Skip Ribbon Commands
Skip to main content

Skip Navigation LinksGeneral Design and Principles of the Advanced Gas-Cooled Reactor (AGR)



General Design and Principals of the Advanced Gas-Cooled Reactor (AGR)

1. Introduction and History
2. Outline Design of the Advanced Gas-Cooled Reactor
3. The Reactor Core
4. Core Restraint Structure
5. Control and Primary Shutdown Systems
6. Secondary Shutdown Systems
7. Post Trip Cooling
8. Fuel Assemblies
9. Primary Heat Cycle (Gas Sided)
10. Secondary Heat Cycle (Steam Side)
11. Pressure Vessel


1. Introduction and History

The Advanced Gas Reactor (AGR) represents the generation II of British Gas-Cooled Reactors, developed from the earlier generation I Magnox reactor design (see “Magnox Design” on page 7) of gas-cooled reactors. In the UK there are currently seven AGR nuclear power stations (5 in England and 2 in Scotland) each with two operating reactors. They are all owned and operated by British Energy, and are located and known as Dungeness B, Hartlepool, Heysham 1, Heysham 2, Hinkley Point B, Hunterston B and Torness. These generate nearly 20% of the electricity supplied in the United Kingdom, and the seven power stations of this type are amongst the world’s most efficient means of converting the energy of nuclear fission into electricity, as well as being extremely safe.


The first commercial AGR power stations, those at Hinkley Point in Somerset and at Hunterston in Ayrshire began operation in 1976, the latest, at Heysham in Lancashire and Torness in East Lothian in 1988.


Each of the 14 separate reactors is linked with its own 660 MW turbine-generator unit. These turbine-generator units are exactly the same as those to be found at a coal or oil-fired power station of equivalent electrical output (the only difference being the source of heat i.e. at an AGR power station this is nuclear energy and at a conventional power station is the energy of combustion). The end result is the same - the production of heat to generate steam. The components of an AGR reactor core are discussed in further detail below.

2. Outline Design of the Advanced Gas-Cooled Reactor

In a typical AGR system the reactor core, boilers and gas circulators are housed in a single pre-stressed concrete cavity known as the pressure vessel. The reactor core is a sixteen sided stack of interconnected graphite bricks maintained in position by a steel restraint structure surrounding the core, and is supported by a system of steel plates known as the diagrid. The complete assembly of the core comprises of 332 fuel channels.


The active region of the core containing the fuel consists of 10 years of graphite bricks, each about 825mm high. The active core is enclosed by further graphite which makes up the upper, lower and side neutron shields – the shield is provided in order to safely gain access to the boilers and plant within the pressure vessel when the reactor is shut down and de-pressurized. Additional shielding is provided by steel plates fasted to the steel restraint system.  The whole structure is approximately 11m in diameter and 9.8m high and weights around 1300 tonnes.


The graphite core and shielding is surrounded by a steel envelope known as the gas baffle. The diagrid forms the lower end of the gas baffle cylinder. The function of the gas baffle is to provide a route for the coolant gas through the core, and is composed of three main sections: the dome at the top, the cylinder and the skirt. At the dome there are a number of penetrations to allow the passage of fuel assemblies, control rods and to other interstitial channels in the core.


General Schematic of AGR Penetrations at Reactor Dome


Each reactor has 308 fuel channels, 263.5mm in diameter, that run vertically down through the reactor. They are arranged in a regular, square, lattice pattern with a distance of 460mm between their centers. In between the fuel channels there are 81 control rods channels, each 127mm in diameter. Each reactor sits on 109 steel support plates which are in carried in turn by a steel “eggbox” structure – the aforementioned diagrid. The diagrid is supported from pressure vessel floor by 16 diagrid supports.


Virtual Build up of Reactor Core Showing Peripheral Structure Virtual Build up of Reactor Core Supporting Diagrid


Heat is produced within the core of the reactor, and is transferred to pressurized carbon dioxide gas, which is pumped over the fuel by gas circulators around the gas baffle. The gas, in turn, transfers its heat to boilers. Feed water is pumped into the boilers and is converted to superheated steam, which is taken to a conventional 660 MW(e) steam turbine-generator. The energy in the steam is converted to electricity in the turbine-generator, which is supplied to the National Grid. The steam is condensed and supplied back to the boilers.

Nuclear Power Generation Process


Plan of Reactor Building Ground Level

Plan of Reactor Building Level 09

Elevation through Reactor and Fuel Handling Building

Elevation through Reactor Building

Core Channel Utilisation Diagram

3. The Reactor Core

As aforementioned the core graphite structure is designed to act as a moderator and provide individual channels for fuel stringer assemblies, control devices and coolant flow for a minimum period of 30 years – these in fact are part of the high level requirements of the reactor core for reactor operation, which are to:

  • provide a structure which, under all operating and fault conditions, ensures that the reactor can be shut down;
  • ensure adequacy of cooling for fuel, control rods, neutron sources, steel absorbers and the graphite must be maintained so that all materials remain within their design limits;
  • maintain unimpeded movement of fuel and control rods to ensure that the above requirements can be met.

Within the core two types of brick are used: one type is a basically circular and contains the fuel channels; interstitial bricks that are basically square and contain either control rods, secondary shutdown or coolant holes make up the second type.

The circular bricks are interconnected by loose graphite keys, while the connection between circular and interstitial bricks is through a key that is an integral part of the interstitial brick. Brick shapes and loose keys are optimized for strength and keying is designed to accommodate core movement and seismic loading.

Graphite Keying Structure

Between each layer of graphite, inter-brick seals are incorporated for the purpose of maintaining the pressure differences generated across the walls of the fuel channels by the appropriate sizing of the flow control ported located in the top and bottom layers of bricks for cooling.

The overall graphite structure is highly redundant, so that any local failures resulting from, for example, a local high load situation, tight clearances or a faulty component, will not result in gross channel distortions or in the failure of surrounding components. The integrity of the core structure as a whole varies over the life because of the effects of irradiation and radiolytic oxidation. 

4. Core Restraint Structure

The core restraint system is located between the outside of the core and the surrounding boiler shield wall. It maintains the stability of the reactor core. As shown by Figure 12 the lower 3 layers of the core are connected to the vertical restraint columns. These are located at their lower end on keying blocks mounted on the diagrid and at their upper end to a circumferential stiffening ring welded to the boiler shield wall. The remaining layers of the core are connected directly to the circumferential rings welded directly to the boiler shield wall.

Virtual Build up Illustrating Details of Restraint Structure at Core Bottom

At the core periphery, steel puller rods are located in recesses in selected bricks at each inter-layer position. The rods are in turn attached to restraint beams that form sixteen-sided polygons around the core at each inter-layer position, the beam ends being loosely connected as a secondary restraint feature. Each beam is inset and supported by the graphite and attached to the restraint tank by two spherical ball-ended links (Warwick links), which provide the means of load transfer from the core.   


Restraint Structure Showing Connection to Graphite Bricks
Restraint Structure Showing Warwick Links and Restraint Beams


The restraint structure also exhibits a high degree of redundancy to cater for the possibility of local component failure. These structures do not suffer from the same strict geometrical and material limitations of the core and their integrity can be guarded by the optimum choice and materials and design.

5. Control and Primary Shutdown Systems

Reactor power is directly proportional to neutron density. In order to control the neutron density neutron absorbing control rods are raised and lowered within the core either by operator or automatic control according to the amount of power required.

There are a total of 89 control rods that are housed in standpipes in the top cap of the reactor vessel. These can be divided into the following categories:

  • There are a total of 45 “grey control rods” which are weak neutron absorbers and are used as regulating rods i.e. to regulate the power output of the reactor core. These are constantly moving in and out of the core under automatic control to provide a steady power output. 16 of these rods are used as a safety group which can be withdrawn when the reactor is shut down to provide a safeguard against inadvertent criticality.
  • “Black control rods” are the primary system for shut down for the reactor core by absorbing a large proportion of thermal neutrons. There are a total of 37 black control rods. This is referred to as the Primary Shutdown Device (PSD). 
  • There are 7 “sensor rods” that exist as part of the black control rod assembly: these are designed to detect any misalignment in the guide tubes between the reactor core and the steel structures above it when the black rods are being raised or lowered. 

The control rod assembly in each standpipe consists of a control rod, a control plug unit, a control rod actuator and the standpipe closure unit. The assembly is designed to be removed with the refueling machine under operating or shut down conditions. The control plug unit is designed to reduce to acceptable levels, radiation streaming from within the core through the standpipe penetrations in the vessel roof. It consists of a steel plug with a central hole through which passes the control rod suspension chain.

Each actuator is complete with motor-operated winding gear and suspension chain storage, electromagnetic clutch, hand-winding drive to the clutch, rod position indicator and limit switches. The actuator and rod drive is design frequent small movements, and the speed of the control rod is controlled by the induction motor. In the event of a reactor trip, the clutch is also tripped to allow control rod insertion by gravity, which is controlled by a carbon disc brake and centrifugal mechanism. 

6. Secondary Shutdown Systems

As a back up measures against failure of the PSD there exist several Secondary Shutdown (SSD)   devices. Different systems have been installed in different reactor cores, the two most common being the following.

The nitrogen injection system consists of an arrangement of pipework and control values linking a permanent nitrogen gas store to a group of 165 interstitial channels in the core. The store, consisting of banks of high pressure cylinders, is common to both reactors and holds sufficient nitrogen in gaseous form for the shutdown and subsequent holddown of one reactor, provided it remains pressurized. Approximately half an hour is required for full deployment of this system, which is completed in two stages.

The first stage is on trip where the nitrogen flows through feed pipes to each of the interstitial secondary shutdown channels. When the first trip values are opened the initial flow purges these channels of carbon dioxide and fills them with nitrogen. The second stage provides additional flow to each channel as the nitrogen flow from within the channels is forced into the re-entrant passages and through the fuel channels, thus gradually building up the nitrogen concentration in the coolant gas circuit until it is sufficient to hold the shutdown core in a sub-critical condition for several hours.

The boron bead device consists of a store loaded with boron glass beads of 3mm diameter which are injected into 32 of the secondary shutdown channels in the reactor core. Each of these channels has an associated delivery pipe, with one end terminated at the top of the channel and the other end connected to one of the bead storage hoppers. The hoppers are located below the reactor pressure vessel in the secondary shutdown room.

On trip, the carbon dioxide is used to convey the beads pneumatically from the bottom of the hopper to the top of the channel. The beads are directed downloads into the channel from the open end of the delivery pipe until the channel is filled. Sufficient gas is provided by the station coolant system to inject all the stores of beads simultaneously. For their recovery, the beads are passed from the bottom of each channel through a vertical recovery pipe terminating with an isolating ball valve in the secondary shutdown room.

7. Post Trip Cooling

As with all nuclear power stations after shutdown and fault-initiated trips the fission products still undergo radioactive decay and continue to release heat (decay heat) even after fission. Post-trip heat removal systems are provided to removal the residual heat so as to prevent the fuel overheating and damage to the reactor structures. The main functions of these systems are to firstly circulate sufficient gas coolant to transfer the heat from the fuel and elsewhere to the boilers and secondly to provide sufficient feedwater to the boilers to enable this heat to be transferred to the environment. Note that AGRs are capable of using natural circulation to prevent overheating, provided the main boilers have feedwater available.

There are two main systems for post trip cooling: these systems are arbitrarily designated ‘X’ and ‘Y’ systems. The ‘X’ system is the preferred system for cooling when the reactor is pressurised. Heat is removed from the fuel by forced circulation provided by the gas circulators operating at low speeds. The heat is removed form the reactor coolant by the decay heat boilers, or which there are four, situated directly below and separate from the main boilers in the boiler annulus.

The ‘Y’ system is used if either de-pressurisation of the reactor occurs or if the ‘X’ system fails. Heat is removed by forced circulation provided by the gas circulators operating a high speeds to compensate for the lack of pressure. The heat is removed from the reactor coolant by the main boilers fed by the emergency boiler feed system.

8. Fuel Assemblies

The slightly enriched (2 – 3%) uranium used as fuel in AGRs is made up in the form of cylindrical pellets. These are contained within sealed, stainless-steel tubes to form fuel pins. Each pin is approximately 980 mm in length. 36 fuel pins are arranged in clusters within graphite sleeves to form each fuel element, eight of which are linked together with a tie bar to form a fuel stringer.


Schematic of Fuel Pin Fuel Pin


At the top of the fuel stringer is the fuel plug unit. This comprises of a closure unit, a biological shield plug to limit neutron and gamma radiation through the standpipe, a gag unit to adjust the coolant flow through the individual channels, and neutron scatter plug to prevent neutrons streaming up the fuel channel. Gas-sampling instrumentation and channel gas outlet thermocouples are also incorporated into the unit to provide monitoring.


Insertion of the fuel assembly into the reactor core is handled by a single re-fuelling or charge machine. The machine runs on a traveling gantry that spans the width of the charge hall and is supported on rails running parallel along the length of the hall. The charge machine is designed to handle both fuel and control assemblies and is capable of refueling 120 channels per year.

During normal operation, the fuel plug closure units seal the reactor pressure vessel boundary at each fuel standpipe. A primary lock is provided on the refueling machine grab, and a secondary lock is provided by a retractable drive mechanism built into the standpipe extension tube.

Charge Machine


 Torness HeyB Fuel Handling Flow Cycle


9. Primary Heat Cycle (Gas Sided)

Carbon dioxide is the gas used for transferring the heat produced form the reactor core to the boilers. The gas is pumped through the gas baffle channels of the reactor core at high pressure by 8 gas circulators. There are two circulators per each quandrant. These are housed in large penetrations towards the bottom of the pressure vessel, and each one forces the CO2 around the primary circuit at a rate of around 500kg every second.

Cooled gas is drawn from the bottom of the boilers by the gas circulators and is discharged into the space below the core. About half of this has flows directly to the fuel channel inlets, while the remainder, known as the re-entrant flow, passes up the annulus surrounding the core and returns downwards through the core in the passages between the graphite bricks to rejoin the main coolant flow at the bottom of the fuel channels. The re-entrant flow is thus to cool the graphite bricks, the core steel restraint system and the gas baffle.

The combined flow passes up the fuel channels and is led through the space between the top of the core and the gas baffle in guide tubes that discharge the hot gas into the area above the gas baffle and thence to the top of the boilers.

Gas Circulator


Primary Heat Cycle through Reactor Core

10. Secondary Heat Cycle (Steam Side)

Each reactor has 12 boilers situated between the gas baffle and the pressure vessel liner and is partitioned into four quadrants, each containing one boiler and two gas circulators. Within each quadrant there are three main boiler units, each consisting of an economizer, evaporator and superheater sections. The boilers are of the once-through type.

Boilers Around the Periphery of the Reactor Core

Each high pressure unit is supported by bearings mounted on a pair of boiler support beams, and the reheater units are suspended from the roof by four support slings.


Reheater Hangers Boiler Support Beam


The hot carbon dioxide from the reactor flows into the boilers at around 4 tonnes per second and at a temperature of approximately 600°C. When the CO2 exits the boiler it is some 300°C cooler – this generators half a tonne of steam every second.

11. Pressure Vessel

The design of a gas-cooled reactor calls for pressure vessels to withstand the pressure of circulating coolant gas used to transport heat from the reactor core to the main boilers and thick concrete shielding to absorb neutrons and gamma radiation given off during fission. These requirements can be combined and in the AGRs, prestressed concrete pressure vessels are used to carry out these functions.

The pressure vessel walls are approximately 5.8m thick, the top slab 5.4m, and the bottom slab 7.5m. The prestressing and post-tensioning system consists of around 3,600 steel tendons in helical formation threaded through 76mm mild-steel tubes that are embedded in the concrete during construction. The tendons are anchored in stressing galleries at the top and bottom of the vessel, which provide access to the tendons and from which insertion and stressing can take place. Routine checking of the tendons is carried out throughout the life of the reactor. All the tendon strand anchorage loads can be checked individually and tendons re-tensioned or replaced if need be. The number of tendons is very much in excess of those necessary to provide the requisite strength, so that in theory many could fail without fear of pressure vessel failure.

Corner Liner Details Inside Pressure Vessel

On the inside of the vessel there is a steel liner that is gas-tight – the main purpose of which is to provide a leak-tight membrane to prevent the release of hot carbon dioxide gas through the concrete and penetrations, therefore minimizing the risk of release of radioactivity from the plant, as well as serving as a foundation for the cooling and insulation systems that protect the concrete from excessive temperatures and temperature gradients.

Layers of the Liner

The liner is insulated on the inner face and water cooled on the outside. The insulation consists of layers of ceramic fibres held in contact with the liner by a system of steel foils and cover plates, each of which is retained in position by a series of primary and secondary retention systems.


The cooling system is a closed-loop system circulated treated water through a network of pipes welded to the vault liner plate so as to provide full-length heat transfer paths from the metal surface to the circulating water. The water circulating through the network of pipes removes the heat passing through the insulation attached to the inner surface of the liners. Together, the insulation and liner cooling system ensure that the liners and vessel concrete are maintained at acceptable temperatures.