Nuclear Power Demonstration Reactor

1962 - 1987

The Canadian Nuclear Society held two very successful events celebrating the 40th Anniversary of NPD, on May 31 and June 1 2002

NPD from the air
NPD from the air in its later years, looking north. The Ottawa River is in the background, flowing left to right. Note the large training centre. (Courtesy of AECL)

NPD from the air
NPD from the air, looking west-northwest.   The Ottawa River is in the foreground and background (note the log boom).   Ontario Hydro's 440 MWe Des Joachims (Da Swisha) hydro dam appears beyond the reactor.   Note the training centre is smaller in this earlier photo. (Courtesy of AECL)

NPD start-up, April 11 1962
NPD start-up, April 11 1962 (Courtesy of AECL)

Inaugural Ceremony
Inaugural ceremony for the Nuclear Power Demonstration reactor, 1962

NPD today
NPD in March 2001

The following article, on Canada's first power reactor and the prototype CANDU, was published in the October 1962 edition of Nuclear Engineering and is reprinted here with the permission of Nuclear Engineering International.  

The main organizations involved in the NPD (Nuclear Power Demonstration) reactor were:

Fig. 1. - NPD site and plant

Nuclear Power Demonstration Reactor

From our Canadian correspondent

Canada's first step toward the utilization of nuclear power for the generation of electricity is the design, development and operation of the NPD (Nuclear Power Demonstration) Station.   This station was built to demonstrate the reliable operation of the heavy water moderated natural uranium type power reactors, to provide useful cost information on the construction and operating costs of a nuclear station and to serve as a prototype for future full-scale nuclear power stations.

Early in 1954 a study group of engineers and scientists was assembled at Chalk River under the auspices of Atomic Energy of Canada Ltd. (AECL) to investigate the feasibility of building natural uranium heavy water reactors.   The recommendation of this group was to build a small reactor of the pressure vessel type.

In 1955, AECL invited proposals from Canadian industry to participate in the design and construction of a nuclear power station and from Canadian utilities to operate the station and to utilize the power produced.   From the proposals received, the Canadian General Electric Co. Ltd. (CGE) was selected to design, develop and act as prime contractor for the construction of the station and the Hydro-Electric Power Commission of Ontario (HEPC) was selected to provide the site, design the buildings and to operate the station as part of its system.   AECL would provide the funds to build the station.

Work commenced immediately on the design and development of the station and by 1957 construction was well underway.   In the meantime, further studies by AECL suggested that sufficient advances had been made in the suitability and availability of zirconium alloys to permit the construction of a pressure tube reactor which would be more applicable to large power reactors.   In 1957 work was stopped on the pressure vessel design and a detailed study was done on the pressure tube design.

In late 1957 work was started on the detailed redesign of the NPD station under the same contractual arrangements between the three partners.

Construction at the site was initiated in 1958 and the station was completed in 1962, and first critical was achieved on April 11, 1962.   Full power generation of 20 MW(e) gross was first achieved on June 28, 1962.

General Description

The NPD station is located in the Province of Ontario on the west bank of the Ottawa River about 140 miles upstream from the nation's capital.   It is close by the AECL research establishment at Chalk River and the HEPC hydro generating station at Des Joachims.   It has a maximum continuous output of 22 MW(e), which gives a net station output of 19.5 MW(e).

The basic station thermal cycle (Fig. 2) consists of the natural uranium fuelled heavy water moderated reactor, with 132 horizontal Zircaloy-2 pressure tubes, the heavy water coolant circulating pumps and piping, the heavy water-light water heat exchanger and steam generator, the turbine with coupled electrical generator, the main condenser, the reject condenser and the feed water pumps.

Fig. 2. - Flow diagram

The reactor is fuelled with 40 000 lb of UO2 in the form of ceramic pellets sheathed in Zircaloy-2 tubes.   The fuel is subdivided into nine bundles per channel and fuel changing is designed to be carried out on power by remotely operated machines which push a fresh bundle in one end of the channel and remove the spent bundle from the other end.   The fuel changing will be scheduled in such a manner that each bundle will receive approximately the same burn-up.

The operating conditions of the primary heat transport system are 1 034 psig at the outlet header, a flow rate of 10 000 gal (Imp.)/min with a bulk average coolant inlet temperature of 485oF and bulk average coolant outlet temperature of 530oF.   At the normal gross station output of 20 MW(e) steam is raised in the boiler at a rate of 296 000 lb/hr and at a pressure and quality of 415 psig, dry saturated.   The turbine is a single cylinder 3 600 rev/min unit with three extraction points for feed water heating and is operated at an exhaust pressure of 1.5 in Hg(abs).

The station is designed for centralized control, to be served by a normal watch of four men consisting of the shift engineer, the control room attendant, and two field operators on roving duty.   Maintenance and other services are provided in support of this basic staff.

The station is intended to operate at a minimum capacity factor of 80% with an availability factor for the reactor plant of over 90%.

Fig. 3. - Cut-away of core including the calandria which is about 17 ft in dia. and 15 ft long.

Reactor and Core

The reactor consists of a 132-tubed, double-walled aluminium calandria vessel which contains the heavy water moderator, the heavy water side reflector, the light water side and end reflector and the coolant tube assemblies which pass through the calandria tubes and which, in turn, contain the fuel and heavy water coolant (Fig. 3).   The moderator dump ports, which provide the liquid-gas interface necessary for support of the heavy water moderator by a differential helium gas pressure system, are also an integral part of the calandria structure (Fig. 4).    The vessel is hung on four 3½ in d. spherically ended steel rods which accommodate the movement of the reactor due to thermal expansion.

Fig. 4. - This view of the calandria shows the three dump pipe connections, the inner wall, the spacing ribs and end walls.   Holes in the end plate indicate the location of the coolant channels.

The reactor is connected to an aluminium dump tank by three 24-in pipes which permit the dumping of the moderator at 100 000 gal (Imp.)/min to provide a reactor scram.   The dump tank contains a baffling arrangement to deflect and dissipate the energy in the falling water.   There are no control rods used in the reactor.

The helium balance line is connected between the dump tank and the top of the calandria vessel.    It is equipped with three rapid-opening butterfly valves which equalize the helium pressure to initiate the moderator dumping action.

The theoretical core is cylindrical, 11 ft in diameter and 12 ft 7 in long.   The core consists of 132 cells on a square lattice with a pitch of 10¼ in.    These cells are arranged in an irregular octagon having alternatively eight and three cells to a side.   The eight cell sides are horizontal and vertical.

The core is encircled by a heavy water reflector which has a radial thickness of 21.6 in at the core midpoint and tapers off to a 5.6 in thickness at the core ends.   The reflector is not separated from the moderator.   The core and reflector are, in turn, surrounded on the side and ends by a minimum 1 ft of light water which serves as both a neutron reflector and shield.


The natural uranium dioxide pellets are sheathed in Zircaloy tubes which are welded at each end to Zircaloy end plugs, to form fuel elements.   The elements are, in turn, assembled into cylindrical bundles by welding the end plugs to Zircaloy end plates (Fig. 5).    Each bundle is 19.5 in in length and sized to slip easily through the 3¼-in inside diameter of the pressure tube.   Selected elements in each bundle have two spiral wire fins which are welded to the sheath and which provide positive spacing between elements and the coolant tube, and which also provide mixing of the coolant among the elements.

Fig. 5. - Ground pellets are loaded into Zircaloy tubes capped at one end by an automatic welding process.   After purging air and refilling with inert gas the other end cap is welded on.

Three basic fuel designs were used in the first fuel loading.   A charge of 720 seven-element bundles with 1 in o.d. elements has been loaded into the 80 outer or lower power channels in the reactor core (Fig. 6), while 468 nineteen-element bundles with 0.6 in o.d. elements which have been loaded into the central 52 channels (Fig. 7), out of which 62 contain 0.242% U235 by weight depleted fuel as a means of reducing reactivity until equilibrium operating conditions are achieved.   The element spacing in both the seven-element and nineteen-element designs is 0.050 in.

Fig. 6. - Vertical cross section through core.

Other variations in the fuel bundle designs include 0.015 in sheathing in place of the standard 0.025 in sheathing, nickel-free Zircaloy-2 sheathing and riveted rather than welded bundle assemblies.   These variations in the fuel design represent the continued advances in fuel technology which permitted improvements in design to be incorporated during the two-year period of manufacture.

Fig. 7. - Two types of fuel bundle, one subdivided into nineteen elements, the other comprising seven.

The power rating of the fuel at 20 MW(e) gross station power output is ∫ kdθ = 30.5 W/cm for the most centrally located seven-element bundle or 2.68 kW/cm of bundle length while ∫ kdθ = 18.1 W/cm for the central nineteen-element bundle or 4.04 kW/cm of bundle length.   The fuel is considered, in the light of present technology, to be conservatively rated.

The two-zone core was selected for three main reasons: -
(1) The knowledge of fuel performance at the time the fuel designs were finalized placed the limit on power rating at ∫ kdθ = 35 W/cm, which necessitated a big degree of fuel subdivision.
(2) The seven-element fuel was less expensive to fabricate than the more highly subdivided nineteen-element fuel.
(3) The resonance neutron capture in the fuel bundle is a function of the internally exposed perimeter and the average distance between fuel faces or, in other words, the subdivision of the fuel.    The lower the subdivision of the fuel, the lower the resonance capture which results in better reactor economics and lower reactivity increase in case of a loss of coolant incident.

The longitudinal subdivision of the fuel into short bundles permits each fuel bundle to be taken to a uniformly high burn-up.   The bundle length was selected on the basis of an optimization study of reactor economics.

The fuel bundles in adjacent channels in the reactor are fuelled in opposite directions.    This has been termed bi-directional fuelling, which results in a very symmetrical axial flux distribution.   The radial flux distribution will be approximately sinusoidal but can be flattened by slowing down the fuel throughput in the central channels and increasing the throughput in the outer channels.

The maximum residence time in the reactor for the outermost fuel bundles is approximately six years.   The expected average fuel burn-up is 6 300 MWd/t(U) and approximately two bundles will be changed for each full-power day of operation.


The calandria (see Fig. 3) is a horizontally positioned cylindrical Alcan C54S aluminium alloy vessel with double side and end walls.  It is about 17 ft in outside diameter and about 15 ft long.   The inner side wall is ¼ in thick, the inner end walls ½ in thick, while the outer side wall is ½ in thick and the outer end walls 1½ in thick.   The side walls are supported by internal stiffening rings or ribs made from ¼-in aluminium plate.   The end walls are joined by 132 stepped tubes which are ¼ in thick and are welded to both the inner and outer walls.   Each stepped tube has an inside diameter of 6 in at the outer end and is stepped down to 4½ in at the inner end.   The Alcan 57S calandria tubes, which are joined at both ends to the inner end of the stepped tubes with mechanically rolled joints, are 4 in in inside diameter and 0.054 in thick.

At the bottom of the calandria are two parallel reverse bend dump ports.   The dump port structure is an integral part of the calandria vessel and is a double-walled vessel with provision for an internal light water reflector.   The dump port provides a minimum flow area of 20 ft2 and is connected directly to the dump tank via three 24-in aluminium pipes.

The calandria vessel is penetrated where necessary to accommodate the piping connections for the moderator, helium and reflector process systems.   An internal piping manifold at the top of the calandria supplies heavy water to a system of spray nozzles whose function is to cool any section of the calandria wall or calandria tubes which are not submerged in the heavy water moderator either during operation of the reactor or during reactor shut-down periods when the main heat transport system is still hot.

The calandria vessel contains more than four miles of welds which were 100% radiographed during construction.

Coolant Tube Assemblies

There are 132 coolant tube assemblies each comprised of a 3¼-in inside diameter, 0.063-in thick, 13¼-ft long Zircaloy-2 coolant tube attached at each end to a type 410 ferritic stainless steel end fitting (Fig. 8) by an expanded joint.   Each assembly is held in place in the calandria by two end supports and by an annular spacer at the centre of the calandria tube.   The end supports are tubular in form, are flange mounted to the outer end walls of the calandria, are concentric around the coolant tube end fittings and have four internal axial female splines which mate with four external male splines on the end fittings.   This design permits relative axial motion between the coolant tube assembly and end supports due to differential thermal expansion and, at the same time, prevents induced torque or bending moment loads to be placed on the coolant tube and expanded joint from the externally connected feeder pipes.   The feeder pipe joints are two-bolt metal gasketed fittings which are accessible for tightening if required.

Fig. 8. - End fittings on each end of the 132 coolant tubes couple up to the fuelling machine snout.

The end fittings are equipped with remotely removable closure plugs which are fitted with high-pressure self-actuating flexible diaphragm seals.   The closure plugs are held in place by a breech-lock type thread and are equipped with positive locking devices.

The flow inlet end fitting which also forms the fuel discharge end of the coolant tube assembly, is equipped with a flow distribution orifice attached to and removed with the closure plug.   The flow outlet end fitting, which forms the fuel insertion end of the coolant tube assembly, is equipped with a spring-loaded six-fingered fuel latch which prevents the fuel from moving out against the closure plug due to the hydraulic pressure drop along the fuel channel (Fig. 9).

Fig. 9a. - Detailed arrangement of end fitting.   Removable plugs seal the end of the tubes to maintain coolant pressure.

Fig. 9b. - For alignment of fuelling machine snout assembly four sensing fingers indicate relative positions.

The replacement of the coolant tubes, end fittings and calandria tubes by remote maintenance equipment is a design feature of the NPD reactor.   This equipment is in the final stages of manufacture and proving.

Fuel Handling

Refuelling is carried out by means of two remotely operated fuelling machines which are located at opposite ends of the reactor (Figs. 10, 11).   The machines are identical and are capable of either inserting new fuel or receiving spent fuel, which is consistent with the bi-directional fuelling concept.

Fig. 10. - Cut-away drawing of NPD fuelling machine head.

Fig. 11. - Reactor vault showing the ends of the coolant tubes with sealing plugs in place.

When not in use they are withdrawn from the reactor vault by shielding doors.   To carry out the fuel-changing operation the machine heads are lowered into the vault on telescoping tubes and are homed onto the same fuel channel by horizontal motion of the carriage in two planes and by vertical motion of the telescoping tube, Fig. 12.   The machines lock on the end fittings, form a high pressure seal and are then pressurized, following which the closure plug is removed from each end of the channel.   A new fuel bundle which has been previously stored in the fuelling machine magazine is inserted in one end past the fuel latch and the spent fuel bundle pushed out the other end is accepted and stored in the magazine of the second machine.   The closure plugs are replaced, a check for leakage is made around the plug and after full depressurization the machines are uncoupled from the end fittings.   The spent fuel bundle is discharged into a temporary storage fuel bay via a fuel chute.

Fig. 12. - Each fuelling machine comprises a carriage mechanism and a head.   The carriage assembly supports the head by means of a vertical telescopic column.

The machines are designed to carry out the fuel-changing operation with the reactor shut down or with the reactor at full power.   Special equipment has been provided also for remote maintenance of the head, which may become contaminated with radioactive particles.   In case of a severe maintenance problem the machine head may be removed and immersed in a water bay for easier dismantling and maintenance.

The fuelling machine controls are all electric and are located in the main control room.   The fuel-changing sequence is initiated by the operator and from there proceeds fully automatically from step to step.   The steps completed and the step in process are displayed visually by indicating lights and the operation in the vault may be observed by remotely operated television cameras.   The controls for each machine are identical and stepping switches are used to provide interlocking of the control sequence.

The fuelling machines were used to load the first fuel charge of 1 188 bundles.    The reactor was pressurized but not at temperature, although fuel-changing operations have been carried out under operating pressure and temperature.

Primary Heat Transport System

The primary heat transport system consists of a boiler, three main pumps, carbon steel piping, valves, the reactor coolant tube assemblies, the heavy water coolant, and associated auxiliary circuit.   To deliver 22 MW(e) from the turbine generator at 27% efficiency requires 81 MW of thermal energy or 2.76 x 108 Btu/h.   The operating conditions of the primary circuit which gives this energy output are:

Coolant inlet temperature to reactor 485oF
Coolant outlet temperature from reactor 530oF
Operating pressure at reactor outlet 1 021 psig
Flow rate 10 000 gal/min
Max. coolant velocity over fuel 15.6 ft/s
Max. coolant velocity in steam generator 15 ft/s
Coolant circulation time 14 s/cycle

Boiler.   The boiler consists of a U-shaped shell and tube heat exchanger and a steam drum connected by risers and downcomers.   The heavy water coolant circulates through the tube side of the heat exchanger.   The tubes are inconel and are expanded and seal welded to an inconel-clad carbon steel tube sheet.   The rest of the unit is carbon steel.

The centre line of the boiler is located 12 ft above the level of the top coolant tube in the reactor to maintain natural circulation of the reactor coolant by thermal siphoning.   This should remove about 2 MW of heat from the reactor with the primary pumps stopped and a maximum fuel sheath temperature of 540oF.

Primary Pumps.   These pumps are vertical, single-stage centrifugal pumps which are equipped with dual mechanical shaft seals.   The pumps are constructed from 11-13% chrome steel and are each designed to supply a flow of 5 000 gal/min against a head of 400 ft.   The pumps are driven by 800 hp motors and are equipped with flywheels to provide an extended rundown in case of power failure.   The system normally operates with two pumps working and the third as stand-by.

Piping.   The primary circuit piping is ASTM A106 schedule 80 carbon steel, with the exception of the Zircaloy coolant tubes and the inconel heat exchanger tubes.   All joints except the flanged connection from each feeder pipe to the end fitting are welded.

Piping expansion is accommodated by flexibility of the piping arrangement and by some movement of the pumps and boiler.

The feeder pipes from each end fitting are 1½ in in diameter and are connected to stepped headers which are, in turn, connected to 16-in mains.   The feeder pipes are not provided with isolation valves to eliminate any source of stoppage of flow in a coolant tube.   Maintenance can be carried out if necessary on the feeder pipe connection by freezing fluid in the piping.

Valves.   Three main isolating valves are provided in the system.    They are located on the reactor inlet, the reactor outlet and on the connection between the boiler and the pumps.   There are also three check valves located on the discharge side of the pumps to prevent back-flow through the idle pump.   The material used for valve construction is generally carbon steel with hardened surfaces.

Surge Tank.   A surge tank is provided to accommodate small changes in the liquid volume in the primary circuit.   The pressure in the primary circuit is controlled by maintaining a vapour space above the liquid level in the surge tank by electrical heating.   A spray injection system at the top of the surge tank can collapse the vapour and quickly reduce the primary circuit pressure.

The surge tank has a liquid level control which governs the admission or discharge of coolant from the system.   It is connected to the outlet header from the reactor and is located above the primary circuit.

Stand-by Cooling.   Maintenance to the main pumps, or other parts of the primary circuit during a shut-down period requires an alternative system for removal of heat.   This is provided by an auxiliary circuit equipped with two 1.5 hp centrifugal pumps and a 250 ft2 cooler which provides a cooling capacity equivalent to 1% full power.

The stand-by coolant circuit is placed into service by closing the two main isolating valves on the inlet and outlet headers.

Primary Make-up Circuit.   The primary make-up circuit is used to maintain the water level in the surge tank.   The outflow from the system to the ion exchange circuit is balanced by a small centrifugal pump which is capable of delivering 48 gal/min against full system pressure.   In addition, two auxiliary pumps having a total capacity of 300 gal/min are provided to supply make-up coolant in case of a break in a coolant tube or feeder pipe.   The source of heavy water for the make-up circuit is the moderator dump tank.

Primary Purification Circuit.   A flow of 40 gal/min is diverted from the primary circuit outlet header to the purification circuit.   The water first passes through a regenerative heat exchanger and then to a depressurizing valve which reduces the pressure to 60 psia.   The water then passes through a cooler and on to the filters and demineralizes at 120oF.    Two edge filters capable of being cleaned during reactor operation are located in parallel.   Two alkali type down-flow mixed bed demineralizers maintain the pH of the system at 10.5.   The resin beds may be changed during reactor operation.   The water is returned to the primary circuit through the regenerator by the make-up pump.

Moderator System

The reactivity of the reactor is controlled by adjusting the level and temperature of the moderator in the calandria.   The moderator height is set by the difference in helium pressures between the calandria and the dump tanks.   The dump tank helium pressure will be 6.5 psi above atmospheric while the pressure in the space above the moderator will be less than the dump tank pressure by an amount equal to the head of water over the dump port.   Under normal operating conditions a moderator flow of 1 300 gal/min spills over the dump port continuously and is circulated through the cooler to remove the heat picked up by the moderator and to maintain the temperature between the control limits of 120oF-180oF.

A heat load equivalent to about 6.1 MW is deposited in the moderator - 5.6 MW generated by radiation adsorption and neutron collision and 0.5 MW transferred by thermal leakage.   This heat is removed in the moderator cooler, which is of shell and tube construction with a surface area of 2 350 ft2.   The cooling water comes from the river and varies seasonally from 34oF to 72oF.   The moderator flow through the cooler is 1 300 gal/min and the temperature is controlled by adjusting the cooling water flow rate.   The cooler is designed to remove up to 6.5 MW of heat.

Three moderator pumps of the shaft seal type are provided, each of which has a capacity of 920 gal/min against a 105-ft head.   In normal operation the flow rate of 1 300 gal/min is provided by two pumps in parallel with the third on stand-by.

A small flow of 40 gal/min maximum is directed through the purification system, consisting of an edge type filter and two mixed bed demineralizers which maintain the chemical purity of the moderator and remove foreign materials.

The 12 000 gal dump tank is located 20 ft below the calandria centre line to provide the required hydrostatic head necessary to achieve the dump flow rates.    The dump tank is fabricated from C54S aluminium and was subjected to the same rigid quality control inspection as the calandria.   It has sufficient capacity to store the full load of heavy water moderator and reflector.

Helium System

Helium was selected for the cover gas because it is inert, does not acquire induced radioactivity, may be readily maintained in a chemically pure condition and its cost is not excessive.   It is stored in a 1 500-ft3 gasholder which is replenished from helium cylinders automatically.

The helium pressure differential is maintained by two blowers which each pump 90 ft3/min; the pressure is modulated by six 2 in by-pass control valves.   Six quick-opening 10 in helium dump valves are provided to equalize quickly the helium differential and dump the moderator.

Light Water Reflector

The light water reflector located between the inner and outer walls of the calandria is supplied from the demineralized boiler water supply; the total light water volume is 10 000 gal.   During operation a flow of 1 000 gal/min is drawn from the reflector and passes through a cooler which is fabricated from aluminium and has 1 000 ft2 of surface area.   This slow circulation of the reflector prevents stagnation and stratification.

A small by-pass flow of 5 gal/min is passed through a demineralizer to maintain the chemical purity of the water.

Turbine Generator Plant

The turbine (Fig. 13) is a single-cylinder 3 600 rev/min unit with 15 pressure-compounded stages.   Steam is supplied from the steam drum at 400 psig dry and saturated and the exhaust pressure is 1.5 in Hg abs.   Three extraction points are provided for feedwater heating.   The maximum capability of the turbine is 22 000 kW and the turbine cycle heat rate at this level is 12 550 Btu/h.

Fig. 13. - The single flow, impulse type, turbo-generator is rated at 20 MW with a max. capacity of 22 MW.

Special drainage belts are provided in the wet area of the turbine and the collected water is led either to a heater belt or to the exhaust chamber.   In addition, all the steam is taken after stage 8 and passed through an external moisture separator.   Practically dry steam is returned to the low-pressure section of the turbine which, in turn, is exhausted to the condenser with 11.5% moisture content.

The condenser is a 22 000 ft2 two pass central flow deaerating unit.   Two full-duty motor-driven air pumps are provided for evacuating the condenser and there are two full-duty vertical two-stage 550-gal/min extraction pumps.

The feedwater heating plant consists of one low-pressure heater with drain cooler, a glands condenser and two high-pressure heaters which raise the final feedwater temperature to 300oF.   Three 50% duty boiler feed pumps, each rated at 150 000 lb/h, pump the feedwater to the boiler steam drum which mixes with the circulating boiler water.   The steam drum contains cyclone separators which detrain moisture from the steam before it leaves the boiler.  The steam flow to the throttle at 22 000 kW electrical output is 296 000 lb/h.

The generator is rated at 20 000 kW at 0.85 power factor, 13 800 V three phase, 60 cycles with main and pilot exciters.   The generator is of the synchronous type with air cooling.   The voltage is stepped up to 115 kV by a transformer for distribution purposes.

A reject condenser with a 200 000-lb/h capacity is connected in parallel with the turbine.   This enables the reactor to operate independently of the turbine, which is a convenience in operating the reactor during the initial start-up period and in providing an alternative load for the reactor for short periods when the turbine is off line.

Control System

The station is fully controlled from a central control room (Fig. 14), which contains the control switches, the indicating or recording instruments, the indicating lamps, annunciators and plant communication facilities.   Those controls requiring frequent attention by the control operator are located on a control console, while those requiring less frequent attention are located on panels which are grouped around the console.

Fig. 14. - NPD's main control room is adjacent to both the reactor and turbine halls.

The significant features of the NPD control system are: -

  1. The protective and regulating systems are substantially independent of each other
  2. All critical trip instruments and the regulating system are triplicated and are connected in coincidence and work on the two out of three principle.
  3. It is possible to disconnect the equipment on one channel for maintenance while the reactor is operating.   It is also possible to check out the equipment one channel at a time without causing a plant shut-down.

Station Regulation

The reactor regulating system (Fig. 15) controls pile reactivity, provides an automatic means for starting up and shutting down the reactor, and controls the reactor at a demanded power level.   This system is triplicated and is connected to the helium control valves which regulate the rate of rise of the helium differential pressure and thus the moderator level in the calandria.   Three modes of regulation may be selected - low moderator level, low log power, and steam pressure.   Low moderator level is used during start-up with an empty calandria and covers the period for the moderator to reach a level sufficient to cover the ion chambers, which then begin to operate.   Low log power control is used to hold the power in a low power state, three decades below full power.   Steam pressure control is used for all loads from 10% to 100% full power.     This control automatically determines the neutron power set-point.

Fig. 15. - The reactor regulating system, which controls pile reactivity, provides an automatic means of starting up and shutting down and it will also keep the reactor at a particular power level.

The reactor regulating system is designed to ensure that the following plant variables remain within safe limits;

  1. Neutron power level.
  2. Coolant outlet temperature.
  3. Rate of change of coolant temperature.
  4. Rate of change of log neutron power.
  5. Coolant outlet temperature from each fuel channel

The steam cycle regulating system protects the turbine equipment from excessive low or high steam pressures.   It provides a means of changing plant power output and controls steam pressure when the reactor is tripped or reduced to low power control.   This system is connected to the turbine throttle valve via a load controller and to the reject condenser throttling valve via a pressure controller.

The turbine load controller is set with demanded power output and loading rate.   The load controller then automatically loads the turbine by adjusting the turbine governor which, in turn, opens the throttle valve.

The steam pressure to the turbine is set on the pressure controller which opens or closes the throttling valve to the reject condenser.    If the reactor is tripped, the reject condenser is unloaded first to maintain steam pressure and turbine speed and generator synchronization as long as possible.

Station Protection

The main purpose of this system is to protect operating personnel, process equipment, the power plant in general and the public from harm or damage due to malfunctioning of the reactor.    The protection system supervises such critical system variables as neutron flux, reactor period, coolant temperature, coolant flow, and when these variables exceed preset limits the reactor is automatically shut down.  The system is connected to the helium dump valves which open when a trip situation occurs; this allows the moderator to drain into the dump tank.   The rate of reactivity change due to moderator dumping is greater than 3 mk in the first second but after 5 s the reactivity will have decreased by more than 58 mk.

The critical instruments are triplicated in a manner similar to the regulating system.     The dump valves are arranged in a series/parallel arrangement with two valves in series in each of three parallel lines.   The opening of a trip contact due to a signal from any one instrument will open two dump valves in two different parallel lines.    Therefore no dump takes place.   However, if a second instrument opens a trip contact the remaining four valves will immediately open, causing a moderator dump.

The turbine generator is provided with the usual protective equipment, including an emergency stop valve, overspeed trip relays, shaft vibration, etc.

Electrical System

The electrical system has been kept as similar as possible to that found in the average thermal power station.   It has been designed to have a high degree of reliability to assure reactor safety during all phases of operation, including periods when the reactor is shut down.

Power is generated by the station at 13.8 kV.   A 13.8/115 kV step-up transformer will connect the station with the Ontario Hydro system.   Two breakers separate the NPD generator from the transformer, and the station service take-off is between the two breakers; thus, the station auxiliaries may be powered from either the turbine-generator or from the Hydro system.   There will be two three-winding station service transformers to provide 2 400 and 600 V.   Power at 2 400 V will be supplied to the main heavy-water circulating pumps and other large loads.   Diesels and batteries supply emergency power for essential loads under shut-down conditions if Hydro power is unavailable.

There are three station service buses, numbered in order of decreasing reliability.   The class four buses are 600 V and 2 400 V a.c. fed from the station service transformers.   The class three bus will normally be fed from the class four 600-V bus, but on failure of the class four bus will be supplied from a diesel generator set.   The class one bus is a d.c. bus normally fed from the class three bus through rectifiers, but on failure of the normal source it will be supplied from batteries; it will supply all the station d.c. loads.

Waste Disposal System

Solid wastes from NPD may consist of spent fuel, damaged equipment parts, or other material.   All material from the active section of the station will be suitably wrapped or boxed to confine the real or potential contamination, and will be sent to the AECL disposal area.   Inactive items may be disposed of by normal methods.

Liquid wastes from NPD will be handled in a manner similar to that used for other reactor installations.   Liquid wastes known to be inactive will be discharged to the river.   Domestic sewage will be treated and ultimately discharged to the river.    Wastes which may be active will be monitored; if the level of activity is less than 10-6 micro c/ml it will be released to the process drain.   Liquid wastes too active for this method of disposal will be collected in storage tanks and transported by truck to the Chalk River disposal area.

The liquid wastes from the reactor zone are sampled for heavy water content.   If heavy water is present it is recovered and the remainder disposed of by the above methods.

Ventilation System

The ventilation system is designed to handle both the normal requirements of a conventional station and the special requirements of the nuclear components in a single integrated system.   The conventional requirements are those involving air changes in working areas, and control of the ambient temperature and humidity.   The special requirements involve control of the movement of airborne radioactive particles, the recovery of heavy water escaping from the process, and heat removal from some of the process components and their structures.

The system has been divided into five main circuits to provide for the special requirements for heating, cooling and heavy water recovery.   They are the main ventilation circuit, office area circuit, control and relay room circuit, boiler room circuit and reactor vault circuit.

The boiler room circuit and the reactor valve circuits are capable of moisture removal and collection for heavy water recovery.   The reactor vault is a closed circuit with provision for removal of nitric oxide and active particles from the air.

All exhaust air from the station is vented to the atmosphere through the 150-ft stack after passing through filters which remove all particles down to 0.3 micro m in size.


The reactor-boiler circuit is enclosed within a massive reinforced concrete combined structure and shielding walls.   The reactor vault room is designed to withstand an internal pressure of 10 psig and the boiler room to withstand 5 psig.   To keep the pressures within these limits in case of a serious rupture in the cooling circuit, the station is equipped with a combined pressure relief duct and dousing system.

The dousing system is connected to a storage tank of light water located outside the building which provides a static head of water.   The reactor vault pressure may be kept below 10 psig in the case of a header failure by the condensing action of the dousing system.   However, to keep the boiler room pressure below 5 psig in the case of the fracture of a 16-in pipe, it is necessary to provide a pressure relief duct to cope with the initial rise in pressure before the dousing system is fully operative.   This duct is normally sealed with a rupture type diaphragm designed to fail at 1.5 psig.   Downstream from the diaphragm is a horizontally hinged steel gate which is held open so that it will fall shut when tripped by the rupture of the diaphragm, thus sealing the duct to the escape of gross quantities of radioactivity.

Two biological shields are provided: a shut-down shield, equivalent to 4½ ft of heavy concrete, around the reactor vault and an operating shield equivalent to 2½ ft of heavy concrete around the process equipment.   The Administrative Wing (see Fig. 1) on the west side houses the offices, lunchroom and first aid rooms.    Normal personnel entry and exit to the building are through this wing.   The Control Wing, between the Administrative Wing and the Main Section, contains the control room, control laboratory, heating plant, change rooms, and other service rooms.    The Service Wing on the east side houses facilities for equipment maintenance, storage of spare parts, and the ventilation ducts, fans and filters.


At the end of December, 1961, the major construction phase was concluded and system testing was in full swing.   Criticality was first attained on April 11, 1962, and first steam of nuclear origin was produced on May 8.   The first electrical power was fed to the system on June 4, and on June 28 first full-power generation of 20 MW(e) gross was attained.

Primary Heat Transport

Before loading the reactor with fuel it was necessary to clean and dry the system thoroughly to prevent deposition of dirt on the heat transfer surfaces and to avoid contamination of the heavy water with light water vapour.   The drying was carried out by evacuating and flushing with helium.   The heavy water coolant was admitted immediately after evacuating and was used to clean the system.   The pH was raised to 10 with lithium hydroxide; hydrazine was added to serve as a low temperature corrosion inhibitor.

After hydrostatically testing the system, the system temperature was raised using the primary pumps as a heat source.   After three days of circulation and filtration the impurity level was reduced to 0.05 ppm, which was considerably better than the anticipated impurity of 1 ppm.   The choice of carbon steel for the primary piping has raised no problems in operation to date.

Fuel Loading

The fuel loading was carried out with the system at pressure and was fully completed with the fuelling machines.    The fuelling machines have also been used to remove and reload irradiated fuel with the reactor pressurized but not at temperature.   At the time of writing no fuel-changing operations had been carried out with the reactor at temperature and pressure but such operations have been carried out in the laboratory.

The total fuel charge was loaded in 19 days, which included a two-day interruption for servicing the machine heads.

Heavy Water Security

No long-term loss rate has yet been established but early operation has given favourable results.    The leak detection system has given early warning of several leaks and provided an opportunity to minimize the leakage.

In the first five months of operation the leakage has been less than 1 000 lb, or about 25% of the budgeted first-year leakage of 4 000 lb.   Most of these leaks occurred due to specific incidents which are not likely to be repeated so the performance to date is very encouraging.

Station Start-up

It was found during the start-up tests in confirmation of earlier calculations that sufficient energy would be introduced into the main circulating pumps to raise the temperature of the system to about 300oF and to produce steam in the boiler.   It has been found that with the system at 300oF the reactor may be started, brought to full power and the turbine loaded in 100 min.   After a scram the reactor can be restarted against xenon poisoning in up to 33 min with equilibrium conditions.

The reactor has performed very well and, in the opinion of the operators, is much easier to operate than a conventional steam station of similar capacity.

At the time of writing (August) the station was undergoing a comprehensive shake-down programme to correct difficulties which have shown up for the first time after the full-power operation; for example, the moderator charging pumps are being modified to correct a loss of prime situation occurring because of high gas concentration in the moderator circuit.

It is expected that by early autumn the station will be in full power operation on a regular basis at 80% load factor.


NPD is a demonstration plant.   As such it will not generate low-cost electricity nor was it designed to do so.   Low-cost generation will, however, be achieved by larger plants rated 200 MW(e) and upwards, such as Douglas Point, now under construction and scheduled for operation in 1965.   In such large ratings, total unit energy cost will be sharply competitive with that from a large conventional fossil fuelled plant.

  NPD HWR-80
Primary Circuit
   Inlet, oF
   Outlet, oF
   Flow, lb/h

   Horsepower (each)

Heat Exchanger
   Area, ft2
   Mean length of tubes, ft

Secondary Circuit
   Steam flow, lb/h
   Temperature (TSV), oF
   Pressure (TSV), psia
5.14 x 106


6 393
2 099

296 000
10.78 x 106


3 x 13 350
3 x 6 242
3 x 26

1 110 000

A review of the design and performance of NPD by CGE, AECL and a Canadian consulting firm, Montreal Engineering, has shown that the power output may be increased to 80 MW for a second plant built to the same general design as the NPD station for an increase in cost of about 25% (see Table 1).   Although the unit energy cost of about 9 mill/kWh is, perhaps, twice that expected form larger units, it is sufficiently low to be of interest in high fuel cost areas.   Four parameters, summarized in Table 2, have been changed to make this increased power possible.

Parameter NPD HWR-80 Ratio
Maximum fuel rating, kW/cm
Radial flux factor
Axial flux factor
Combined radial and axial factors
Number of fuel channels
Gross station efficiency
Electrical output (MW(e) gross)

Fuel Rating

In 1959, when the fuel design for the first NPD fuel charge was finalized, the maximum rating of UO2 fuel, based on the state of knowledge at the time, was considered to be:

∫ kdθ = 35 W/cm, for Ts = 400oC.

Further research and development work has now been completed on UO2 fuel and the acceptable design rating is now considered to be:

∫ kdθ = 40 W/cm, for Ts = 400oC.

This increase in rating has been achieved by using higher UO2 densities and closer control of UO2 microstructures which are important factors in limiting the fission gas release.   Improvements have also been made in the quality of Zircaloy sheathing available from the suppliers and in the non-destructive testing techniques for fuel sheathing, which are important because of the increased stress imposed on the fuel sheathing.   Irradiation tests at Chalk River have established the validity of these higher fuel ratings.   The increase in fuel rating will permit the power to be increased 2.20 times that of the NPD core.

Radial Flux Flattening

In the up-rated NPD it is proposed to achieve a degree of radial flux flattening by having a two-zone core with a different average fuel burn-up in each zone.    Combined with a slight increase in the axial flux factor, the power will be increased to 1.255 times that of the NPD core.

Number of Fuel Channels

It has been found feasible to add 32 more fuel channels to the existing 132 in the calandria vessel and thereby the power will be increased to 1.24 times that of the NPD core.   This will result in two minor mechanical changes.   The tapered inner wall of the calandria, which acts as a barrier between the light water reflector and the heavy water moderator, will need to be moved out at the ends of the taper by approximately 20 cm.   Also, the stainless steel end fittings at both ends of the coolant tubes will need to be increased in length by 5.7 cm to permit room for the extra feeder pipes.   Neither one of these changes requires new development work.

Thermal Efficiency

Coolant pressures and temperatures have been increased to the levels used in the design for Douglas Point, which permits a 25oF increase in steam temperature.   Furthermore, the HWR-80 steam cycle would use five stages of regenerative feedwater heating together with live steam reheat rather than the three feedwater heaters without reheat used in NPD.   Altogether these factors permit a net gain in turbine cycle efficiency.   With Canadian cooling water conditions NPD thermal efficiency is 0.264.   With the HWR-80 and similar conditions the efficiency would be 0.324.

Mechanical Design Changes

In up-rating the original NPD design it is obviously necessary to make changes appropriate to the increased output.   In general such changes are relatively simple and involve only extrapolation of proved designs largely in the conventional field.   Modifications would also be necessary to the fuel handling system.   In the fuelling machines, mechanical changes are necessary in the hydraulic ram system to increase the length of stroke to meet the increased length of end fittings.   The carriage and vertical telescope system require design changes to extend horizontal and vertical travels to cover the added sites on each dimension.   Such design changes are easily made and no development work is necessary.

The increased neutron flux in which the machines would operate combined with heavier fuelling duty cycles poses a slightly more serious problem concerning the operation and life of components incorporating organic materials, such as seals, pneumatic and hydraulic valves, electrical devices, wires, cables, and hoses.  The solution lies in:

  1. Replacement of organic by metal seals.   An active development programme already exists to provide metallic seals for the NPD fuelling machines and the results will be applicable to the up-rated design.
  2. Reduction of exposure time by speeding up the refuelling operation.   Although the controls and drives for the NPD fuelling machines were designed to minimize the time cycle of the fuelling operation, some scope exists for a further reduction.

Foldout Supplement:
Technical details (58 kB)
Sectional views (183 kB)

[CNS home page]