The dominant production route of Xe-135 is as a daughter product of a radioactive decay chain starting with Те-135 as follows:

Te-135 -> 1-135 — Xe-135 -* Cs-135 — Ba-135(stable)
2 min 6.7 h 9.2 h 2 x 104 у

As the half-life of Te-135 is very short it may be neglected. Iodine has a low neutron absorption cross — section and its concentration is therefore determined

The xenon concentration at equilibrium can be found from Equations (3.7) and (3.8) by putting 6N/ot and 5N[/5t equal to zero. This gives:

X + ] cra(E)o(E)6E

It is seen from Equation (3.9) that the saturation level depends on neutron flux via the fission rate term on the top and the neutron capture in xenon term on the bottom of this expression. As the flux is in­creased the saturation level increases until the neutron capture term dominates the radioactive decay term X. Neutron flux levels differ in the different thermal reactor systems, for example, typical total fluxes in magnox, AGR and PWR are 0.5, 1.7, 3.0 in units of 101J neutrons/crrrs, and the saturation xenon con­centration for operation is the reactivity effect of the xenon which depends on the neutron absorption rate in xenon relative to other materials. The resulting differences in xenon reactivity-worth between reactor systems is much less, typical figures for magnox, AGR and PWR for full power xenon-worth being 1.7, 2.1 and 2.5 Niles.

In order to provide sufficient reactivity control to accommodate the change in reactivity due to xenon build-up a coarse control system is required. This is provided in magnox and AGR reactors by bulk rods in banks. The use of partially-inserted bulk rods during reactor start-up causes distortion of the axial and radial flux shape. These distortions may prevent full reactor power output being achieved because of the need to stay within operating rules limits on can temperature. In PWR the xenon reactivity poisoning is counteracted by reduction of boron level in the water moderator which provides a uniform reactivity control and therefore does not itself cause significant flux distortions. The distribution of xenon throughout the reactor is not uniform since it depends on the neutron flux distribution; hence changes in reactor power level can lead, via xenon level distribution changes, to flux distortion, particularly in the axial direction. These distortions are controlled in the PWR by balancing movements of control assembly banks against boron concentration control.

When a reactor has been operating at power for some time and is then shut down, the removal term in Equation (3.8) relating to neutron absorption in xenon goes to zero but the production term, XiN;, due to radioactive decay of iodine, does not imme­diately reduce and the net result is an increase in xenon level. A peak in xenon level is reached after about ten hours and then decays away over two or three days as shown in Fig 3.3.

In order to override the xenon build-up extra re­activity is needed. Without it the reactor may not be capable of achieving criticality and starting up dur­ing the few days immediately following a reactor shutdow n.

Fic. 3.3 Xenon and iodine concentrations following
shutdown from equilibrium conditions

The effectiveness of a control rod varies with the relative neutron flux in which the rod is operating; ‘relative’ in this case means relative to the reactor as a whole, i. e., moving a given rod up or down in a high relative neutron flux has more effect than in a low relative neutron flux. The variation is manifested in a number of ways.

First, and most important, is the variation of ef­fectiveness with rod insertion. The neutron flux is higher in the middle region of the core than it is at the top and bottom, therefore moving a control rod a given amount around mid-insertion will give rise to a larger change in reactivity than moving it the same amount around the top or bottom of its travel. This variation in effectiveness is shown in curve 1 of Fig 3.23.

Second, the shape of the curve of rod worth versus rod position depends on the axial flux shape in which the rods are working. The slope of the curve at a given rod position is related to the relative neutron flux at that axial position in the core — the greater the neutron flux the greater the slope. Thus if the axial flux shape is distorted from the natural cosine­like shape, for instance by virtue of a group of bulk rods being inserted halfway into the core as may be the case shortly after start-up, then the curve of rod worth versus rod position will be as shown in curve 2 of Fig 3.23.

Third, if the neutron flux is uniformly affected (relative to the reactor as a whole) in the region in which our control rod is operating, i. e., the magni­tude of the flux is modified but it retains its natural cosine-like shape, then the rod worth will be modified. In the example shown in curve 3 of Fig 3.23 a group oi bulk rods is fully inserted in the core as may be the case shortly after start-up of a magnox reactor; the neutron tlux is depressed in the immediate vici­nity of the inserted bulk rods but the average flux across the core is unchanged (because reactor power

is unchanged) so the neutron flux in the regions in which the sector rods are working is higher than the average, therefore their worth is increased. Similarly the worth of an individual rod is influenced by con­ditions in its immediate vicinity, for example, if a sector rod is close to a bulk rod which is fully in­serted, or if it is close to a channel of absorber (see magnox fuel cycle in Section 8 of this chapter), then the worth of that sector rod will be reduced because it is working in a lower-than-average relative neutron flux. Figure 3.23 is plotted for the group of 28 sector rods at Dungeness A.

In practice the second and third effects described here are barely noticeable to the reactor control en­gineer amongst the many other parameters affecting the reactor performance, but they are taken into account in theoretical analyses for reactor optimisa­tion and fault studies. The first effect, however, the basic S-shape, is very apparent and important to him.

As mentioned in the last paragraph, temperature co­efficients of reactivity of the moderator in a magnox reactor and the fuel outer sieeve in an AGR become positive as fuel irradiation proceeds. The bulk mod­erator temperature coefficient of reactivity in an AGR is positive from start of life. Let us consider these three positive feedback effects together. As tempera­ture rises, this gives a positive change in reactivity which tends to increase the neutron power which in turn increases temperatures, and so on. The effect in the bulk moderator will be much less pronounced in an AGR than in a magnox reactor because moderator temperature is maintained largely constant by the re­entrant gas flow, but the principle is the same. This is a destabilising effect, causing the power to increase further. These effects have a time constant of 10-20 minutes.

A second destabilising effect is that of Xe-135, As reactor power increases, xenon is destroyed faster than it is created so the concentration decreases. Since xenon is an absorber of neutrons this results in a positive change in reactivity which tends to increase reactor power further. This initially reinforces the ef­fect of the graphite temperature coefficients mentioned in the previous paragraph.

A burst cartridge (or can) detection system is neces­sary to detect any defect in the cladding of a nuclear fuel element which is leaching radioactive material into the coolant gas. Radioactive isotopes in the coolant gas will be deposited over the reactor internals and the boilers and boiler voids which will give two main problems:

• Access to boilers is limited because of high radia­tion levels preventing easy access for maintenance.

• Blowdown of pressure vessels and boilers will be limited bv gas containing radioactive material.

The most serious problem associated with a burst fuel element in a magnox reactor can, is that uranium metal will quickly oxidise and there is a high risk that rapid swelling will occur which will restrict the coolant gas flow up the fuel channel. Swelling of fuei will quickly block the channel and unless action is taken immediately a channel fire could result.

With AGR fuel the principal risk arises from con­tamination of the circuit and boiler.

Burst cartridge detection equipment Burst cartridge detection equipment detects the amount of radioactivity in a sample of gas drawn from a fuel channel, a group of channels or even the general bulk coolant stream. The principle of operation is that decay products from radioactive gas in the sample are attracted to an electrostatically-charged wire which passes through a scintillation counter where the radio­activity is measured. The level of activity is counted on a ratemeter and indicated to the operator as a number. The number displayed is used relatively as each counting period ends.

The series of readings are fed to a computer where it is stored for historical display providing differential relative values against previous readings and compared with alarm settings to make the operator aware that a faulty element may exist.

Normally a number of systems exist for BCD moni­toring and each serves a different purpose. These are listed as follows:

• A system for interrogating each channel, which may sample each channel in turn or which may have to be preset to an identified channel. The routine sampling of each channel may take several hours and is therefore a slow system.

• A system for interrogating a group of between 4-16 channels which will survey the whole reactor in minutes.

• A bulk system which monitors the bulk coolant on

a continuous basis.

The slow system is primarily used for detecting very small leaks to give warning that a more serious burst may be developing, and for monitoring a single chan­nel during fuelling operations to look for damaged fuel that may be in the process of being loaded.

The fast system, which monitors groups of chan­nels, gives information to the operator of impending fast fuel cladding failure, although it will not identify the culprit channel.

The bulk gas sample gives information on fast fail­ures and any indication here must be serious because the gas sample only contains heavily diluted activity from an affected channel.

The BWR differs from the PWR in two very important respects:

• The water is allowed to boil as it flows through the core.

• The steam thus evolved passes directly to the turbine; that is, the BWR is a direct-cycle design and the need for intermediate steam generators is eli­minated (Table 1.8 and Fig 1.29).

It follows that the coolant pressure is only slightly higher than that required at the turbine, typically about 80 bar, and the temperature 300°C. The lower coolant pressure means a lower core power density (about 50 MW/m3) and is generally about half that for the PWR. Hence the core volume of a BWR is correspondingly larger, being about double that of a PWR to give the same thermal output.

The BWR pressure vessel however is more than double that of the PWR because of the additional need to accommodate steam separators and dryers above the core and a number of jet pumps located in the annulus between the core and the vessel wall. The steam separators take out most of the water phase, allowing it to return for recirculation via an annular downcomer which surrounds the core. The steam, which leaves the separators with a low mois­ture content still entrained, passes through mesh type dryers at the top of the pressure vessel before leaving, >99.5% dry, to drive the turbine. The jet pumps, numbering twenty or more, form part of the coolant recirculation system. This system normally consists of two independent loops, each having a recirculation pump and associated control valves located external to the pressure vessel. Although the pressure vessel for a BWR is appreciably larger than for the PWR, the lower operating pressure results in a wall thick­ness about half that of the PWR vessel.

The turbine steam is radioactive, mainly due to the radionuclide N-I6 formed from an (n, p) reaction with the 0-16 in the water, requiring the high pressure end of the turbine to be shielded. Fortunately, N 16 has a short half life of 7.1 seconds and maintenance work during shutdown is not unduly hampered as the activity soon decays away.

The BWR fuel rod is similar to that of the PWR but it can tolerate a larger diameter because of the lower power density. A fuel bundle is based on an 8 x 8 array. The whole is encased in a square zircaloy channel to prevent lateral flow between adjacent as­semblies, in contrast to the open array of the PWR fuel assembly.

Control rods are cruciform in shape and move in the interspace between four fuel assemblies. They are driven hydraulically from below the core and thus counter neutron flux distortion due to steam voids in the upper region of the core. Also, as refuelling is off-load (burn-up values approaching 30 000 MWd/t) and requires access to the core from above, bottom entry for the control rods does not interfere with the annual refuelling. Reactivity is also regulated by varying the coolant flow through the core and hence the value of the void fraction as the degree of boiling is increased or decreased.

BWRs are generally housed in primary and second­ary buildings. The former, termed the ‘drywell’, is a steel pressure vessel surrounded by reinforced concrete and is designed to withstand LOCA pressure transients. The secondary containment building, termed the ‘wet — well’, completely encompasses the drywell. The base of the wet well contains a pool of water connected to the drywell via large submerged ducts. In the event of a LOCA, high pressure steam in the drywell is directed into the pool where it condenses. For this reason the wetwell is sometimes called the pressure suppres­sion chamber. The pool of water is also available for use by the Emergency Core Cooling System (ECCS).

Effective weight loss

A full assessment of core behaviour is extremely com­plex involving the many parameters already described. In some circumstances, such as coolant optimisation (see Section 10.3 of this chapter), a simpler parameter is required related to core integrity. A major change occurring during core life is the reduction in reserve factors caused by loss of strength due to radiolytic oxidation.

As described in Section Ю.4 of this chapter, changes in gas composition through the brick result in com­plex weight loss profiles through the brick, high carbon monoxide concentrations giving flatter profiles than low carbon monoxide concentrations. A series of ex­perimental strength measurements, together with two and three-dimensional finite element strength calcula­tions were carried out to determine the major para­meters determining brick strength.

The basic strength of a 25 mm brick slice was measured by applying a load via four keyways in a plane perpendicular to the brick axis. Relative loss of strength was measured on slices profiled to repre­sent various weight loss distributions, the reduction in thickness modelling the loss of modulus across the brick and the strength at the point of failure. All of the test slices failed by crack propagation from the root of a keyway generally propagating to the inner channel wall but in a few cases the failure was from key way to key way. The two — and three-dimensional calculations were carried out using the finite element program ‘BERSAFE’ with loadings as used in the brick slice tests. The two-dimensional mesh was suc­cessively refined to give a minimum radius at the keyway corner of l mm as this is the maximum notch radius which has a stress raising effect in graphite.

The two-dimensional mesh was ‘grown’ into the third dimension such that the thickness at any point was proportional to the modulus at that point. Both series of calculations showed that there was a stress concentration around the keyway root in agreement with the site of failure in the slice tests.

The relative failure loads for the three sets of data were plotted against various weight loss functions to determine any correlation with the strength loss re­lationship described in Section 10.4 of this chapter. It was found that the loss of strength can be deter­mined for all the weight loss profiles from the effective weight loss (EW) as given by: EW = 1/2 (mean weight loss — t — keyway root weight loss).

After breakdown of irradiated AGR stringers in the irradiated fuel dismantling cell, individual elements are discharged to a cooling pond, where they are stored vertically in compartmented skips to allow cooling and radioactive decay before transport off site.

To provide criticality control, the pond water con­sists of demineralised water dosed with boron in the form of boric acid at a level of 1250 ppm boron, the pH being adjusted to 7 with caustic soda. Chillers are provided to maintain pond water temperatures of approximately 25°C.

In anticipation of release of activation products from fuel cladding, especially from the clad oxide, ion exchange plant is provided to remove active cations such as Co-60, and also to provide some measure of chloride-ion control to meet the specification of a target value of 0.5 g/m3 (upper limit 10/m3) to sup­press corrosion. For sulphate ion a target level of <0.5 g/m3 is set. In practice to date (1988), the graphite sleeves rather than fuel cladding have been found to be the major source of activation products. The activation products form in the graphite itself, and also in originally inactive material deposited from the reactor coolant on the large graphite surface area. Caesium 134 is among the activation products found

and this element is not removed by the cation ex­change resin to any extent. However, caesium selective resins are available if required. Fuel pin failure is not expected in station ponds where storage periods are of limited duration. However, activity release from deliberately defected pins has been examined and is expected to be slow, and limited in amount to the cap inventory (i. e., the amount released from the fuel pellets to the pin interior during service). The fuel does not react significantly with water.

Slight visual signs of corrosion are observed on elements where the steel components have been irra­diated in a temperature range conducive to sensitisa­tion. For example, parts of the brace which include welded strip sections, are affected and occasionally pin end caps. Metallography has shown slight inter­granular corrosion after typical pond residence times of about 250 days but pin and brace integrity has not been significantly impaired.

At this early stage of the AGR programme, ex­perience of fuel behaviour in pond storage [45] is necessarily limited and inspection of fuel behaviour both in pond by underwater techniques and by de­structive examination of a limited number of pond — stored elements is planned to continue.

MAGNOX REACTOR

The main auxiliary supply required is electrical power to drive gas circulators, feed pumps (where not steam turbine driven), CW pumps ventilating fans and other motor driven systems, electric actuators (including control rod drives), instrumentation control systems and main and emergency lighting. As with conven­tional stations, an auxiliary electrical system uses sup­plies from the station busbars and the generator (unit) busbars which are fed via transformers to 11 kV (or

6.6 kV), 3.3 kV and 415 V busbars. Large loads are fed from high voltage busbars and smaller loads from lower voltage, as this gives a more economical motor and cable design. Some magnox stations used separate small turbine-generators to supply the gas circulators, as turbine governor control could be used to vary the gas circulator speed and gas flow for low load opera­tion, but in practice the reactors have been operated at maximum output.

Some electrical supplies are required under fault and post-trip conditions to ensure safety, to remove reactor shutdown heat and to maintain control, in­strumentation and lighting services. These essential supply systems are described in the next section.

Other auxiliary supplies are also required. Cooling water is required on certain plant and this is supplied by electric pumps from a reservoir (or from the sea). Air supplies may also be required, e. g., pneumatic operation of valves or actuators, for which diesel or electric compressors may be used. In all cases, ade­quate redundancy must be designed into the systems in a similar way to that required on electrical sys­tems. The supply systems should be operated in sepa­rate sections with an adequate number of pumps or compressors to cover maintenance and failures. Causes of common mode failure must be removed. It may

be practicable to provide local storage and so reduce the urgency of restoring supplies after a fault.

6.1 Layout and radiological protection

As understanding and confidence grew with experi­ence, the widely separated reactors, turbine hall and auxiliary buildings exemplified by Hinkley Point A were replaced at the later magnox stations in the interests of capital economy and ease and efficiency of operation by a closely-integrated compact layout. This concept was also adopted for the AGR. AGRs also use prestressed reinforced concrete pressure vessels with boilers enclosed in an annular space protected by an internal shield as developed for the later mag­nox stations (Oldbury and Wylfa).

The basic radiological problems and needs of mag­nox reactors apply also to AGRs but there are some differences. It was deemed necessary to provide shield — mg over the reactor core to permit access to the ■ op dome area of the gas baffle. It is also necessary ■o dismantle irradiated fuel stringers and to carry out routine maintenance on fuel assembly components such as the plug units. During operation of the pro­totype AGR at Windscale, extensive deposits of active dust were discovered in the boilers. It was thought that a similar problem would arise in the CEGB re­actors and provisions were made to counter this. These included the installation of dust collectors (central inertial collectors) at the outlet of some reactor chan­nels and means for handling and storage of full col­lectors. Fairly elaborate facilities for handling and decontaminating gas circulators having significant de­posits of activity have been provided, but the active dust problem has not yet manifested itself.

Figure 2.1 shows the difference between the AGR В station and the magnox A station site arrangement at Hinkley Point. With the exception of the gas cir­culator workshop, the emergency generators (gas tur­bines) and waste incinerators, all ancillary services for the AGR reactors are inside the main building. This obviated the need for a controlled area fence and internal road system. The A station laundry was ex­tended to deal with the В station.

Although unique, it is worth noting the design of the boiler shield wall which comprises three layers of steel plate separated by steel tubes 89 mm diameter filled with calcium hydroxide. All the other AGRs, except for the ‘pod’ boiler arrangement at Hartlepool and Heysham 1, employ steel and graphite for this purpose.

Figures 2.70 and 2.71 show the disposition of most of the ancillary facilities in the Hinkley Point В sta­tion. The integral building includes a central shielded block^ between the two reactors and a reactor services building between the reactors and the turbine hall. The irradiated fuel cooling pond on the centre line of the building extends into an annexe beyond the reactor building which also houses the pond water and active effluent treatment plant. The central shielded block includes fuel handling and maintenance fa­cilities, waste stores, radiochemistry laboratory, CO2 coolant treatment plant and plant decontamination centre. The reactor services building includes the sta­tion control room and instrument rooms, main change room and ventilation plant. Although the control room is outside the controlled reactor area there is an emer­gency access route from the control room to the charge hall.

There are three extract ventilation plant rooms each with a separate discharge stack. There is one for each sub-pile cap interspace and one for the reactor ser­vices building and central shielded block.

Filtration comprises absolute filters with pre-filters. The areas covered include combustible active waste sorting, waste stores, radiochemistry laboratory, CO2 treatment plant, plant decontamination centre, reactor area workshop, charge machine maintenance, gas cir­culator penetrations, irradiated fuel handling, new fuel assembly and sub-change rooms. There is a separate

system for the cooling pond and effluent treatment plant areas.

Active waste stores are as follows;

• Irradiated fuel debris vault for reflector sleeves, tie rods and central inertial collectors,

• Maintenance cell debris vault for control rods and chains, tie rods and flux measuring heads. [19]

materials which are too active to burn, e. g., con­taminated clothing, filters.

• A ‘transient’ store for combustible waste.

• A ‘permanent’ store for combustible waste.

• A store for spent desiccant.

• Sludge and resin tanks.

• CO2 blowdown filter tank.

Within the main building there are four permanent sub-change rooms:

• One for the pond water and active effluent treat­ment plant, cooling pond and irradiated fuel trans­port flask area.

• One for the charge hall and contaminated ventila­tion plant extract rooms.

• One for the plant decontamination plant centre, reactor area workshop, charge machine service well and maintenance area.

• One for the new fuel assembly room and low-er new fuel assembly cell (primarily to maintain clean conditions).

Routes between the sub-change rooms and main change room can be isolated to prevent the spread of con­tamination.

Separate personnel control and change facilities are provided for the external circulator workshop and com­bustible waste sorting and incinerator facility.

For access to the pressure vessel there is a spe­cial penetration in the pile cap over which a tent is erected. Suction pumps discharge air from the ves­sel through the CO2 blowdown filters to ensure that no contamination escapes to the environment. The tent operates as an inlet filter to protect the core. There is a permanent gallery round the gas baffle from which access can be gained to permanent lad­ders between the boiler casings for inspection and maintenance. A system for the supply of air for pro­tective suit cooling and breathing from special com­pressors is provided.

Each station has a storage pond with a boronated water depth of some 7 metres. The pond water is circulated through filtering, cooling and water treat­ment plant with provision for the disposal of filter and resin arisings. There are two bays where the fuel is stored in compartmemed boron steel skips for up to 100 days decay before dispatch. Typically, the two sloping discharge tubes in the IFDF each connect to one bay. The fuel element/fuel bottle arrives in a reception tube which is then tilted to the vertical, so that it can be removed and handled into the ap­propriate skip by a manipulator which bridges the pond. When the skip is full it is parked, and after the decay period the skip is removed to the dispatch bay.

The 50 tonne transport flask after cleaning of road dirt and removing the lid bolts is lowered into the

Nuclear power station design

F;c. 2.112 Irradiated fuel disposal facility at Heysham 2

pond. The lid is removed for maintenance of the lid seal — The empty skip in the flask is exchanged for the dispatch skip. The lid is replaced and the flask is removed for washing, decontamination, water level adjustment, refitting of lid bolts, leak testing, etc before dispatch by road transporter to the rail head en route to Sellafield for re-processing. The Hevsham 2 arrangements are somewhat different in that the Hask is not placed in the pond, instead, the fuel skip is raised from the pond and placed in the flask. This operation takes place immediately adjacent to the pond in a shielded ceil which the flask enters on a trolley through a sliding shield door.