and ductwork and, with the ex­ception of Sizewell At are made from a fully-killed mild steel. The Sizewell A boiler shells are made from a low alloy steel (Ducol).

The boiler tubes pass through the boiler shells via thermal sleeves to accommodate the temperature dif­ference between the tubes and the shell.

Since the water/steam pressure in the boiler tubes is generally higher than the coolant gas pressure, any leaking tube will raise the pressure in the gas circuit. Safety valves are therefore provided on each boiler gas circuit capable of handling the full discharge from both ends of a completely severed tube. The safety valves discharge through filters to remove any par­ticulate matter which might possibly be radioactive.

Genera! requirements for the cabling systems for neu­tron detectors are:

• Maintenance of high insulation resistance at high temperatures and in the presence of high levels of neutron and gamma radiation.

• Immunity from electromagnetic interference.

• Low microphony, i. e., do not generate spurious signals when moved.

• Mechanical characteristics appropriate to the ap­plications.

Cables for DC ionisation chambers have to ensure that leakage currents are small relative to the detector ion current so that leakage does not cause significant error. The techniques used are:

• Special insulant materials, e. g., PEEK.

• Special configurations with multiple screening, e. g., superscreen trilaminax and colaminax.

Cables for pulse counters are important because the charge in the electrical pulses caused by the ionisation are relatively small and with the wide band width of

the counting equipment, the whole system is suscep­tible to electromagnetic interference.

This may be introduced through the earrhins sys­tem. This problem occurs in other measurement sys­tems and is discussed in Chapter 8 of Volume V The su^cepu’bilits to interference is reduced to accept­able levels by a combination of multiple screening configurations combined with very careful attention to the earthing and bonding practices.

There are a number of different types of penetrations which pass through the concrete vessel wall and top and bottom slabs. These provide access for each fuel stringer and control rod; water and steam lines from the boilers and reheaters; gas circulators; entry to the reactor when shut down; safety valves; reactor instru­mentation; nitrogen injection; remote in-service inspec­tion equipment and coolant gas ducts to and from the gas treatment plant.

The penetration diameters vary in size from 75 mm to 2000 mm. They form part of the primary coolant pressure boundary and are designed, constructed and inspected to high standards. Failure of a penetration could result in a reactor depressurisation rate suffi­cient to cause the fuel pins to overheat and fail, thus releasing activity. The resulting high temperature gas could cause malfunction of safety and shutdown equip­ment, and the gas pressure rise could result in structural damage.

It is now practice to have, where possible, second­ary retention and/or flow-limiting devices to make the failure of a primary closure or penetration a tolerable

event.

6.5.1 Liner

The function and design of the liner for the concrete pressure vessels is similar to that for the magnox stations. There is now a considerable amount of in­formation available from test work on which to base the design of the liner and its lugs.

Immediately post-trip, the steam main balancing header valves are closed to split the boiler steam systems into two pairs in line with the emergency boiler feed ssstem (EBF). The LP vent system becomes opera­tional about 35 seconds post-trip and reduces the main boiler pressure to 80 bar at a rate of 0.5 bar/s. The emergency boiler feed, reactor sea water, and other essentia! systems pumps are started in sequence.

Emergency boiler feed will not normally be intro­duced into the main boilers with the reactor pres­surised, unless the following X system failures are detected:

• Loss of starting and standby feed whilst the steam temperatures are high.

• Loss of decay heat boiler system after termination of starting and standby feed.

• Loss of all circulators.

• Reactor pressure is less than 28 bar.

Detection of these conditions, on a quadrant basis, signals the EBF feed isolating valves to open. Once EBF is introduced it will continue until terminated by the operators.

For long term cooling with the EBF system, the operators must conserve water and take action to make additional water supplies available or establish a re- circulatory route via the start-up vessels and the main feed and condensate systems.

In addition to in-service inspection, periodic in-service testing of components and systems important to safe­ty is carried out to verify their operational readiness during service life. The frequency of testing a com­ponent depends on whether or not it can be fully tested on-load, and on the operational experience of the component or system.

Pumps are tested at specific reference conditions related to pre-service or commissioning tests, when the performance of the full system is verified against the system functional requirements. Generally, safety sys­tem pumps not normally running during plant opera­tion are tested at monthly intervals.

Valves are tested during plant operation by stroking over their full range of movement, where practicable, or otherwise by partial movement. Full movement and leakage tests are carried out if necessary during shutdowns, where on-load testing is not practicable.

Where practicable, remotely controlled systems will be installed to allow for testing of components such as pumps and valves without a need for the plant operator to be in the close vicinity of the component under test.

Following a shutdown for refuelling or in-service inspection, a visual examination is carried out to detect evidence of leakage prior to restarting the reactor. This examination is performed during the system leakage or hydrostatic tests. In areas associated with high radiation levels, remote viewing equipment is employed in order to minimise the radiation dose to operators performing the examination.

The primary containment is leak tested approximate­ly every three years to ensure that its leak-tightness has not degraded significantly.

A major proportion of the site licence deals with re­quirements to measure, instruct in precautions, control, and record the dose of nuclear radiation absorbed by personnel engaged in duties on the site. These require­ments are the responsibilities of a department staffed by Accredited Health Physicists to give the appropriate technical and scientific advice required by the licensee. The following represents the bulk of the activities of Health Physicists:

• Film badge and dosimetry services are required un­der the site licence to monitor the radiation dose to persons employed or visiting the site. Additionally, measurements of dose to the public outside the perimeter fence need to be kept under review.

• Monitoring of radiation levels and radioactive con­tamination within and without the site boundary is necessary, both for action and to collect a history of events. Off-site monitoring encompasses both land and sea samples for dairy farming, agriculture and marine life. Airborne activity is measured by the use of" tacky shades to collect dust samples. Gaseous samples are taken and measured at source.

• Zoning of the site into the four types of radiation areas and the four types of contamination areas is a requirement of the site licence. Generally all these zones are found on a nuclear power station, and notices are displayed to identify them.

• Radioactive waste products occur in the normal operation of a nuclear power station. High dose waste is normally stored in shielded vaults pro­vided for that purpose, but the bulk of consumable waste (e. g., paper filters, paper handkerchiefs, waste oils, effluents, etc.) is of a low active content and is therefore stored on the site until disposal. The health physics department would normally provide the service to store and dispose of this waste.

• Instruction to personnel on the hazards of nuclear radiation and contamination, and the precautions that must be taken to either eliminate or reduce the effects of the hazard, have to be given to persons

engaged in activities on the site. Usually the health physics section would carry out this responsibility.

• Records of all the aforementioned items have to be recorded for the site licence and again this is done by the health physics department.

• Advice to other departments is given by Accredited Health Physicists to enable them to use and set up suitable precautionary measures for access to zones or work on plant and equipment that may have an associated radiological hazard. This advice is par­ticularly important to senior authorised persons who may use it for issuing safety documents.

• Laundry services for coveralls, underclothing, socks, surgeon’s boots, rubber gloves and any other item provided for radioactive contamination control are necessary. This laundry service needs to be able to deal with radioactive contaminated clothing and re­duce that contamination to an acceptable level for the clothing to be re-used. It is usual for the laun­dry service to be extended to wash overalls worn by staff in the course of their normal duties.

• Emergency schemes are required on all nuclear sites to cover both on-site accidents including those which may have a direct effect on the public. Such schemes lay down the procedure for all employees, including the formation of health physics monitoring teams, incident survey, damage control, first aid, fire teams and the emergency controller and his supporting staff. The emergency handbook details the structure of the emergency scheme, including the reporting procedures and the use of the police, fire brigade, and the local health authority services. Normally the head of the health physics department will put this scheme together.