SNAP Experimental Reactor (SER)

The SNAP-2 Experimental Reactor (S2ER or SER) was built to verify the core geometry and basic reactivity controls of the SNAP-2 reactor design, as well as to test the basics of the primary cooling system, materials, and other basic design questions, but was not meant to be a good representation of the eventual flight system. Construction started in June 1958, with construction completed by March 1959. Dry (Sept 15) and wet (Oct 20) critical testing was completed the same year, and power operations started on Nov 5, 1959. Four days later, the reactor reached design power and temperature operations, and by April 23 of 1960, 1000 hours of continuous testing at design conditions were completed. Following transient and other testing, the reactor was shut down for the last time on November 19, 1960, just over one year after it had first achieved full power operations. Between May 19 and June 15, 1961, the reactor was disassembled and decommissioned. Testing on various reactor materials, especially the fuel elements, was conducted, and these test results refined the design for the Development Reactor.

S2ER Schedule and Timeline

S2ER Reactor and Cooling Systems

Fuel and Core Geometry

The S2ER was built to verify the core geometry and basic reactivity controls of the SNAP-2 reactor design, as well as testing the basics of the primary cooling system, materials, and other basic design questions, but was not meant to be a good representation of the eventual flight system. The fuel itself was 6.88% (by weight) 235U, enriched to 93%, with the rest being ZrH(1.07) measuring 10”x0.975”x0.975”. Moderator strips were included in each fuel element, made out of beryllium, with a 1 1/2” solid Be plug on each end, surrounded by a 1” welded stainless steel can with a 10 mil thick wall and 1/2” endcaps. The fuel-moderator-can element assembly measured 14”x1”x1”, not including central axial pins for fuel element placement at each end. The core itself was made up of 61 of these fuel/moderator elements, arranged in a triangle lattice pattern to form a hexagon that was approximately 8” across flats by 9” across corners. The core contained a total mass of 3 kg of 235U.

The external reflector was made up of multiple parts, roughly divided into a hexagonal shape surrounding the core. Three flat parts were Be reflectors, made up of individual plates 1/4” thick. There were two reasons for the plates: first, the number of plates could be varied to control the excess reactivity in the core, and second, the plates could be removed by the safety system (this reactor’s equivalent to a scram rod). The other two parts surrounding the reactor worked together: the control drums were made up of partial cylinders of Be, which would rotate away from the core to control reactivity in normal operation, and a set of specially shaped “shims” filled the gaps between the core and the control drums when fully rotated to face the core. End-cap reflectors were made up of the aforementioned fuel element end plugs, as well as the NaK plenums and end grids to support the fuel elements.

Due to the reactivity of the reactor, a Pu/Be neutron source was added to one of the control drums to initiate criticality of the reactor during a test, with an activity of 1.68×10^6 n/sec.

Reactivity was controlled by rotating the drums with a direct motor-drive gearing, set to a maximum reactivity insertion of 2.5 c/sec, with each drum (at beginning of operation) had a total worth of $3.82 over its 180 degree range. Scram functionality was provided by removing the flat reflector plates, mounted on hinges at the bottom of the core with an electromagnetic release (through power interruption to the magnets) at the top. Each plate, when removed, had a worth of about $5.40 in reactivity, and the scram function took approximately 240 ms to complete. All drive mechanisms were behind secondary shielding to prevent gamma and neutron irradiation complications.

The reactor vessel itself was constructed out of ¼” carbon-steel boilerplate, 15’ 5 1/8” deep, with a variable inner diameter stepping three times from 48” at the top to 38” at the bottom. The vessel was filled with N2 at a slight overpressure to prevent oxygen fires from either the NaK or the hydrogen leakage from the ZrH. Water coolant pipes were installed surrounding the containment vessel, both to cool the vessel and to prevent overheating to the concrete surrounding it. A lead shield resting on the installed thermal shielding, topped with a shutdown shield of borated concrete that extended up to the second step, was installed just above the core, with the safety drives and control drives installed above this. The NaK coolant pipes ran through this shield as well, before exiting the containment vessel. A concrete biological shield was installed within the containment vessel at ground level, resting on the highest step, followed by the end cap for the containment vessel above ground level.

Cooling Systems for the Reactor and Associated Installation

The coolant system used NaK-78 (78%K by mass) as the working fluid, in a two loop system with fill-and-drain tanks in both loops, pressurized by He, to allow for complete core draining of coolant and emergency shutdown. The primary loop had an inlet temperature of 1200F and an outlet temperature of 1000F, while the secondary loop operated at 950F-1150F. A counterflow heat exchanger rated to 100 kW was used to transfer energy between the two loops. The secondary loop was pressurized 1 psi over the primary, so any coolant leakage would be secondary to primary to ensure no irradiated material passed into the secondary system. Surge tanks were installed in both systems, as were permanent magnet flowmeters, electromagnetic pumps designed by Atomics International, and other equipment needed for the monitoring and operation of the coolant system.

All tubing in the primary loop was 1 1/4” stainless steel (schedule 40, type 304) except for plugging indicator and fill-and-drain lines, which were 1/2” piping. Everything except the pump and the instrumentation was thermally insulated, which required a 400 F maximum temperature.

The secondary loop rejected heat through an airblast radiator, which took air from inside the facility, ran it through a 20” diameter, ½ hp fan, then pumped it out of the building. This would take the nominal flow of 11 gpm of NaK from 1155 F to 1105 F. The pump was mounted immediately downstream of the heat exchanger, and was identical to the primary coolant pump. An electrical heater was initially installed after this, which was capable of 50 kW of heating to the NaK system if needed, however, after the original ended up failing this was replaced with another one immediately downstream of the intermediate heat exchanger. All piping was essentially identical to the primary loop, except the heat exchanger, which was 1” OD type 304 stainless, with a wall thickness of 0.083”.

Three different cover gases were used to prevent oxygen contact to either the fuel or coolant in case of a leak. The containment vessel used N2, the primary loop used He, and the secondary loop used Ar (the primary loop originally used Ar as well, until the irradiation of the gas bacame a concern, leading to its replacement with He). Occasional N2 venting was necessary for moisture control, this used the same discharge system as the Ar valve to allow for the decrease of radioactivity to acceptable levels.

The containment vessel and surrounding concrete were kept cool using a water coolant system. This had two parallel pipes of 1/2” schedule 40 mild steel, with each measuring a total of 150 ft. A second system of 3/4” pipes was used to cool the outside of the vault liner. During normal operation, the containment vessel pipes had a flow rate of 2.6 gpm, while the vault coolant system used 6 gpm. A forced-draft evaporative cooler provided heat rejection, and a closed cycle water distillation system prevented scaling.


Two fission chambers were used in parallel for reactor startup measurements (from 5×10^-4 to 5 watts), one linear and the other logarithmic. Once the logarithmic count rate meter showed a reasonable count rate, the control drums and/or safety elements would be moved from their shutdown positions.

From 5×10^-1 to 5×10^5 watts, two compensated ion chambers were used to cover core activity, fed into parallel log current amplifiers and period amplifiers. The period amplifiers were fed into the safety amplifier on the scram system, which was bypassed once the reactor was at greater than 10 kW to prevent unnecessary scrams at full power.

Three incompensated ion chambers were used with three safety amplifiers for high-speed flux level scrams for the 5×10^2 to 5×10^5 watt power range, with one supplying information to a recorder controller for flux level information needed by the automatic power level controller. One final uncompensated ion chamber with a micro-microammeter was in place for the operator to see the necessary flux level information.

The coolant systems were fitted for instrumentation for temperature, pressure, liquid level, and flow rate.

All of this came together on the control console, which allowed control of:

1. Control drum drive position

2. Safety element drive position

3. Primary and secondary NaK coolant pump power

4. Manual-to-automatic reactor control switch

5. Airblast heat exchanger fan

6. Emergency scram.

Meters and indicators were available on the console for:

1. Reactor period

2. Reactor power level

3. Control drum positions

4. Safety element positions

5. Primary and secondary NaK loop flow rates

6. Airblast fan speed

S2ER Operational History

In September 1959, the core was ready for its first fuel element, after completing all facility construction requirements. September 17 saw the first loading of 14 fuel elements, with dry criticality achieved two days later with the load of 55 FEs. This allowed for initial control and safety calibrations to be made, and facilitated the loading of the full 61 element load on 19 October. The next day, NaK was loaded into the reactor, and wet critical testing was completed on the same day.

This led to the first high-power operation on 9 November, which led to an intensive testing period to determine the particular behavior of this reactor under a wide range of operating conditions. The aforementioned secondary heater failed and was replaced during these tests, and after it was replaced the reactor’s thermal reactivity coefficient was determined. The #3 safety element malfunctioned, was modified, and brought back into operation, with a follow up dry critical experiment at the end of the year.

1960 started for the S2ER with a test of xenon transient behavior in the core, a cooling air line failed and was replaced, additional criticality test were done following tweaks to the core geometry and equipment, and various 600 F test runs were completed.

Finally, on February 22nd, the reactor was ready for its’ first extended full power run. This lasted until March 12th, when an equipment malfunction caused a scram in the reactor. After resetting the safety equipment, the reactor was brought back up to full power later that same day, and ran without major incident until April 23rd, when it achieved 1000 hours of continuous operation at design power and temperature. The reactor was manually shut down, and the next round of tests were conducted.

The testing that was done on the reactor included several extended tests at 600F-900F, one power test at 950F, and zero power testing 900F. This was interspersed with temperature coefficient tests, power coefficient tests, ramp insertion and transient experiments, with only brief shutdowns until October 1960, when once again the secondary NaK heater failed. This was then just removed, and replaced by a length of straight pope, to continue low temperature testing (this time at 960F), during which control simulator studies were conducted.

Perhaps one of the most interesting parts of the S2ER testing regime was the use of a Donner analog computer for reactor transient testing, with inputs for core inlet temperature, NaK flow rate, and control drum reactivity. After calibration of the mathematical model in the computer with reactor behavior, a number of tests were conducted. These included control drum calibration and reactivity behavior tests using neutron kinetic equations with results varying between predicted and actual of about 10% of drum rotation position – considered reasonably accurate enough for automatic operation. Other tests included: prompt power reactivity coefficient testing, NaK flow rate and inlet temperature variation testing, and others. By the end of the analog computer testing regime, the calibrations on the computer were accurate enough that there was almost no ability to distinguish between the computer’s predictions and the reactor’s actual behavior.

By the time the S2ER was shut down the final time on November 19th, it had operated for 10,306 hours, with 1,877 hrs (31% of total operating time) at full power and temperature, 2,290 hours (38% of total operating time) at full power but low temperature, and 1,868 (30.9%) less than full power and temperature. It produced a total of 224.6 mWhr, with a total of 4,493 hours at full power. Of the downtime on the reactor from its first criticality, 30.1% was for holidays and weekend, 29.2% of it was for routine repair and maintenance, 24.8% was for miscellaneous component failures, and 15.9% was due to the heater bundle failures in the secondary NaK loop.

On May 12th, the reactor started being dismantled, with the last fuel/moderator element removed on May 31st, with no complications (the FEs could only be removed six at a time for safety reasons, which made it a slow process). No obvious physical changes occurred in any of the fuel elements when compared to installation, and six of them were sent to AI’s Santa Susana mountain field laboratory (Component Development Hot Cell) NW of Los Angeles for post-irradiation testing.

The full round of testing on the S2ER showed that the reactor was inherently stable, with negative thermal power coefficients at all power levels up to 110% of full power. All scrams of the reactor were either initiated by the operators, or were caused by anomalous instrumentation readings, not abnormal reactor conditions during testing. The majority of the maintenance problems on the S2ER were caused by a single component, the secondary NaK loop heater, which is a piece of equipment that was only for ground testing and would not be present in a flight system. This success validated the SNAP-2 reactor design, provided information to the SNAP-2 Development Reactor that was currently under construction, and prepared the way for the first high-power reactor core to be ready for launch as part of the SNAPSHOT program, which at that point was already under discussion.

References and Further Reading

The quarterly status reports on the SNAP Reactors Program also carry more valuable details about the development, testing, and operation of these reactors, available on the SNAP main page.

Overall Program



Fuel Element Development and Analysis


Nuclear Design and Analysis


Decommissioning and Safety