Fire is recognised as a potential threat to the safe op­eration of the power station and to plant operators. For this reason, the design and layout of the main plant have been established with particular emphasis on the segregation of plant by fireproof barriers or by distance, the elimination of combustible material and ignition sources wherever practicable, and the provi­sion of extensive fire detection and suppression sys­tems. The latter include fixed water and halon gas systems, supplemented by portable fire extinguishers. The fixed water system comprises three direct diesel — driven pumps and two external hydrant pumps, with distribution pipework to sprinklers and fire hose hy­drants respectively. The pumps draw from towns water reservoirs and the former set of pumps are auto­matically started on receipt of appropriate fire detec­tion signals.

The reactor primary equipment, the pressure circuit and associated systems are provided with high integrity containment, which forms the final principal barrier against release of radioactive material to the environ­ment in the event of a fault. The containment is de­signed to accommodate without failure the affects of LOCAS, steam and feed line breaks (SLB, FLB) aris­ing from equipment failures inside the reactor con­tainment building, and various other faults involving mass and energy release. The containment function is performed primarily by buildings and structures, but various systems and equipment also contribute to containment integrity and effectiveness.

The requirements which the various elements of the containment must satisfy are in three main categories:

• Control of radioactive release.

• Control of containment atmospheric conditions.

• Control of hydrogen.

These requirements are satisfied in the following manner,

Flattening of the radial rating profile is achieved in the initial core by differential enrichment in several zones, and the use of burnable poisons in place of

Fig. 3.13 PWR axial rating shapes

some fuel rods. During a cycle the assembly power ratios change as burn-up proceeds, the peak positions tending to reduce further than the average and hence partially ‘burning out’ the peaks. On refuelling, the fuel assemblies are rearranged and some new assem­blies loaded in such a way as to minimise the rating

profile.

Flux fine structure effects within an assembly are smaller than in AGRs because of the presence of moderator throughout the fuel bundle and the absence of a separate moderator.

The temperature raising phase will have been carried out at a low gas flow. Although reactor power will have increased considerably during this phase, the term ‘power raising’ is normally used for the phase in which gas flow is increased. This is carried out at nominally constant temperature, important on a magnox reactor where the number of variables must be kept to a minimum because of the effect of the moderator temperature coefficient of reactivity.

On a magnox reactor the blowers may be on pony motor drive at the start of the power raising phase. As gas flow is increased, initially by raising pony motor speed, the circulators are transferred to main motor drive at a stage appropriate to the particular station design. Gas flow is increased by the method appropriate to the particular station design, for ex — ample, by increasing motor speed where this is a variable (at Bradwell and Hinkley Point A, auxiliary turbine speed is increased; at Dungeness A and Old­bury, the blowers are steam-turbine-driven), by increas­ing blower speed at constant motor speed where a fluid coupling is fitted {Berkeley and Dungeness £). by manipulation of throttle and bypass valves at constant motor, blower speed (Trawsfynydd), by opening inlet guide vanes at constant motor/blower speed (other stations). Where the blowers are steam-driven (either directly, or indirectly via auxiliary turbine-generators), blower speed is dependent on steam generation.

As gas flow is increased the core temperatures tend to decrease (see the formula in Section 5.1 of this chapter relating power, gas now and temperature rise) therefore control rods are operated as required to ensure that reactor power is increased in step with the increase in gas How, so that reactor gas outlet temperature is maintained constant. Net rod movement is much less than in the temperature raising phase because moderator temperature changes are small. During this phase the reactor gas outlet temperature will be on auto control on those stations which have the facility, but temperature control cannot be left entirely to the regulating rods because the reactivity changes required as the xenon concentration builds up would cause the regulating rods to run outside their useful range of operating height, therefore bulk rods are used for overall control and regulating rods for local trimming. On the stations without auto control the principles are the same. Xenon build-up is rela­tively slow’, taking two days to completion, and is therefore easily allowed for.

As gas flow is increased, reactor gas inlet tempera­ture will tend to rise and is allowed for in the re­actor by control rod movement to maintain constant gas outlet temperature.

When full gas flow has been achieved the power raising phase is complete. At various stages during the start-up other checks and operations may be car­ried out. for example, checks on the BCD readings for failed fuel as power increases (fission product activity increases with neutron power), transfer from intermediate to high range flux instrumentation as appropriate, changes in protection devices as shutdown and start-up protection gives way to full-power pro­tection. The start-up will be halted and the reactor power held steady while these checks and operations are carried out. The plant and start-up procedure are designed such that these checks and operations are grouped at convenient stages, for example, the tem­perature raising phase is very demanding on the re­actor control engineer so it is arranged to proceed ‘-moothly with minimum interruptions.

On AGRs, however, although the temperature rai­ding and power raising phases are easier from the reactor point of view they are more difficult from the boiler point of view. As mentioned before, there is no clear distinction on an AGR between tempera­ture raising and power raising phases. The operating regime required on AGR boilers strongly influences the reactor start-up procedure in as much as the target rate-of-rise of temperature is determined largely by the start-up technique required on the boilers. The rate-of-rise must, of course, be within the limits im­posed by the plant and flux and temperature pro­tection as mentioned in preceding paragraphs. The turbine can also influence the start-up procedure, see Fig 3.26 which shows two routes for the start-up of Dungeness В depending on the turbine casing temperature.

Fig! 3.26 Reactor start-up at Dungeness В
The start-up procedure on an AGR is influenced by
the state of the main turbine. In the example shown,
Route A is for a turbine casing temperature below
150°C and Route D is for a turbine casing
temperature above 250°C. The temperature raising is
shown as a steady linear rise for the purpose of
showing the differences in circulator speed-raising
stages, it is not intended to imply that the actual
temperature raising proceeds in this way.

Bulk coolant gas temperatures are measured in each of the reactor gas circuits, using thermocouples sit­uated in the ducting reactor-side of the main isolating valves. These temperatures are used by the operator to help calculate the maximum operating tempera­tures as previously described. The temperatures are displayed on the low speed scanner and individual circuit temperatures can be monitored on recorders situated in the central control room (see Table 3.4). In addition the temperatures can be obtained from a 16-point recorder in the reactor annexe. This gives sufficient information to the operator so that loss of individual temperature measurement would not cause any problem.

In addition to channel gas outlet (CGO) tempera­tures, measurements are also made at other important points in the individual gas circuits. All circuit inlet and outlet temperatures are measured and displa>ed on a 16-point recorder on the reactor control desk; each circuit also can be individually selected onto an­other recorder (on the reactor control desk) to give a continuous reading during single circuit outages or reinstatement (see Table 3.4),

Reactor core graphite temperatures

The reactor core consists of graphite blocks with chan­nels for fuel elements and control rods. Regular mo­nitoring of graphite temperatures is important for the reasons given in Chapters! and 2. The maximum tem­perature allowed for each section of the graphite core is defined by the station’s ‘Operating Rules’ which are a requirement of the site licence. A large number of thermocouples are connected to the low speed tempera­ture scanner (LSS) and a printout of their readings can be obtained on demand. A selection of graphite temperatures are indicated on recorders in the annexe of the CCR and trends can easily be detected. As there are 243 graphite thermocouples per reactor, loss of any individual thermocouple or recorder would not cause significant problems for the operator. Loss of all these temperatures is not considered possible as this would require a complete failure of the instrument and safety supplies system which is a secure supply.

6.3 Magnox fuel

8.1Л Fuel cycle

Chapter 1 has outlined the fission process and has indicated the energy available during fission. A 10 kg magnox fuel element has the potential to produce about 140 MWd of energy. Half of this energy will be derived from the fission of uranium 235 and half from the fission of plutonium 239. In the process, the element will have been irradiated to the level of 14000 MWd/t.

However, as will be apparent later, it is not prac­ticable to reach this level of irradiation. Indeed, a channel average irradiation of 45% and an individual element irradiation of 60% of this level, is probably the maximum that can be obtained.

The fuel cycle, which determines the rate and man­ner in which the fuel is exchanged, must maximise the production of energy from the fuel and has, amongst others, the following objectives:

• To ensure that the fuel reaches its target irradiation.

• To keep local flux perturbations to a minimum so that the flux and coolant flow distributions are always matched, avoiding power losses due to any channels running hot or cold.

The fuel cycle has a relatively simple concept insofar as it is required to exchange the fuel continuously at the various rates, which are dictated in the most part by the target irradiation and rating of the fuel, in the different zones of the reactor.

Reactivity changes of the fuel

Reactivity may be regarded as the ability of the fuel to produce energy and this ability is not constant throughout the irradiation of the fuel. The reacti­vity life cycle of magnox fuel is illustrated by Fig 3.35 and is characteristic of all magnox fuel.

The curve shows the change in reactivity of the fuel as the irradiation of that fuel progresses. Initially, there is a small drop in reactivity brought about by the production of high cross-section fission products. This is followed by an increase in reactivity due to the production of fissile plutonium 239. The reactivity reaches a peak at about 1000 MWd/t irradiation. With further increases of irradiation, the reactivity decreases due to the burn-up of uranium 235, the fissile plutonium 239 and the production of non-fissile plutonium 240. At about 3000 MWd/t the reactivity returns to the same level as at the start of irradia­tion. Thereafter, the reactivity continues to fall as irradiation is increased, i. e., the ability of the fuel to produce energy becomes less than that associated with new fuel. All fuel within a magnox reactor behaves in this manner irrespective of its radial or axial position.

Physical and chemical properties

The limitations of uranium fuel naturally led to the~ search for a suitable substitute. A number of com­pounds of uranium have been tried in various ex­perimental reactors but the only one in widespread commercial use is uranium dioxide, UO2. Since the density of UO2 is less than that of pure uranium (ll. O compared to 18.5 g/cc), it is normal to increase the concentration of fissile atoms by enrichment. The dioxide is black/brown ceramic with a melting point of 2800°C; it has a cubic (fluorite) crystal structure and, because uranium can take any of a number of valance states, it has a variable composition capable of existing with an excess or a dearth of oxygen atoms the hyper and hypostoichiometric forms respectively. This variable stoichiometry is important in that it affects diffusion, including self-diffusion, which in turn controls such processes as sintering and creep. Chemically, UO; shows little reaction with stainless steel or zirconium cladding.

Uranium dioxide pellets are generally made from powder by cold pressing followed by sintering at high temperatures, typically 1600-1800°C. The choice of route and the presence or absence of binder material at the pressing stage are likely to affect pellet dimen­sional tolerance, grain size and density. The finished pellet is usually in the form of an approximately right cylinder with less than 5% porosity; PWR pellets have a diameter of about 9 mm whilst CAGR pel­lets are hollow with an outside diameter of 14.5 mm and a bore diameter of 5-7 mm. Most designs re­quire pellets to have slightly dished ends since this ensures that the thermal expansion of the pellet stack is controlled by their surface, rather than their centre, temperature.

Boric acid is dissolved in the reactor coolant to con­trol neutron reactivity in the core, and is referred to as a soluble neutron absorber, soluble poison, or ‘chemical shim’ (Cohen 1964 [26]). Its concentration is adjusted throughout a fuel cycle over a range 0-2500 wppm boron. The high boron level corresponds to the start of a fuel cycle and compensates for excess reactivity in the core. The level is reduced through­out the fuel cycle in response to fuel burn-up and changing core reactivity, temperature and the build up of other poisons such as xenon and samarium. Typical boron concentrations at identified reactor con­ditions are given in Table 1.19.

It should be noted that other soluble neutron ab­sorbers have been considered such as gadolinium or cadmium nitrate. Such elements have large neutron capture cross-sections and could therefore be effective at low concentrations. However there are potential disadvantages in terms of solubility and hydrolysis reactions, and current practice is entirely with boric

d The use of boric acid has the advantages that sufficiently soluble in water to yield the required concentrations, it has sufficient chemicaland physical. ability over the required temperature range, and it has a low propensity for incorporation into oxide films which could result in local concentrations of

neutron poison and acidity.

The high purity boric acid used will consist of nat­ural boron which comprises the two isotopes boron 10 (19 6%), with a neutron capture cross-section of 4000 barns and boron И (80.4%) with a neutron capture cross-section of 0.05 barns. This implies that boric 10 acid levels could be reduced by a factor of up to 5 if enriched boron -10 boric acid was used.

The use of boric acid to control reactivity in this manner is therefore a requirement of reactor op­eration, and the main disadvantage is that it imposes additional difficulties in pH control and purification by ion exchange. Boric acid is a weak acid, and al­though at typical PWR concentrations the acidity decreases with increasing temperature, there is an interaction with the control of pH.

Boric acid H3BO3 is known to form polyborate ions in the presence of hydroxyl ions by the follow­ing equilibria:

H3BO3 + OH — <-► B(OH)4-

2H3BO3 + OH” <-► B2(OH)7-

ЗН3ВО3 + OH" «-> B3(OH)i0"

where only the monomer, dimer and trimer are sig­nificant.

In a study of the boric acid dissociation over tem­peratures relevant to a PWR, Mesmer et al (1971) [27] determined the equilibrium constants and showed that the first reaction of H3BO3 was the significant determinant of pH in primary coolant. This is due 10 the decreasing importance of the polyborates at high temperature, low boron concentration and low hydroxide concentration.

In the presence of LiOH hydrated species of lit­hium metaborate (LiB02) are formed which have a negative solubility coefficient (decreasing solubility) with increasing temperature.

With successive magnox designs, reductions in capital and operating costs were achieved by increasing the reactor sizes and coolant gas pressures. As stated pre­viously, there is a limit to the reactor size and coolant gas pressure if steel pressure vessels are used to contain the core.

Safety could be improved if the boilers were con­tained in the same pressure vessel as the core, thereby eliminating the gas ducts, and if an alternative could be found to the steel pressure vessels. Steel pressure vessels have a single load path to withstand the gas pressure and care has to be taken in the design, selec­tion of material, welding, inspection and testing to ensure that possible failure by fast fracture can be discounted.

Towards the end of the magnox station programme, prestressed concrete pressure vessels were being de­signed and models built and tested. For the last two magnox stations at Oldbury and Wyifa, prestressed concrete pressure vessels were used to contain the core and boilers (Table 2.1).

The Oldbury vessels are cylindrical in shape, with a top and bottom slab, whilst the Wyifa vessels have a spherical internal shape.

The boilers are arranged round the core and are protected against irradiation from the core by a boiler shield wall. The vessel walls contain a large number of penetrations of differing diameters; for fuelling the core, for control rods, gas circulators, for feedwater and steam to and from the boilers, for safety valves and
connections to the auxiliary gas circuits, coolant gas make-up and gas blowdown systems (Fig 2.11).

The detectors and cables discussed in this section operate with electronic amplifiers and ratemeters as diown in Fig 2.49.

Fu; 2.49 Tjpieal pulse counter channel

The pulses due to neutrons at the detector are unall and are usually amplified by a head amplifier mounted near the detector, though the distance can be up to 80 m in AGRs. The output of the head amplifier is led to the main pulse amplifier, pulse height discriminator and ratemeter. This produces a IX’ signal which drives indicators and chart recorders and operates alarm and trip units. The complete chain ь show n in Fig 2,49,

The neutrons produce larger pulses at the detector [han rays and so the effects of 9 rays can be reduced by counting pulses only above a certain lev­el, using a pulse height discriminator. The gain of the pulse amplifier is adjusted to achieve the best discrimination.

With large у ray levels the number of pulses due to 7 rays is large and since they can become super­imposed, their height can add up to a level compara­ble to that due to neutrons. This effect is known as ‘gamma pile-up’ and is reduced by modifying the pulse shape. For example, with a typical AGR sys­tem integrate and differentiate time constants of 40 ns are employed in a wide bandwidth system, typically up to 20 MHz.

In order to cover the low power levels and give good overlap with the logarithmic DC channel, a wide range is required and it is convenient to have a logarithmic scale. Typically the usable range is Ю to 106 pulses/s. This characteristic is obtained by a log­arithmic diode pump circuit, which produces a DC voltage proportional to the logarithm of the pulse rate. This DC voltage is amplified to provide an output to drive indicators, chart recorders and operate alarm and trip units.

In addition the DC output is fed into a differ­entiating circuit and, in a similar fashion to that em­ployed in the ion chamber period meter discussed in the next section, provides a doubling time signal which drives indicators, control and alarm signals. Further details are given in Table 2.6.

A typical pulse counting channel measures pulse rates from fission or BFj chambers in the range of 10 pps to 106 pps. In the Heysham 2 AGR power station, a P8 fission chamber feeds pulses through about 20 m of mineral insulated cable to a head amplifier. The output of the head amplifier at 50 ohms impedance is fed through about 100 m of flexi­ble superscreened coaxial cable to the main pulse amplifier.

The channel measures pulse rate on a 5-decade logarithmic scale and provides outputs to operate a front panel meter and external trip and recording equipment. The log output voltage is differentiated internally to provide an output which is proportional to the doubling time of the reactor power. A front panel meter indicates doubling time (DT) and DT outputs are provided to operate external equipment.

The head amplifier is housed in a sealed metal can and the remainder of the channel is in modular form mounted in a 483 mm chassis. Trip circuits are pro­vided in an external unit.