ZEEP: The Little Reactor that Could

by R.E. Green and A. Okazaki
from CNS Bulletin Vol 16 No. 3, Autumn 1995
Ed. Note: Ralph Green, a former vice-president of AECL Research, gave an oral version of this paper as one of the three invited talks for the Nuclear Heritage celebrations held at Chalk River, August 4-6, 1995, to mark the 50th anniversary of ZEEP, Canada's first nuclear reactor.   He repeated the talk at AECL Sheridan Park in September.   Co-author Al Okazaki is a former senior scientist at AECL Chalk River Laboratories.   Both worked with ZEEP in the 1950s.

On September 5, 1945 the ZEEP reactor went critical for the first time at the Chalk River Laboratories of Atomic Energy of Canada Limited.   ZEEP (for Zero Energy Experimental Pile) was the first reactor to operate outside the USA.   In this paper we recall some of the events that led to the construction of ZEEP, and briefly describe the role it played in the development of the Canadian Nuclear Program.

ZEEP: conception to criticality
The first attempt to achieve a self-sustained nuclear chain reaction in Canada was made by George Laurence, assisted by B.W. Sargent, working at NRC during the years 1940-42.   Their pile consisted of sacks of uranium oxide interspersed with sacks of powdered coke.   Their attempt failed mainly because of impurities in the materials they were using, although it would have been very difficult to achieve a critical assembly using natural uranium oxide and graphite, even with pure materials.

In 1942 it was decided to move the UK Nuclear Energy Program to Canada, and a joint Canada-UK laboratory was set up in Montreal in the fall of 1942.   The work in Montreal, described in a pamphlet entitled Early years of nuclear energy research in Canada, by George Laurence, led to the decision, in mid-April, 1944, to build a natural uranium fuelled, heavy-water moderated reactor, what we know today as NRX.   The design of NRX was based on theoretical calculations backed up by subcritical experiments in the Montreal laboratory with lattice arrangements of natural uranium metal and heavy water.

In late April 1944, John Cockcroft came to Canada to lead the Canada-UK program.   In May 1944, Cockcroft decided it would be desirable to have some operating experience with a low-power reactor like NRX before the latter was built, and to have the capability to alter the reactor core to investigate the effect of changes to the lattice arrangement.

The main reasons for building such a reactor were that it could be constructed quickly and the experience gained during the construction and operation would be valuable for NRX.   It could also be used to measure some materials properties, and to test control, safety and radiation-protection equipment.

So, in July 1944, Cockcroft asked two of his staff to look at the possibility of building a low-power reactor without seriously impeding the NRX project.   In August 1944, approval was received to proceed with the design, and Lew Kowarski, newly arrived from the UK, was asked by Cockcroft to manage the project.   Charles Watson-Munro was Kowarski's second in command, and they were assisted by A.M. Allan, F.W. Penning, G.J. Fergusson, C.W. Gilbert, E.P. Hincks, H.F. Freundlich and H. Carmichael.   The chief designer was George Klein from the NRC Mechanical Engineering Division at Ottawa.   He was ably assisted by Don Nazzer, also of NRC.

During the design phase there was pressure from the research staff for a reactor power of 1 kilowatt, rather than 1 watt, because this would provide neutron fluxes high enough for good cross-section measurements for the chemists to prepare good radioisotope sources, for the engineers to study material properties and for radiation protection work to be done.   However, such a power level would require more shielding to protect the operators, and would preclude the rapid rearrangement of the core to study different lattice configurations.   So, the power level was kept at 1 watt.

Final approval for the construction of ZEEP was given on October 10, 1944.   Construction was complete by September 4, 1945, and the reactor went critical on September 5, 1945 at 3:45 pm, only 16 months after conception and only 11 months after approval of construction.

ZEEP Building on April 10, 1945

One might wonder how long it might take to achieve that today.   Of course, this was before the creation of the Atomic Energy Control Board!

The critical height of the heavy water in the ZEEP reactor was 132.8 cm, compared to the calculated level of 128 cm.   This excellent prediction was made by John Stewart, a long-time AECL employee, working with George Volkoff, who later went to the university of British Columbia.

As noted above, ZEEP was the first reactor in the world to operate outside the USA, and it was a great achievement for the Canada-UK team.   However, it is important to acknowledge the contribution made by the US, in providing key materials, and information from the operation of the CP-3 heavy-water research reactor at Chicago.

Early operation of ZEEP: 1945-47
Once criticality had been achieved, a busy schedule of experiments commenced, and continued up until early 1947, when ZEEP was shutdown so that its heavy water could be used in NRX.   Space limitations preclude our listing all of the experiments done during this initial operating period, but the major ones were as follows:

  • measurement of the buckling, or overall reactivity, of the ZEEP lattice
  • measurement of the relaxation times for various subcritical and supercritical conditions, to determine heavy-water reactor kinetics reactivity
  • measurement of intensities and lifetimes of delayed neutrons and delayed photoneutrons, important for reactor control
  • calibration of ion chambers for the NRX control and safety systems
  • measurement of the reactivity effects of various control rod configurations, including interference effect between rods
  • measurement of the neutron absorption of various nuclear materials, e.g. samples of graphite and uranium for the UK reactors, and thorium for the NRX J-rod annulus, where it was planned to produce U-233
  • various nuclear-physics experiments, e.g. The measurement of gamma-rays emitted during fission, and a search for the negative proton
  • determination of ETA (number of neutrons emitted per neutron absorbed) for U-233
  • neutron activation of various samples for radiochemical studies.   One of these experiments determined the radioactivity produced in Ottawa River water, which enabled an estimate to be made of the activity to be expected in the NRX cooling water
The people involved in these first experiments were: J.G. Bayly, S.W. Breckon, A.J. Cruikshank, F.J.M. Farley, F.W. Penning, G.J. Fergusson, K.D. George, C.W. Gilbert, H.E. Gove, M.W. Johns, I. Kowarski, B. Kinsey, D.J. Littler, B.W. Sargent, I. Siminovich, A.G. Ward, C. Watson-Munro and D.H. Wilkinson.

ZEEP in February 1954 with NRX and NRU (under construction, in background.

Since ZEEP initially had no shielding outside the graphite reflector, it had to operate at first at a fraction of a watt, to protect the operators.   Later on, tanks of ordinary water were stacked around the reactor, wood was placed on top and a small room of masonite and steel blocks was built to house the operators.   In this way the power could be raised to 50 watts for brief periods.   During this first phase of operation ZEEP operated around the clock, except for Sundays, when the reactor was shut down at 7:30 am, presumably to give the staff time to get to church, or to go sailing, or play tennis!

ZEEP was shut down in April 1947, and its heavy water was transferred to NRX.   Much was accomplished during this first period of operation, and much of it was relevant to the operation of NRX.   However, no experiments were done to study the effect of changing the lattice arrangement, one of the original reasons for building ZEEP.   Perhaps there were too many other experiments to be done, and since the ZEEP critical size had been accurately predicted, it may have been decided that the more time-consuming lattice experiments were not required at that time.   These would come in the next phase of operation.

Second period of operation: 1950-56
The ZEEP program started up again during the period April-August 1950, under the leadership of A.J. Pressesky.   During the shutdown new side shielding had been provided so the reactor could now operate at higher power levels, and improvements had also been made to the control system.

The focus for the experimental program now was support for the new reactor NRU, then being planned.   Experiments were done with different numbers of NRU rods and the results were used to optimize the lattice spacing and overall core size for NRU.   Other experiments were done to measure the reactivity effects of empty fuel channels and the split lattice used in NRU to provide the horizontal through tubes for neutron beam research.   Other NRU-related studies involved measuring reactivity effects and neutron flux perturbations due to insertion of guide tubes and various control devices.

At this stage in our power-reactor evolution it was believed important to extract the maximum amount of energy from natural uranium fuel, and to do this would require recycling the plutonium produced in the original fuel.   This led to experiments in ZEEP with close-packed lattices that might be used as a blanket around a reactor core to produce plutonium.

Cutaway drawing of ZEEP as it was in 1950

There was also interest in power reactor cores with fuel rods containing large amounts of uranium, so experiments were done with 3-rod clusters of ZEEP rods to investigate this concept.

In another experiment the temperature coefficient of reactivity for the ZEEP core was measured by heating the reactor to 80 degrees Celsius.   Measurements of the temperature coefficient of uranium were also made, using the "swing" method, in which samples of heated and unheated uranium were alternately inserted into equivalent positions in the reactor core.

Other experiments were done with Pu-Al rods prepared by John Runnalls and co-workers.   This type of fuel was being considered for use in NRX and NRU.

ZEEP was also used during this period by scientists from the UK to measure the properties of fuel rods to be used in a proposed UK heavy-water power reactor.

Near the end of this period lattice experiments were done with 19-rod clusters of uranium metal, similar in size to those used later in NPD and Douglas Point.   This fuel was produced before it was clear that uranium oxide would be the eventual fuel for CANDU reactors.

The key players during this period of operation were D.H. Allen, W. Dickerson, D.W. Hone, J.H. Moon, A. Okazaki, R.M. Pearce, L. Pease, A.J. Pressesky and D.H. Walker.

The second period of operation was now coming to a close as plans had been made to shut the reactor down for another upgrade.   There were several weaknesses in the system that needed fixing.

One was that there was no way to drain heavy water from the reactor at the control desk.   The reactor was normally started up by pumping heavy water into the reactor tank to a level at which the power would increase at a fixed rate.   When the desired power level was reached water had to be drained from the tank to achieve operation at steady power.   However, the drain valve was located at the side of the reactor, 10 to 15 feet from the control desk.   So, one operator had to manipulate this valve on instructions from a colleague watching the power meter at the control desk.   (It should be noted here that the scientific and technical staff were also the operating staff.)

The shielding for the top of the reactor was also primitive compared to today's standards.   There were tanks of boron-loaded paraffin that could be placed on the reactor lid, for operation at high power, but since lifting these was no fun the tendency was to operate as much as possible at low power, or for short periods at higher power.   Once when ZEEP was operating without the shielding in place the NRX reactor tripped due to high neutron flux in the NRX reactor hall.   After that, we were asked to inform the NRX operating staff when ZEEP was going to operate.

There was also a problem with the ZEEP shutoff rods.   These were attached to cables wound on drums mounted on the rod-support beams.   Sometimes when these rods were dropped to shut the reactor down the cables would jump off their drums.   While this wasn't a safety concern, it did delay the experimental program.

There is one anecdote from that period that readers might find interesting.   To pump water into the reactor tank one had to push a button at the control desk to start the pump.   However, the pump ran only for a fraction of a minute at a time, and then stopped.   So an operator had to repeatedly push the button to keep the pump running.   Since this was rather tedious, one operator made a block of wood that could be used to jam the pump button so the pump would run continuously.   One day, a couple of researchers were on the top of the reactor inserting detectors, and an operator was at the control desk pumping up the heavy water, with the pump button jammed.   Suddenly, the phone rang at the other side of the building and the operator left the control desk to answer it, leaving the pump running.   The call took longer than expected and the next thing the researchers heard was the shutoff rods dropping into the reactor.   The reactor had tripped on overpower.

No one knows how much radiation the researchers received since they had left their film badges in their coat pockets on the floor below!   However, it couldn't have been too much since one of the researchers' wives later had a healthy baby.   One might deduce from this that "a little neutron flux never hurt anyone".   This incident was never reported to senior management.

ZEEP was shutdown for several months at the end of 1956.   A new rolling shield for the top of the reactor was installed, as well as new control and safety equipment.   The latter was similar to the instrumentation to be used in NRU, so once again ZEEP was used as a test bed.

Third period of operation: 1957-68
ZEEP started up again in the spring of 1957.   The first series of experiments involved a core of 55 19-rod clusters of uranium oxide.   Although the density of the oxide was lower than that used later in the power reactors, it nevertheless enabled us to obtain the first lattice physics data for uranium oxide fuel.   One experiment involved heating the whole reactor to 65 degrees Celsius to determine the overall temperature coefficient.

Later we acquired a full loading of 7-rod clusters of the original NPD uranium oxide fuel for another series of experiments.   This fuel was in the form of 50-cm long bundles, another first for ZEEP.   Tests were done with heavy water and air coolants, which gave valuable information on the reactivity effect of a loss of coolant, information that was important for the design of CANDU safety systems.

In September, 1960 the ZED-2 reactor started up, and from that point on most of the full-scale lattice experiments were done there.   ZED-2 was large enough that experiments could be done with complete fuel-channel assemblies, i.e. with pressure and calandria tubes. ZED-2 has provided a wealth of information for the CANDU program, and is still doing so today.

However, the role of ZEEP was far from over.   A hot loop was installed at the centre of the reactor and was used to measure detailed neutron-spectrum effects in CANDU fuel at elevated temperatures, in an attempt to more closely simulate conditions in the power reactors.

During this period a series of experiments was done to check the feasibility of determining lattice parameters by using a small number of fuel assemblies located at the centre of a large core of different assemblies.   This substitution technique was of great interest since it would, if feasible, reduce the amount of new fuel required for such work in the future.

Many other valuable experiments were done in ZEEP during this final period of operation.   Some of the more significant ones were:

  • measurement of the reactivity of several NRU fuel assemblies, in an attempt to explain a loss of 7 mk in reactivity when new fuel was introduced in NRU.   The reactivity loss was found to be due to boron in the aluminium coolant tubes
  • measurement of flux peaking at the gaps between the ends of adjacent CANDU fuel bundles.   The fuel engineers were concerned about fuel overheating at the bundle ends
  • a comparison of the neutron absorption of samples of Zircaloy, Zr-Nb and Ozhennite, prospective pressure-tube materials
  • irradiation of sulphur capsules for the commercial products division of AECL to explore ways to enhance the production of P-32
  • tests of self-powered flux detectors being developed by J.W. Hilborn
  • the reactivity of Douglas Point type fuel bundles for the CANDU reactors in India
We are now up to the end of 1968, and from here on ZEEP was used sporadically, as all of the lattice physics work was being done in ZED-2.   From this point until its final shutdown the reactor was used mainly by university students for post-graduate projects.   ZEEP was shut down for good on July 27, 1970, after almost 25 years of outstanding service.

The major players in this last phase of operation were D.H. Allen, G.A. Beer, C.B. Bigham, D.S. Craig, B.G. Chidley, W. Dickerson, R.E. Green, K.J. Hohban, D.W. Hone, B.A. Maciver, A. Okazaki, R.J. Patterson, D.J. Roberts, L.P. Robertson, K.J. Serdula, P.R. Tunnicliffe, R.W. Turner, D.H. Walker and S. Yewchuck.

Other R & D programs
As noted at the outset this has been the story of ZEEP.   However, lots of other things were happening in parallel with the ZEEP activities, and since ZEEP was shut down, which contributed greatly to the success of the Canadian Nuclear Program.

Experiments in the NRX, NRU, and WR-1 reactors, and work in AECL's laboratories and Canadian industry, were used to develop the key features of the CANDU reactor system, such as:

  • highly successful fuel and on-power fuel-handling systems
  • fuel channels, comprising the pressure and calandria tubes
  • heat-transport system components
  • systems chemistry and corrosion-control methods
  • heavy-water production technology
  • computer control and CANDU-specific instrumentation
  • reactor safety technology
  • radiation-protection equipment
  • methods for the handling and disposal of both low-level and high-level nuclear waste
Each of these areas has a great story to tell, and these can be found in a book entitled "Canada enters the Nuclear Age", the technical history of AECL as seen from its research laboratories.   This book is currently in the press and should be available early in 1996.

In this paper we have tried to take you back in time to the early days of the Canadian nuclear program, and to give you a summary of the history of ZEEP, whose 50th anniversary we are celebrating this year.   We hope you will agree that while ZEEP was a small reactor, it was a very versatile one, and made a large contribution, out of all proportion to its size, to the Canadian Nuclear Program.

It represented the first self-sustained nuclear chain reaction in Canada, the first outside the USA, and launched us on the road to CANDU the best power reactor system in the world.

Therefore, we hope you will agree that ZEEP deserves the sub-title given at the outset: "the little reactor that could".   Indeed, perhaps this sub-title should be expanded to "the little reactor that could, and did".

Top view of ZEEP after 1956 upgrading

Canadian Nuclear Society home page

CNS/SNC Webmaster