The primary function of the moderator in a uranium — fuelled thermal reactor is to slow’ down the fast neutrons released in fission to thermal energies of less than 1 eV at which further fission of the U-235 isotope can occur. For best neutron economy, the moderator must therefore contain insignificant parasi­tic absorber and exhibit a high degree of neutron scattering, thereby achieving many slowing down colli­sions and large loss of energy per collision to avoid unwanted resonance capture in the U-238. In addition to this primary function the moderator for a magnox reactor is the structural material for the whole core and must remain effective for the reactor life. Its physical integrity is important in maintaining the align­ment of fuel and control rod channels to ensure free­dom of charge/discharge and safe shutdown. Graphite meets these requirements and can be processed in a pure form for magnox reactors from petroleum coke (PGA graphite). It does not lose strength as its tem­perature is raised and with its low coefficient of thermal expansion and high thermal conductivity it offers excellent resistance to thermal shock.

Fast neutrons produced in fission of the U-235 have energies up to about 14 MeV. Moderation of these neutrons down to thermal energies is caused by elastic collisions with carbon nuclei. Since only 25-60 eV is needed to displace a carbon atom in the graph­ite lattice, many such displacements occur, due not only to neutron collisions but also to secondary colli­sions as the recoiling carbon atoms bounce around in the lattice. In the peak flux regions of the core, each graphite atom is on average displaced many times during reactor life. This leads to the core graphite exhibiting:

• Changes in dimension (both growth and shrinkage),

• Changes in physical properties including strength and modulus, thermal conductivity and creep.

• The storage of energy.

The main effects are temperature dependent and are generally most deleterious at temperatures below 200°C. Since graphite moderator temperatures in magnox reactors are generally above 200°C the worst conse­quence of these effects are avoided. Only at Berkeley, the first CHGB magnox station, was it considered necessary at the time of construction to raise the tem­perature of the lower core graphite. This was accom­plished by the use of magnox sleeves placed in the lower sections of all fuel channels to restrict the heat flow from the graphite to the gas coolant.

Another important effect of irradiation on the core, particularly in the later concrete pressure-vessel high gas-pressure stations, is the temperature dependent reversible oxidation reaction between graphite and the reactor coolant. This radiolytic oxidation process in the graphite core leads to:

• Reduction of graphite strength and modulus.

• Formation of carbon monoxide (CO) and hence

possibility of deposition of carbon within the gas

circuit.

• Loss of moderation due to loss in graphite density.

A degree of protection against radiolytic oxidation must therefore be provided particularly in the later stations. This is accomplished by careful choice of the reactor coolant composition. In magnox reactors the primary impurities in the CO2 coolant are carbon monoxide (CO) and hydrogen/water. Carbon mono­xide (CO) is also the primary inhibitor of the graph­ite radiolytic oxidation reaction and is therefore per­mitted to build up to a maximum concentration of

1.5 v/o. Above this concentration little further de­crease in the oxidation rate occurs. The CO inhibition process occurs due to the short-lived ‘reactive oxidising species’, formed as a result of irradiation, combining with the CO before it has a chance to attack the graphite. Hydrogen/water also acts as an inhibitor with no further effective increase above 150 vpm; the inhibition process is due to the small quantity of hydrogen/water (total of 25-100 vpm) forming a slightly sacrificial protective layer on the graphite sur­face and reacting preferentially with the incoming ‘oxidising species’. (This subject is discussed fully in Chapter 1.)

For nuclear physics considerations the mean density of the graphite structure should be as high as possi­ble. Removal of graphite for fuel and other channels is inescapable, but all other voidage should be mini­mised. For this reason, and also to minimise the cost of graphite wastage associated with machining, a structure composed basically of bricks of the largest size that is compatible with reactor physics considera­tions is preferred. The need to reduce leakage of coolant between a channel and the interstitial spaces between bricks, of particular importance where the spaces are vented to the outlet header as is normal with vertical channels, dictates that the channels should be bored through the centres of the bricks rather than machined at the interfaces between adjacent bricks. To reduce leakage further, the number of brick junc­tions in the length of a channel should be small. This leads to a generally preferred arrangement in which bricks of cross-sectional dimensions approximately equal to the lattice cell of about 203 mm are bored longitudinally to the required channel diameter. A brick length of some 762-838 mm gives a unit of weight and size convenient to manufacture and han­dle, and a complete reactor core can be formed to the required dimensions by stacking these unit bricks. Magnox reactor cores have about 11 or 12 brick layers and contain between 2500 and 6000 fuel channels.

The magnox reactor core and restraint structure at Oldbury is shown in Fig 2.4.

Additional channels are required for control rods, absorber, flux scanning and for specimen irradiation purposes. In a vertical channel design these channels would normally be parallel to the fuel channels and could either replace some of them, to give a uniform structure, or occupy intercell positions off-pitch from the fuel channels. The latter arrangement is acceptable since exactness in dimension is less critical with these channels and the possible leakage of coolant from them through the longitudinal brick junctions in the channel walls is not important: indeed, the arrange­ment is preferable in some way since it gives a uni­form fuel lattice, less radial flux distortion and a greater power density from the core. Bradwell is the only station without these intercell (interstitial) chan­nels. Intercell channels will tend, however, to increase the outleakage of coolant from the fuel channels un­less a reliable seal can be included at each interbrick junction.

Thermal expansion and either growth or shrinkage of the graphite in a core having vertical channels can be easily accommodated in the vertical direction by allowing each column of bricks to move independently of its neighbours and to be restrained only by gra­vitational forces. In this type of structure dimen­sional changes in the radial direction in the individual graphite bricks, arising from thermal expansion and irradiation, are conditioned by the properties of the graphite in the direction perpendicular to the extrusion axis of the bricks. The thermal expansion will vary at different levels of the core depending upon the axial temperature distribution, whilst the irradiation — induced growth or shrinkage will vary in both the radial and vertical directions depending upon the flux and temperature distributions. A moderator composed simply of plain rectangular graphite bricks would not be able to accommodate these changes and remain structurally stable, therefore an interlocking or keying system has to be adopted. This system, in addition to accommodating the dimensional changes, has to ensure alignment of channels and a uniform lattice

pitch during all reactor operational conditions. The early magnox cores employed the brick/tile column configuration whilst the later magnox cores used a fully interlocking keyed structure. Both of these de­signs are described as follows.

The graphite core sits on a mat of support plates which provide a level surface for the graphite struc­ture. Each plate is keyed to the main core support structure, the diagrid, with gaps between the edges of adjacent plates so that each plate is free to expand and contract from its centre.

The early magnox cores are designed to thermally expand as graphite in the radial direction and employ the brick/tile column configuration with the graphite columns being mounted on ball thrust bearings resting on the support plates. The active core containing the fuel is surrounded by a 600-900 mm thick graphite reflector which is kept in the solid cylindrical form by encircling elastic garter restraints. The structure is
then free to undergo movements that are associated with the thermal expansion of graphite. This reduces the thermal radial disturbance of the core to a mini­mum. However, in view of the slight possibility of core bursting pressures arising under fault conditions, an encircling rigid steel restraint structure is needed as a ‘back-up’.

The elastic garter restraints referred to consist of a number of steel tie-bars pinned together. The ar­rangement is shown in Fig 2,5. Each tie-bar comprises a central rod inside a series of concentric tubes. The central rod and alternate tubes are made from mild steel and are in tension whilst the other tubes are made from stainless steel and are in compression. The design is such that the load is transferred from the central rod to each tube in turn to obtain a suitable spring rate for the tie-bar and an overall coefficient of thermal expansion which approaches very closely that of graphite. This ensures that stress changes in

he elastic garter restraints are minimised throughout I e range of temperatures concerned.

In early magnox reactor cores the graphite columns e formed from alternate layers of graphite bricks arid tiies keyed together as shown in Fig 2.6. Lines of contact were established through every column of Graphite across the reactor at each tile layer along Juo perpendicular axes in the horizontal plane. This was accomplished by using pairs of tiles keyed to — ther between the brick layers with their longer sides being parallel to the extrusion direction and at right ansles to each other. Space was left for growth of the shorter sides of the tiles and around the four vertical sides of the bricks. This design was adopted since at the time of construction it was believed that the dimension parallel to the extrusion direction would remain stable, but as experimental evidence became available it was realised that the tiles would shrink in this direction and therefore cause slackness in the araphite structure which might endanger fuel cooling and control rod insertion. Therefore, at a late stage in core construction, zirconium pins were inserted into horizontal holes and slots in both the moderator bricks and reflector tiles to maintain the channel lattice pitch when the inevitable shrinkage of the tiles occurred.

An alternative solution for maintaining the channel lattice, which avoided the wasteful absorption of neu­trons by the zirconium pins, was evolved for the later magnox reactor cores using an interlocking vertical keying system for the graphite bricks as shown inset in Fig 2.4. All the later magnox cores adopted inter­locking keying systems except for Wylfa which retained the solid cylindrical reflector, but the keys were now incorporated in both the active core and side reflector with the peripheral reflector bricks being attached to the steel restraint structure at each graphite layer. In addition, the bottom graphite bricks are spigotted directly to the steel support plates and bricks within the same column are spigotted together. This con­struction results in the effective radial thermal expan­sion of the core being controlled by the encircling rigid mild steel restraint structure which is mounted on the diagrid. As stated earlier, the steel restraint structure, in addition to controlling the stability of the core under normal operating conditions, would also act as a re­straint under fault conditions that give rise to radial bursting forces in the core.

There is a temperature variation up the height of the restraint structure which will determine its geome­tric profile. The peripheral reflector bricks will follow the movement of the restraint structure and as the temperature rises the graphite columns will be pulled radially out with maximum displacement at the core periphery and at the hotter upper part of the structure. The distortion of the columns from the vertical will progressively reduce to zero at the core centre and voll be further diminished by the clearances in the ‘nterlocking brick keying system. It is important to rmnimise this distortion during the operating condition

so as to allow unimpeded entry of control rods and other shutdown devices, and the safe loading and unloading of fuel, and some graphite cores are preset during assembly. This is accomplished by means of adjustable links between the restraint structure and the graphite side reflector, the amount of preset at

each graphite layer being dependent on the calculated amount of radial movement that will occur at the particular level when at operating temperature.

The top of the core is covered with an array of charge pans to facilitate fuel and control rod entry and also to protect the graphite bricks from damage. In the earlier magnox stations the charge pans were designed to rest on the top graphite reflector bricks but later stations adopted designs where they were suspended from the standpipes. These designs allowed for the maximum axial graphite shrinkage that would occur over reactor life so that the charge pan sleeves would remain inserted into the fuel and control rod channels to ensure charge path continuity.

Other magnox reactor core design features include:

• A duplicate diaphragm seal between the pressure vessel and the core periphery to minimise the leak­age of coolant around the outside of the core.

• A means of setting the gas flow through each in­dividual fuel channel. This is achieved in magnox reactors by the use of preset gags which are located at the bottom of the channel and are replaceable so that re-gagging can be accomplished if necessary (see Fig 2.7).

• A shock absorber which is incorporated in the fuel element support at the bottom of the channel to take the impact of an accidentally dropped fuel element. This feature is also shown in Fig 2.7.

The basic DC14 chamber unit is a conventional DC fission chamber consisting of a thin walled, stainless steel, cylindrical envelope containing two coaxially mounted electrodes, the HT and the collector, These electrodes are individually mounted off the envelope on alumina insulators and connecting leads are brought out of the envelope via metal-ceramic seals, two for the HT and one for the collector. Three stainless steel, coaxial, mineral insulated cables are joined to the chamber unit with their screens welded to the chamber envelope and their cores welded to the metal — ceramic seals. They provide HT, HT tell-tale and signal connections.

A guard ring construction is employed in the cham­ber unit so that direct insulation leakage current from the HT electrode to the collector electrode is elimi­nated. If a virtual earth input amplifier is used to measure the signal current, leakage to or from the collector electrode is reduced to a minimum.

Neutron sensitivity is provided by coating the fac­ing surface of the electrodes with uranium oxide of

0.75 mg/cm2. The chamber is filled with Xenon +

1 % Helium. The DC14 is available in two standard sensitive lengths which give neutron sensitivities of

1.2 x 10-!4 A/nv and 1.6 x 10-14 A/nv.

The DC14A chamber is a DC14 chamber unit en­closed in and insulated from a pressure-tight stainless steel containment. Three triaxial mineral insulated ca­bles with copper cores, copper inner sheaths and stain­less steel outer sheaths are joined to the assembly, so that the copper cores and inner sheaths go to the chamber unit and the stainless steel outer sheaths go to the containment. At the end remote from the chamber each cable is terminated in a special cold end termination. This hermetically seals the cable end and carries a coaxial connector joined to the cable core and inner sheaths, which is insulated from the outer sheath. Connection to the cold end termination is made via a coaxial cable with an insulating outer covering so that the DC14 chamber unit inside its containment is completely insulated from the reactor structure.

The containment increases the pressure rating of the system to 45 bar at 550°C.

The cold end termination can incorporate a seal plug which seals to the reactor pressure vessel.

The cold end termination is not demountable and the cable end is pumped, baked and filled with dry gas during assembly of the cold end termination so that the exact length of mineral insulated cable must be specified when it is ordered.

A significant advantage of the DC fission cham­ber over the boron type, is that the fall in sensiti­vity due to neutron irradiation (‘burn up’) is much less.

As in the case of the boron DC chamber, the ac­curacy of measurement at low neutron flux after ir­radiation is limited by internal radiation effects. These arise mainly from the longer half life fission pro­ducts in the uranium coating and give rise to cur­rents of about 0.2^0 of the maximum current at w hich the chamber has been working when in the neutron flux.

DC fission chambers are therefore mainly used in high levels of neutron flux and where measure­ment is required only over the upper two decades.

Cables for use with ionisation chambers are dis­cussed in Section 5.2.9 of this chapter.

The core for the first commercial AGR at Dungeness В was structurally similar to the latest magnox station at WyHa. The increased power density was matched by a proportional increase in radiation heating in the graphite moderator (about 6% of total heat gen­erated). Graphite moderator temperatures however could not be allowed to increase much beyond those reached in the magnox stations, about 400°C maxi­mum, thermal oxidation becoming significant above about 600°C.

Moderator cooling by direct flow through the core parallel to the fuel cooling flow, as in the magnox reactors, would result in severe dilution of the fuel outlet gas flow which reaches 650-675°C with mod­erator cooling gas at, say, 450°C. System thermal efficiency clearly demanded a design which avoided this dilution. The solution chosen was that of re-entrant moderator cooling, i. e., a downwards flow of coolant through the moderator in series with the fuel. To achieve the required graphite temperature, about 40% of the total circuit flow is directed through the re­entrant moderator cooling path. The balance is either allowed to enter the fuel channels directly through the bottom of the core where it mixes with the re­entrant flow or is directed to cool the side shield and restraints. Re-entrant moderator cooling allows the whole of the circuit mass flow to be raised to the maximum temperature permitted by the fuel.

Figures 2.75 (a) and (b) show for comparison the main flow and temperature patterns of typical mag — nox and AGR stations.

It can be seen above, that coolant flow is somewhat more difficult to control in the AGR than in the mag — nox reactors. The feature requiring significant inno­vation relative to magnox is the hot box membrane, which has to separate the relatively cool re-entrant gas from the fuel outlet gas. Not only does it have to cope with the high differential temperature but it also has to carry a large part of the circuit driving pressure, say two or three bar, or the full pressure drop across the fuel and core.

There have been three different designs of ‘hot box’ or ‘gas baffle’ as the membrane separating reactor inlet gas from outlet gas is called (Fig 2.76).

Figure 2.76 (a) shows Windscale (WAGR) a feature of which is the stainless steel hot box which isolates the fuel outlet gas from the containment. Although a successful embodiment of the AGR principle the flexi­ble gas-tight connections to the core, to the boilers and to the top cap (for fuel and control rod insertion) were very difficult to extrapolate to commercial AGR size. Figure 2.76 (b) shows the first commercial solu­tion to the hot box design problem. It turns the hot box inside out and, as a gas baffle, it surrounds most of the cooler gas volume. The gas baffle is at­tached to the bottom of the concrete pressure vessel and the main gas seals are those between the fuel and the gas baffle dome. By applying thermal insu­lation to the outside of the upper region of the gas baffle, it was possible to keep its temperature close to that of the core support and restraint steelwork and use similar mild steel material; thus obviating many
of the differential expansion problems associated with a stretched WAGR arrangement.

Figure 2.76 (c) shows the second commercial solu­tion to the hot box problem, differing from the first in that the connection to the vessel is near the top of the structure rather than at the bottom. In this case the separating membrane is called the ‘hot box dome’ and the ‘hot box’ has a partly steel, partly concrete boundary with insulation on the hot faces.

To ensure adequate cooling of the structural steel­work, shielding, graphite, fuel and core facilities (such as neutron sources, flux detectors and control rods), the flow distribution has to be positively controlled by design. This is done using large two-dimensional flow network programmes in which all the core flow pas­sageways are represented on a radial section through the core.

This type of programme is called a ring model since it represents the whole core as a number of concentric rings, equal to the number of columns on a core ra­dius plus further rings to represent the core/shield and shield/gas baffle annuli where appropriate. Each ring in the model is allocated all the axial flow pas­sages at that radius, and each ring is connected to the adjacent rings by representative radial cross flow links. Flow resistances for the links in the network are obtained mainly by rig testing.

The model is used to obtain axial and radial pres­sure gradients and general cooling flows. These are required for the calculation of gas pressure loads on the graphite and restraints and for component tem­perature assessment. One of the most important para­meters derived from the ring model is the coolant temperature at the inlet to the fuel stringer. This is

a mix of re-entrant flow through the core and direct flow’ through the support plates, the former may be 50°C to 100°C warmer than the latter.

The ring model also provides the boundary condi­tions for single-channel network models, which inves­tigate in more detail the effects of local variations in stringer flows and conditions during fuel charge/ discharge.

A three-dimensional flow network model of an octant of an AGR core has been used to verify the output from the ring model. During commissioning of each reactor, extensive flow measurements are made to confirm that no significant errors in calculation or construction have occurred. Before fuel loading, there is a full inspection of the reactor gas circuit to ensure freedom from erection debris which might obstruct coolant passageways during subsequent operation.

The power density in an AGR is three times that in a magnox reactor and it follows that for a similar design lifetime, say 25 or 30 years, the neutron damage to the materials of the core and its immediate sur­roundings is similarly multiplied.

Neutron damage to graphite is covered in Chapter 1 of this volume. Since the fast neutron dose at the bore of a fuel brick is about twice that at the out­side diameter (due to moderation of fast neutrons), the shrinkage at the bore is initially much faster than elsewhere in the brick. The fast neutron damage gra­dient across the brick wall means that well within the lifetime of an AGR, the bore of a fuel brick begins to swell while the outer regions of the brick are still shrinking. The shrinkage/swelling across the brick wall causes internal stress within the graphite which initially increases but later reduces the strength of the brick that is available to carry externally-applied loads. The highest load applied stress of importance in a brick is at the keyway root. Bearing stresses at the keyw:ay outer corner can be higher, but local load relief by chipping at the corner does not cause brick
failure and leads to sharing of loads between the eight keys on the brick.

Another effect of the radial damage gradiant is that keyways ‘dovetail’ or close-up at the outside dia­meter thus further reducing the effective key/keyway clearance. Similar geometrical distortions occur on brick end features.

The very high radiation level in the moderator re­gion of an AGR means that additional neutron and gamma shielding has to be provided to protect the surrounding structural steelwork and boilers (or con­crete) from neutron damage and gamma activation. Approximately one metre thickness of graphite and denser material like steel round the sides, top and bottom of the active core is provided for this purpose. The steel may contain up to 3% boron to increase its efficacy as a neutron absorber. Another function of the shield region is to reduce gamma radiation levels at positions or routes where man access for mainte­nance or inspection might be required.

The high gamma flux in the moderator causes radiolysis of the coolant and the generation of chemi­cally active species which are corrosive to the graph­ite. The chemistry involved and provisions made to inhibit graphite corrosion are described in Chapter 1. Figure 2.77 shows a typical distribution of graphite weight loss across a brick section. Such weight loss is accompanied by a reduction in effective strength of the graphite as shown in Fig 2.78. This is to some extent compensated by irradiation hardening.

The effects of neutron dose, gamma dose, shrinkage, weight loss, creep and hardening are interactive, and experimental verification of the combined effects, at present calculated using pessimistic combinations of factors, is being obtained by continued irradiation experiments and graphite sampling of the operating reactors.

A typical arrangement of the moderator, side re­straints and shield is shown in Fig 2.79.

Fig, 2."8 Variation of graphite strength with
weight loss

This shows the keyed structure of large (fuel) brick columns and small (primary and secondary shutdown, graphite sample and neutron source) brick columns, surrounded by reflector bricks and reflector/shielding
steel and graphite. Key, keyway and brick sections have been optimised by calculation and test for maxi­mum strengths and provided with minimum clearances compatible with expected brick and keyway distortions through life.

It can be seen from Fig 2.76 that an essential fea­ture of the AGR is the gas tight connection between the gas baffle and the core. The gas baffle or dome is provided with integral nozzles at every core access position and a guide tube is attached to each one.

As well as isolating the hot gas from the cool gas, the other function of the guide tubes is to smooth the route for insertion of the fuel and control rods between the vessel top cap penetrations and the core channels. The detail design of the guide tubes has to accommodate any misalignment between the penetra­tions, the dome nozzles and the core channels. Erec­tion tolerances form a large proportion of this mis­alignment, but pressure and thermal movements in operation must also be allowed for. Various provi­sions are made to deal with misalignment such as the manufacture and setting of parts to site survey, inclusion of articulating joints, offsetting to compen­sate for extremes of thermal movement and in the

FLUX SCANNING
UPPER SHIELDING RING
INTERMEDIATE SHIELDING RING
LOWER SHIELDING RING
TOP STIFFRING ‘
CONTROLROD SHOCK ABSORBER AND GUIDE TUBE
UPPER SHIELDING RING
FUELSUPPOHT
LOWER SHIELDING RING
CORE SUPPORT PLATE

case of Hartlepool and Heysham У, the selection of very flexible tube design.

Vertical movements and tolerances are also impor­tant especially the axial shrinkage of the fuel and moderator under irradiation. Sliding joints take care of all vertical movements.

The post-trip sequencing equipment initiates and co­ordinates the operation of the post-trip cooling me­chanical systems, the essential electrical system and the diesel generators.

The equipment is divided into eight independent sections in line with the overall segregation of the post trip cooling plant, i. e., X and Y systems each consisting of four independent trains.

The X system comprises microprocessor-based logic systems and the Y system conventional relay logic.

All the equipment is located in the essential sup­plies buildings adjacent to the switchgear.

The equipment is normally put into service (armed) during reactor start-up by means of controls in the CCR, immediately prior to feeding the main boilers; interlocks on the feed valves prevent them opening un­til the PTSE is in service. Once the equipment is armed it cannot be switched out whilst the reactor guardlines remain healthy.

The general arrangement of post-trip sequencing equipment is shown in Fig 2.115, Reactor trip sig­nals from two out of four guardlines initiate the sequence.

The X system consists of four independam logic units:

• Decay heat boiler system.

• Main boiler feed.

• Gas circulators.

• Electrical system.

and the Y system of five independant logic units:

• Emergency boiler feed and IGVs.

• Gas plant, auxiliary cooling, and heating and ven­tilating.

• Vessel and turbine protection.

• Main boiler depressurisation and feed system.

• Electrical system.

Each of these units is made up of three independent channels, the reactor trip detection logic and the electrical supply monitoring logic producing separate signals for each of the three channels. The output signals to the plant are generated by combining the outputs from the three channels into a ‘2 out of 3’ network.

Test switches are provided to simulate plant input signals to allow the logic to be tested when the re­
actor is on-load. Testing of one channel of each logic unit at a time does not result in the generation of any plant output signals.

In addition to controlling plant operations the PTSE also isolates the associated CCR controls, to avoid the possibility of spurious signals affecting the plant operation in the event of a fire in the CCR. A secure coded signal device (one for each of the eight trains) is provided, which cannot operate spur­iously in such events, to regain normal control of plant; this switches out (i. e,, disarms) the PTSE, and simultaneously restores the CCR control facilities without affecting the plant operating status.

It is important that adequate provision is made in the design to permit the necessary IS1 to be carried out effectively and with the minimum of radiation exposure to personnel.

During system layout and component design, care­ful attention is given to physical clearances to allow personnel and equipment to perform required in-service examinations. Access provisions to meet the specified inspection requirements are considered in the design of components, weld joint configuration and system arrangement. Space is provided to handle and store insulation, shielding, calibration blocks and similar material related to the inspection.

Where required, access for ultrasonic examination oi the pressure-containing welds is provided from the external surfaces of components and piping by means of removable insulation, and shielding. If practicable, permanent tracks for remote inspection devices are presided in areas where personnel access is restricted. Working platforms are provided at strategic locations in the plant to permit ready access to tho e areas of the reactor coolant pressure boundary which are de­signated as inspection points in the in-service inspec­tion programme. Areas without permanent platforms are provided with temporary platforms as required to facilitate inspection.

10.2.11 Inspection of primary circuit components

Each of the CEGB’s nuclear sites must have a licence, issued to a licensee, under the Nuclear Installations Act 1965 before operations can commence or fuel is loaded into its reactors. The licence is issued by the Health and Safety Executive.

The licence is usually split into two schedules, first­ly that which defines the position of the installation and the land occupied by the licensee, together with a brief description of the reactor/equipment installed on it, and secondly a set of conditions describing the requirements for operating the site.

The second schedule is that which gives instruction to the licensee as to those features which must be part of his management system, although from time to time the Health and Safety Inspector may require action to be taken in areas of control, specification and management that are unsatisfactory. The follow­ing paragraphs indicate the scope and complexity of the conditions for a typical magnox power station, but these may vary from location to location depend­ing on siting and the technical nature of the plant installed. To illustrate the nature of the site licence only brief descriptions of the majority of the condi­tions are given, but later a description of how these are enacted will be explained for the significant ones.

The general conditions make provision for the dis­posal of the site, the prevention of unauthorised per­sons entering the site and the erection of a suitable fence which is properly maintained. There is a re­quirement to appoint qualified persons to perform the functions of ‘Duly Authorised Persons’. In a number of the conditions, responsibilities are given to duly authorised persons to carry out certain functions. Lists of these persons must be sent to the HSE describing which conditions he/she has a function to and his/ her qualifications and experience. The HSE may, if it notifies the CEGB in writing, exclude someone from becoming a duly authorised person. It also requires that all records, authorities and certificates issued in pursuance of the licence be kept for a period of at least 30 years or shorter if the HSE agrees.

Nuclear matter The licensee must make proper pro­vision for the storing, accounting and recording of nuclear matter. If the amount exceeds 5 tonnes of fuel elements then a suitable store acceptable to the HSE must be provided and the licence certificate for its use displayed in the store. This condition excludes any alteration to the store or its equipment without the HSE’s approval.

Carriage of nuclear matter This makes provision for the carriage of nuclear matter on site in containers designed for the purpose, or by arrangements ap­proved by the Executive. Nuclear matter to be con­signed to another site must have the approval of the HSE. There must be proper records for storage and receipt of nuclear matter. Radioactive liquid or solid waste must not be processed on the site except in a place and manner agreed by the HSE, and proper records of waste must be kept.

Dangerous occurrences, incidents and emergency ar­rangements The provision for dangerous occurrences refers to the Nuclear Installation Act and any list pre­scribed under the Act. The condition refers to the prohibition of any interference with plant or equip­ment that may have become damaged or contributed to the dangerous occurrence. It also makes provisions for the recording of any incident, in a Register of Site Incidents, which the HSE considers of a record­able class. Emergency arrangements must be made for the event of an accident or emergency. These ar­rangements must include for the supply and mainte­nance of appropriate clothing, equipment, instruments to measure ionising radiation, and changing, washing, and decontamination facilities. Proper communication facilities must be provided and at ail times readily available. If such arrangements involve the assistance or co-operation or the services of the local authority, or other body, then proper consultation between the licensee and that organisation must take place.

The persons participating in any emergency scheme must be trained and exercised in the use of equip­ment and performance of the arrangements. The fre­quency of exercises is prescribed by the HSE and is usually set at one year. A full record of all training given in this connection must be maintained giving the type of instruction and the dates it is given. All schemes must have the approval of the Health and Safety Executive.

Sufficient warning notices must be displayed to in­form persons on the meanings of any warning system, the location of exits, and the measures to be taken by personnel in the event of a fire or emergency.

Design, operation and maintenance of the plant There is a requirement for the licensee to form a ‘Nuclear Safety Committee’ for the purposes of considering and advising the licensee on the safety of:

• Alterations, additions or modifications to the plant.

• Alterations or amendment to the operating rules.

• Experiments in connection with the plant.

• Any other matter the licensee sees fit to refer.

To satisfy this condition, the CEGB has set up a single committee to serve all nuclear power stations. Those persons who serve on this committee have to submit their qualifications and experience to the HSE who may at their discretion insist they do not sit on that committee. There is a requirement that all business where recommendations are made must be attended by at least five of the members and it must be a properly constituted committee. If urgent safety proposal clearance is required, special arrangements are made with the approval of the HSE. Proper re­cords must be made of all recommendations and the HSE informed within fourteen days. The HSE requires the licensee to submit to them, as they may demand, any plans, designs or specification of the plant. Addi­tionally, where the licensee has submitted plans, de­signs and specifications, then he must ensure that the plant is built to and in accordance with them. He must make no alteration except under procedures de­vised by the licensee and approved by the HSE.

Before operation of the plant commences and to secure safe nuclear operation, the licensee has to com­pile a set of rules known as the Operating Rules which specify the methods, procedures and conditions under which it may be operated. Operation of the plant may only take place once the HSE has approved those rules; if any changes are required, those changes must be referred to the Nuclear Safety Committee and to the HSE for their approval.

To implement the Operating Rules the licensee has to make a set of instructions known as the Plant Operating Instructions, and incorporate those Operat­ing Rules within them. Before operations commence, the Plant Operating Instructions have to be submitted to the HSE for approval. Changes to the Plant Op­erating Instructions have to be submitted to the HSE within fourteen days from when they became effective (changes to the Operating Rules have to be dealt with as previously explained).

The plant may only be operated provided adequate protective devices are in operation and safety mecha­nisms, devices and circuits are duly connected and in good working order. These devices are of course main­ly for protection of the reactors but other protection for personnel and the public may be included in this category where release of radioactive matter could be a risk under fault conditions.

Control and supervision of the plant must be made by a Duly Authorised Person appointed for that pur­pose by the licensee. Under normal conditions this would be the Shift Charge Engineer or Shift Manager.

Adequate records on the operation of the plant must be made and kept and the list approved by the HSE. One particular record mentioned in the site li­cence is that of any article or fuel element loaded into the reactor. A full record must be made of its identity, time, position and condition in the reactor. The site licence contains conditions for the mainte­nance of the plant and reactors. The first requirement under the conditions is the provision of a Maintenance Schedule. The maintenance of all the plant and an — cillaries is not covered by the Maintenance Schedule but includes only those items that have a nuclear significance. It must be submitted to the HSE for its approval and in addition a set of maintenance instruc­tions, which sets out in detail how the maintenance of those items listed in the Maintenance Schedule will be carried out. The following lists generally those items included in the Maintenance Schedule:

• The reactor core and graphite properties.

• The control rod and shutdown system.

• The external and internal features of the pressure vessel including structures, relief valves, core re­straint and thermal insulation.

• The burst cartridge detection system, safety circuits and shutdown contactors.

• Gas circulators and their auxiliaries.

• Emergency boiler feed pumps and certain feed lines and valves.

• Essential supplies systems (electrical, CO2, etc.).

• Reactor refuelling equipment.

• Communication systems.

• Any machinery or equipment used in an emergen­cy or in the event of failure of any other plant or equipment used for the safe operation of the plant.

• Essential indications display or recording system.

• Radioactive effluent system and iodine absorption plants.

The licensee shall submit to the HSE the name of the person appointed to ensure the examination and maintenance of those items listed in the maintenance schedule.

Periodic shutdowns are required to carry out the ’л°гк specified in the maintenance schedule. The site licence makes a condition for each reactor unit to be shut down within a period not exceeding two years toliowing the previous consent of the HSE for that reactor to be started up or shut down. Following the shut down, the reactor may not be started up without the consent of the Executive. Consent will only be given by the HSE providing the Maintenance Schedule has been complied with and the condition of the re­actor boilers and pressure circuit examined.

The appointed person responsible for the perfor­mance of the examination, inspections, maintenance and tests is required to supply the HSE with a certi­ficate that the work is completed, and within 28 days of that certificate a full report showing the results of the examinations referred to in the Maintenance Schedule.

If the appointed person finds any reason or defect which may affect the safe operation or condition of the plant then the licensee must cause the appointed person to bring it to the notice of the HSE imme­diately and follow this with written confirmation and a report.

Full records of an accurate report must be main­tained by the licensee of all examinations, tests, in­spections and maintenance of items specified in the Maintenance Schedule.

For the purpose of the maintenance of safety mechanisms, i. e., those related to reactor safety, all work must be authorised under an authority in writ­ing granted by a Duly Authorised Person. He shall ensure that on completion of any maintenance work, the safety device shall not be put into service until it has been tested and seen to function correctly, and the person who carried out the test and the Duly Authorised Person endorses the authority as to its fitness.

Two other provisions are made as to the fitness of the plant; the first is to check at frequent intervals (usually once per month) the leakage of CO2 coolant from the pressure vessels to assess whether the extent of leakage from the vessel indicates that the plant is not operating properly. The second is any test that the HSE may require in consultation with the licen­see in the interests of safety.

General protective measures There is a need to con­trol the amount of ionising radiation or contamina­tion to persons on the site, and therefore each licensee has to provide a set of Safety Rules and Controls to ensure that this happens. All personnel who either work on the site or are authorised to be on the site must receive adequate instruction on the Safety Rules and Controls so that they may safeguard themselves against receiving unacceptable levels of nuclear radia­tion. The minimum requirements by the site licence are those as recommended by the ICRP (International Commission for Radiological Protection) and are usu­ally given in an appendix to the site licence.

Visitors to nuclear sites are generally in the cus­tody of a member of the staff who has the respon­sibility for their well-being, and are conducted in such a way that no harm befalls them. In this case the formal instruction on Safety Rules is not necessary

but it is the licensee’s responsibility to ensure their safety.

The requirements to give protection and control to personnel on the site are as follows;

• Classified Persons — a classified person is one whose duties require him to spend a significant part of his time in a ‘controlled area’. Such a per­son shall not be less than IS years of age. A con­trolled area is one where specific levels of radiation or contamination exist as specified in the ICRP recommendations.

• A classified person must have undergone a medical examination by an appointed doctor and have been entered in a health register that they are fit for occupation in that capacity. That medical must in­clude a blood test and examination. Such medicals must be carried out with a frequency not exceeding fourteen months.

If the licensee has reason to believe that a per­son has received a radiation dose in excess of that specified in the ICRP recommendations or come into contact or ingested an undue amount of con­tamination, then he must make arrangements for an examination by an appointed doctor as soon as reasonably possible.

If after any examination the appointed doctor has the opinion that a person should not, in the interests of their health, remain a classified person, then the licensee shall have their name removed from the health register and no longer use him/her in that capacity. That condition will remain until such time as an appointed doctor signifies under his signature that the person’s fitness for occupa­tional work is no longer a problem.

• A ‘Health Register’ has to be kept by the licensee to record all medicals and blood tests. If the me­dical is required because of an excessive radiation dose then the entry must be marked accordingly. The safe keeping and storage of records must be made under conditions discussed with the appointed doctor.

• There is a duty on the licensee to ensure by all reasonable steps, that any radiation dose is mini­mised to any person, and personnel are instructed not to expose themselves or others to a radiation dose that is unreasonable for the work or opera­tion thex are doing. There is also a need to control any exposure from any item on the site (including waste) in any form that would give a person a radiation dose in excess of the ICRP limits. [27] on the licensee to notify the Health and Safety Executive of the existence of controlled zones or any changes to them which may be in existence for three months or more.

Reasonable steps must be made to prevent per­sonnel entering a controlled area unless they are either classified persons or persons authorised by a Duly Authorised Person. Such entries are usu­ally controlled by warning notices for the lower radiation and contaminated areas or by notices and locked doors for the higher radiation and contami­nation zones. Contamination must be reduced by reasonable means to levels specified in the ICRP regulations, or if this is not possible then the source must be enclosed hermetically. Contamination can exist as surface or airborne.

If a person should receive a radiation dose or ingest contamination to a level in excess of the maximum specified, an inquiry into the circum­stances has to be made to determine as accurately as possible the radiation dose received. The results of the enquiry have to be entered into a radiation dose register, together with information on the part or parts of the body affected.

The radiation dose register must be open to in­spection by the appointed doctor,

• For any person who is engaged on duties in a classified area and is a classified worker, or is a person directed by a Duly Authorised Person, suit­able means of measuring the radiation dose re­ceived must be worn. This will normally be by a film badge or other suitable dosemeter. Each film badge or dosemeter must be able to be identified with the individual wearing it. Premises and equip­ment have to be provided to process the film badge or dosemeter so as to accurately determine the radiation dose it has received. The results of mea­surements for each person are kept in the radiation dose register and are open to inspection by those persons. When leaving the employment or finishing with the site a record of total radiation dose re­ceived and type has to be given to the person.

Where the examination and measurement of a dosemeter indicates the dose received differs sub­stantially from that expected, an enquiry needs to be set up to investigate and conclude the actual dose received. This may be done by investigating the area and conditions in which the individual was working, by examining the dose received by other personnel working in the vicinity and by a Duly Authorised person making an accurate assess­ment of the dose received. The measurement so produced will be substituted for that originally recorded and entered in the radiation dose register.

Similarly, when a dosemeter is lost an estimation of the dose received will be made and entered in the register. Provided the dosemeter is found within three months from the date it w’as to be processed, then the licensee may use the processed reading and substitute it Гог the reading recorded in the re­gister. After three months it is necessary for the original enquiry, which derived the estimated dose, to be re-opened and determine whether the dose — meter reading should be used or not.

• provisions are made for reports to be made to the Health and Safety Executive when either exces — sie radiation doses are received, or excessive in­gestion of a radioactive substance has occurred, or excessive contamination to a person has occurred. This reporting level will be discussed in the next section dealing with the Safety Rules.

і As a routine the licensee is required to make measurements around the site to determine the lev­els of ionising radiations, surface contamination and airborne contamination. To carry out this func­tion he must provide sufficient, readily available, well maintained and calibrated instruments for ra­diation measurements,

In addition he must make arrangements to deter­mine, in the case of a person who is occupationally exposed, the amount of radioactive substance in the person’s body and the radiation dose received. Such determination must be adequately recorded.

• No person must enter any area or remain in any area which is contaminated if he has suffered an injury which has broken his skin, unless the break or cut is covered to prevent entry into it by any radioactive substance.

In addition to any other regulations, the licensee must appoint sufficient first aiders who can deal with the treatment of wounds that are contami­nated on the site.

Within the CEGB, temperature coefficients of reacti­vity are calculated using the lattice codes ARGOSY, a transport-corrected diffusion code, or WIMS, a full transport code. These codes may be used to calculate a feedback coefficient at each point in the reactor, usually one value for each fuel element. These point coefficients are only suitable for use in a 3-dimen­sional reactor kinetics code. In order to use the co­efficients in a simpler code, some form of radial or whole-core averaging is required giving a statistical weight to each point value. The most appropriate weighting function contains terms dependent on the square of the local neutron flux, enrichment and local fuel temperature [1].

In its simplest form a reactor shutdown is effected by allowing the control rods to run into the core. Before describing additional shutdown systems, two particular features of control rod systems should be mentioned.

First, the magnox reactors with steel pressure vessels have secondary shutdown rods as described in Section

5.2.2 of this chapter. This is to give a more rapid shutdown in the event of a major depressurisation incident.

Second, the AGRs have articulated control rods, ie, they are made in several jointed sections, This feature improves the probability of the rods entering the core in the event of their paths becoming non­linear due, tor example, to distortion of above-core

components.

It has been acknowledged that certain situations could develop which could impede the free entry of control rods into the core, therefore a number of systems have been incorporated to ensure that the reactor can be effectively shut down and be main­tained subcritical in all credible situations, Some of these systems are called secondary and tertiary shut­down systems, so in this respect the control rod system may be regarded as the primary shutdown system although it is not normally known by that name.

General gas composition Primary coolant gas is ana­lysed by the Station Chemist when it is delivered to site and sample checks are made of both gas in store and in the reactor pressure envelope. This information would be made known to the operator along with any recommendation necessary. However, continuous mea­surement of the gas purity is essential to comply with the Operating Rules and moisture-in-ССЬ detectors are situated in each of the gas circuits downstream from the boilers. The measurements are displayed on a re­corder in the CCR and a high level alarm feature is incorporated. Should there be a requirement to verify the accuracy of these readings, or in the event of an instrument failure, manual sampling and checking fa­cilities are available local to the reactor. Methane indicators in the blower control room provide a mea­sure of oil ingress to the gas circuits but more accu­rate values are obtained from routine chemical analysis of the coolant gas.

Boiler leak detection A variety of techniques are avail­able for the measurement of water in CO2 resulting

from boiler leaks, and the instruments used are de­scribed in Chapter 2.

If a boiler leak occurs it causes a rise in water — in-CCb concentration and at some stations the leaking boiler is identified by reducing feedwater pressure below reactor pressure and observing any change in water-in-CO: concentration.

Mention has been made of the cross-pin temperature gradients which exist in CAGRs. In general these are such as to produce cross-pin temperature differences of up to 30°C, though the exact value varies from ring to ring and even changes direction part way up the channel. This temperature gradient is believed to be the primary cause of pin bowing (Crossland, 1982 [12]). At clad temperatures below about 650°C axial creep elfects are negligible and the main cause of pin bowing is differential thermal expansion (thermal bowing) of the pellet stack and/or the clad. Such bowing is restrained by the grid and braces which cause the pin to be stressed; it is the relaxation of these ‘■tresses v-hich makes the bow permanent.

In this type of bow the hotter side of the pin be — vomes convex. Since this reduces the cross-sectional area available for coolant on this side of the pin, the v’ltect is to further increase the cross-pin temperature difference, i. e., positive feedback occurs. Fortunately the design of the fuel is such that thermal bows, even including feedback, are very small in CAGR.

At higher clad temperatures (typically in the upper half ot the fuel stack) the main source of pin bow­ing is differential creep shortening. As already men­tioned, creep shortening is the process whereby axial creep occurs at inter-pellet gaps. If there is a cross — clad temperature difference, this process will occur more rapidly on the hotter side, so that the inter — pellet gaps will become wedge-shaped and the pin will bow. The process depends upon there being sufficient numbers of inter-pellet gaps and high enough tem­peratures to produce axial creep. These provided, the potential for pin bowing is much greater than with thermal bowing, although there is a major difference: in the case of differential creep shortening, the hotter side of the pin becomes concave so that the bowing itself tends to reduce the cross-pin temperature dif­ference, i. e., negative feedback occurs. This ensures that such bowing is self-limiting and relatively benign.

At burn-ups beyond about 10 to 15 GWd/t differ­ential fuel swelling (due to the inexorable production of solid fission products) takes over from differential creep shortening as the major source of pin bowing. This produces a small but steadily increasing pin bow which is directed outwards towards the graphite sleeve.