2.3.1 Basic requirements

Many of the requirements are associated with fuel element temperature assessment but a large number of measurements are made of the temperatures of reactor graphite, the pressure vessel and parts of the boiler in addition to measurements of auxiliary plant operation.

Another set of temperature measurements is re­quired for connection to the temperature trip units in the reactor safety circuit discussed in Section 5.4 of this chapter.

The measurements are normally made using indus­trial-type, metal-sheathed, mineral-insulated thermo­couples manufactured generally to BSl041:l943, with batch accuracies of ±4°C up to 500°C. The thermo-

couples are supplied in accordance with the CEGB specification for reactor instrumentation, and illus­trated in Fig 2.55; further details are given in Volume F, Chapter 4. Stainless steel sheathed chromel/alumel, Type К thermocouples are used mainly in the reactor and on the reactor vessel structure, iron/constantan Type J thermocouples being used in the concrete biological shield. In a power reactor there may be several thousand thermocouples.

To maintain the reactor coolant compositions within predetermined tolerance levels, bypass gas processing plant comprising filters, a recombination unit, a dry­ing plant and associated control systems, pipework

and valves is required. The bypass system is connected

to:

• The carbon dioxide storage and distribution system for reactor filling and make-up.

• The methane store or production plant for coolant chemistry control.

• The reactor blowdown system.

Just under one per cent of the total reactor gas flow is diverted through the gas bypass plant. A schematic arrangement of a typical bypass plant is shown in Fig 2.87. Flow to the bypass plant is taken from the gas circulator outlet. The gas first passes through a filter to remove any particulate matter which could contaminate the plant downstream, then to the re­combination unit and drier unit, and finally through another filter to remove any fine desiccant from the drier towers before returning to the reactor via an­other penetration in the pressure vessel wall. The pene-

trations in the vessel wall for the inlet to and outlet from the bypass plant are fitted with either venturis or orifices to restrict the maximum discharge area in the event of the rupture of the bypass plant pressure boundary. Isolating valves are provided in the circuit. To support the plant, carbon dioxide storage, meth­ane storage or production, and oxygen production plant are necessary.

The earlier AGRs were not designed to the ‘CEGB Design Safety Guidelines’, but as this document is largely a statement of requirements which arose out of the Board’s experiences with these stations, there are many similarities in the design of the post-trip heat removal, systems. None of them, however, have the extensive provisions made at Heysham 2.

Hinkley Point В is the basis of the station design for Heysham 2 and the systems for post-trip heat removal are similar, but without the extensive redun­dancy and segregation. The main heat removal route is via the main boilers, with steam venting to atmos­phere. The boilers are fed either by the two starting and standby boiler feed pumps or a station-based emergency boiler feed system comprising three 10% pumps drawing from the reserve feedwater tanks. The gas circulators only run at full speed, thus only one of the two circulators is normally operated in each quadrant, with the inlet guide vanes partially closed. Automatic sequencing equipment is provided to estab­lish post-trip cooling and organise the electrical sup­plies, which are provided from a conventional station and unit boards system, backed-up by four 17.5 MW gas turbines for emergency duties. The decay heat boilers and the feed system are introduced manually about 20 minutes after the reactor trip; a single feed and condenser system serves both reactors.

Hartlepool and Heysham I

These two stations are quite different in concept to Hevsham 2 and do not have the benefit of decay heat boilers. The main post-trip heat removal route ■ ^ja the main boilers, fed either by the single start — in(T ancj standbv feed pump or two 8% emergency boiler feed pumps, these latter mav draw water from the de-aerator, the RFTs or the condensate polish — jne plant outlet. The steam is dumped into a single dump condenser which serves both reactors. The gas circulators are driven at low speed through pony motors. Automatic sequencing equipment establishes these post-trip heat removal systems and the essential electrical system. The latter comprises two unit trans­formers supplying two 11 kV boards, with a generator switch to enable them to remain energised with the unit shutdown; emergency power is provided by four

17.5 MW gas turbines. In addition there is an entirely separate back-up cooling system which is initiated solely by operator action; this is to cater for certain hazard conditions which may result in total loss of the normal systems. The system comprises two HP diesel-driven pumps supplied from RFTs to provide feed to the main boilers, and two LP diesel-driven pumps drawing from towns water reservoirs to supply the primary sides of the PVCS coolers and reactor auxiliaries cooling system coolers; gas circulation is achieved through natural convection.

Dungeness В

Dungeness В again is without a decay heat boiler sys­tem. The main boilers fed by either the single starting and standby feed pump, a 15% steam-driven emer­gency boiler feed pump or two 2,5% electrically-driven emergency boiler feed pumps, all drawing from the de-aerator. The gas circulators, operating at low speed through a pony motor drive, provide the main post­trip heat removal route. The electrical system is a conventional station and unit transformer system, with lour 3.5 MW diesel generators suppling 3.3 kV station boards. Again an entirely independent back-up cool­ing system is provided, similar in concept to Hartle­pool. However, the Dungeness В reactors do not have sufficient natural convection capability, so three diesel generators and associated distribution boards are provided to supply the gas circulator pony motor drives and three electrically-driven HP and three LP emergency cooling water pumps.

A position detector is supplied for each control rod cluster drive rod. Each rod position detector consists of a tube, with 42 detector coil assemblies mounted on the tube and spaced out along its length at 95 mm intervals. The tube fits over the rod travel housing. Power at 12 V AC from a constant voltage source is applied to the coils. The electromagnetic flux generated by the coils penetrates the thickness of the drive rod travel housing. Since the rod within the housing is ferromagnetic, the magnetic flux rises steeply as the rod moves through a coil. The concentration of flux results in an increased impedance to the AC current flow in the coil. To prevent total loss of position in­formation due to a single coil failure, the outputs of every other coil on the tube are connected as coil set X, while the outputs of the remaining coils are connected as coil set Y.

The detector electronics samples the voltage at the junctions of coils in each set. When the rod is within either both or neither of the coils in the pair, the voltage sampled is ‘high’.

When the rod is within only one of the paired coils the sampled voltage is Mow’. A differential amplifier compares with voltage samples for each coil pair with a fixed reference, and provides an output which can be digitally encoded to represent each rod position.

11.1.4 Thermal power measurements

The N-16 power monitoring instrumentation provides an input into the station automatic control system and the primary protection system (PPS) for calculation of the thermal power.

this instrument monitors the thermal power ot the reactor by detecting the level of N-16 present in the coolant system. N-16 is an isotope of nitrogen, venerated by neutron activation of oxygen contained in the water. The level of N-16 present in the primary

oiant is directly proportional to the fission rate (and hence power) in the core. Decay of the N-16 isotope produces high energy gamma rays which penetrate the wall of the high pressure piping. Therefore the fsl-16 concentration in the primary coolant can be monitored by measuring the characteristic gamma ra­diation outside the primary coolant piping.

N-16 gamma radiation is monitored by two ion chambers located on the hot leg piping of each cool­ant loop. The ion chambers are located as close as physically possible to the reactor vessel but outside the biological shield, and shielded from reactor neutrons.

The two N-16 detector signals are processed and corrected for N-16 decay to give an average signal. The transit time around the loop (about 11 seconds) and the N-16 half-life result in a saturation signal, giving a computed signal which varies directly with reactor power.

The Troid signal is used to compensate the N-16 signal for fluid density variation and the final signal of thermal power is computed within the PPS, for use (together with system pressure) in the departure from nucleate boiling (DNB) and linear power density (kW/m) calculators of the PPS,

When any plant is put to work it generates the re­quirement for maintenance. On a nuclear site, main­tenance is considered in two separate philosophies, these are:

• The commercial risk in not doing it, e. g., loss of generation from failure of plant.

• Maintenance which is part of ensuring that plant operates safely (safety of the public and persons engaged on the site).

In the first item a programme of maintenance has to be designed to ensure that the risk of failure is acceptable. Advantage can be taken of the provision of standby plant where this exists to do regular main­tenance, and most major plant auxiliaries are dupli­cated. The problems occur with those major equipment items, such as the main turbine-generators and gas circulators, that need the unit shut down for access to carry out the repairs and maintenance.

The requirement under conditions of the site li­cence to shut the reactor down for inspection purposes within a maximum period of two years from the time it was previously given consent to start-up, provides a window for the overhaul and maintenance of these major items. Unfortunately, this gives a very ‘peaky’ resource requirement for maintenance during this win­dow of about three months. Most nuclear sites keep the maintenance in-house rather than using contract labour. The reason for this is that the work requires specialist knowledge and expertise coupled with (in many instances) a good knowledge of health physics procedures and practices.

In a reactor with a negative fuel (fast) temperature coefficient any rise in fuel temperature is accom­panied by a drop in reactivity. This tends to reduce the rate of rise of temperature thus stabilising the reactor. There will be a consequent rise in moderator temperature but the time scale for such changes in the graphite is long and the de-stabilising effect of the positive moderator is easily dealt with.

2.3.1 Positive fuel temperature coefficient

If a reactor were to have a positive fuel feedback co­efficient it would be highly unstable to temperature perturbations. It would be possible to control such a reactor only with a highly sophisticated, fast-acting, auto-control system. It is important, therefore, to main­tain the fuel coefficient at a suitably large, negative, value.

Changing a fuel stringer on an AGR has a much greater effect on core reactivity than changing a chan­nel of fuel on a magnox reactor, partly because it represents a much larger fraction of the total core fuel inventory and partly because of the much greater reactivity difference between start of life and discharge. The effect on an AGR core is about 60-90 mN per stringer, compared with about 0.5 mN per channel in a magnox reactor.

As the irradiated stringer is discharged the adja­cent fuel channels run hotter, as in a magnox reactor, so the regulating rods will be driven in by the auto control system to maintain average channel gas outlet temperatures constant. As the new stringer is loaded, the regulating rods will be driven in further to main­tain criticality due to the greater reactivity worth of the new fuel. Over the two days following refuelling the regulating rods will be driven out slightly to match the build-up of Xe-135 to its equilibrium concentra­tion іплЬе new fuel. Also some trimming will be re­quired to optimise the reactor output with the new fuel, not only the new fuel stringer itself but also the nearby fuel stringers because of the effect of the new stringer on them (see Section 8 of this chapter). The strategy by which the temperature and power distribution is optimised is carefully chosen to mini­mise temperature cycling of the fuel.

Consideration of certain faults which may arise during refuelling on an AGR currently require that this operation be carried out at reduced power, there­fore the concern for exceeding temperature limits on fuel in the vicinity of the refuelling is greatly re­duced. Because of this, and the fact that moderator temperature is maintained largely constant by the re­entrant gas flow, and the greater reactivity invest­ment in regulating rods (supplemented by trim rods at Dungeness B), the kinetics of reactor parameters during refuelling is much less onerous on an AGR than on a magnox reactor from the control engineer’s point of view. His attention is concentrated more on the safety of the fuel stringer being handled, parti­cularly in the removal of the irradiated stringer, and on the effects on the boilers of the gas at reactor gas inlet temperature which streams from the empty channel.

Start-upAvill be defined as the point of commencement of control rod withdrawal or at the commencement of withdrawal of a specific group of rods, Normally, a reactor is considered shut down when the core is 1000 milliNiles reactivity below critical condition. Therefore start-up will be deemed to have commenced at some convenient stage of control rod withdrawal. This will normally be chosen at worst reactor condi­tions to ensure that the reactor remains subcritical in any event.

Checks on the state of the plant’s readiness will be made just prior and after the start-up has com­menced. The majority will be before start-up has com­menced as defined in the operating documents. The operating documents must clearly define the sequence and checks to be carried out and take the operator through the correct route for a successful start-up. Acknowledgement at each stage must be made by a signature at each step and major stages acknowledged by a senior person, usually the shift manager, for further progression.

The purpose of the coolant is to remove heat from the fuel elements and to transfer it to the boilers. With most designs of gas cooled reactor some leakage of coolant occurs and availability of supply has to be considered. If it were not for this factor, the gas helium would be ideal: as it is carbon dioxide is chosen for magnox and AGR stations.

Carbon dioxide is a minor constituent of the atmos­phere: it is colourless, odourless and heavier than air. Chemically speaking it is an acid anhydride, i. e., it will react with water to form an acid-carbonic acid. Although carbonic acid is a weak acid, moist carbon dioxide is corrosive to some metals, particularly mild steel. In addition other minor constituents found in the reactor gas, e. g., ammonium chloride, reduce the relative humidity required to produce condensation and it is therefore essential to control the moisture concentration to prevent dewing out of water.

The individual elemental constituents of carbon di­oxide, carbon and oxygen possess the main nuclear property of low neutron absorption cross-section. Car­bon dioxide is also chemically compatible with the reactor environment, since it decomposes only slowly under irradiation. However, it can react with the core graphite, which can be expressed by the simple reaction C + CO; ~ 2CO.

This reaction can proceed by either of two processes — radiolytic or thermal reaction. During normal operation the former is important for both reactor designs but the latter has also to be considered for the highest temperature graphite components in an AGR, i. e., the upper inner graphite sleeves and in some reactor designs — graphite boiler support pads. The equation is inadequate to describe the complicated sequence of reactions w’hich take place in the core
but it serves to illustrate the net reaction by which the carbon dioxide can attack the graphite. The re­action can proceed to an equilibrium with the reverse reaction (hence the double arrow’ symbol), termed in the thermal case as the Boudouard Reaction. This is subject to the Law of Mass Action which means that as the concentration of carbon monoxide is increased the forward reaction is progressively reduced.

Since the graphite is a structural member, and graphite strength is reduced by loss of carbon, the reaction can only be permitted to proceed at an ac­ceptable rate. In addition the graphite is the reactor moderator and hence significant loss of carbon would affect the nuclear properties of the reactor. In magnox reactors where the rate of the reaction is low, due to the low dose rate and low total gas pressure, it is necessary merely to maintain a small proportion of carbon monoxide in the coolant, in order to limit the rate of the forward reaction, and so to reduce the loss of graphite to acceptable levels. In AGR reactors where the dose rate and total gas pressure are higher, the reaction would proceed to an unacceptable loss of core integrity if limited by carbon monoxide alone and further protection has to be provided.

The rate of the reaction is dependent on the type of graphite and the material used in AGR (giiso car­bon graphite), which was primarily developed to reduce the fast-neutron-induced dimensional change, exhibits a rate approximately one-half of that of the graphite used in magnox reactors (pitch coke graphite). Even this is not sufficient to reduce carbon loss to accept­able values. It was demonstrated that low concentra­tions of methane can further significantly reduce the rate of reaction. However, the methane itself under­goes radiolytic oxidation according to the equation CH4 + 3C02 — 4CO + 2H20.

The consequent rate of formation of carbon mono­xide and water is significantly higher than in magnox reactors and hence in the AGR it is necessary to install large recombination units (2CO + 02 2C02)

and driers (H20 removal) to maintain the coolant composition within specification.

A further important consideration is that, if the concentration of methane or carbon monoxide are too high, carbonaceous deposit may begin to form on fuel pin and boiler surfaces, inhibiting heat transfer from the fuel to the coolant. The detailed choice of coolant requires optimisation between core corrosion, steel corrosion, fuel pin deposition, boiler deposition, and coolant composition control.

Helium is also added to the coolant in both mag­nox and AGR reactors being maintained at a concen­tration between 50 and 500 vpm. The reasons for the addition are:

• Asa measure of the reactor carbon dioxide leak rate.

• As a measure of the reactor carbon monoxide leak

rate which is used as a measure of the core corrosion

rate.

Due to the typical parabolic rate of increase of sur­face corrosion and oxide film thickness with time (Mann 1976 [36]), it is logical to allow the surface oxide to establish before fuel is loaded. The period of most rapid corrosion and corrosion product release therefore occurs when no fuel is loaded and the sys­tem can be cleaned up by the CVCS. This can be achieved during hot functional testing when the water temperature reaches 300°C due to heat input from the pumps, by operation with lithium hydroxide and boric acid present. Corrosion and corrosion-product release is then reduced during subsequent reactor operation at power.

A more positive chemical pre-treatment under de­velopment has the objectives of controlling the rate — of-growth, structure and composition of indigenous oxide films. The process is required to be simple and to produce oxide films that are stable, resistant to uptake of Co-60 and repairable after shutdown.

The production of a stable oxide film, known as pre-filming, can be achieved by treatment with ethy — lenediamine tetra-acetic acid (EDTA) at 200°C. This produces a glassy film of СГ2О3 and FejCU, followed by conversion to black crystalline material containing Fe}04 and FeCr204 by raising the temperature to 250°C.

However, the process kinetics and the stability of the resultant films at 300°C are unknown and further development work is required, although the process has been used on Canadian BWR systems (CANDU).

A further pre-filming development is the controlled growth of nickel ferrite in an alkaline/reducing envi­ronment, with an excess of benign cations present (e. g., zinc or manganese) to fully substitute the lat­tice structure of nickel ferrite. This produces a stable nickel ferrite film which is unable to take up Co-60 cations and therefore cannot accumulate activity due to grown-in active species. However, as above, the long term stability of the film at 300°C is unknown.

All pre-operational chemical treatments are designed to produce an oxide layer of enhanced adhesion and chemical stability, which is resistant to uptake of active species such as Co-60.

Start-up

Early establishment of the co-ordinated primary cir­cuit specification is judged to be important, although any benefit will only emerge after several fuel cycles, when radiation levels may be shown to be lower than on a reactor where early establishment of co­ordinated chemistry was not adopted. The lithium hydroxide concentration should be established during pre-operational testing through to normal operating temperatures, and oxygen levels should be below the recommended value as soon as possible. The latter may require nitrogen purging during start-up and the use of hydrazine to ensure less than 0.1 ppm oxygen above 80°C.

Charging the RCS up to pressure may cause a transient in the dissolved hydrogen level, due to a changing CVCS volume control tank pressure. This will subside when charging is complete.

Shutdown

The reactor shutdown causes transients in the primary circuit chemistry due to the removal of hydrogen and the addition of shutdown levels of boric acid (2000 ppm), necessary to suppress residual core reactivity. As noted earlier, the end of fuel cycle chemistry will already be characterised by low boron and low or zero lithium. In addition, there are temperature, pres­sure and hydrodynamic changes, and ultimately the primary coolant water is oxygenated by exposure to air via the refuelling pond water.

It is not surprising that these transient conditions cause an order-of-magnitude increase in circulating activity levels due to the release of crud to the pri­mary coolant. Currane, Stagg and Shaw (1979) [37] have identified three phases in this increased activity release. Initially, an increased Co-58 solubility ac­companied the cooldown and boration to 2000 ppm, followed by a maintenance of the Co-58 level while reducing chemistry was retained. Finally, when oxy­genation of the primary coolant occurred, there was a large increase in soluble Co-58 levels.

In terms of PWR shutdown, it is possible to deli­berately generate oxidising conditions by the addition of hydrogen peroxide (H202), and remove the re­leased Co-58 activity by use of the CVCS ion ex­change units before the system is opened and allowed to oxygenate via the refuelling pond water. However, this technique is not universally used.

It should be noted that normal short duration non­refuelling shutdowns would not be subjected to this oxygenation, and efforts would be made to retain the normal operational chemistry.