RE: Evidences of a leak in the Gas System of the Raleigh Research
Reactor; Plans for Repair.

This memorandum describes in further detail the matters discussed in recent
telephone conversations relative to a leak which developed in the gas system of
the Raleigh Research Reactor and procedures which have been followed, subsequent
to its discovery, in preparation for repairs.

On Friday, May 6 the reactor was operated in a series of experiments at
various power levels up to 7000 watts. At intervals over the preceding weeks
there had been some tendency of the fuel to precipitate when the reactor was
operated at high power and moderate temperatures (i.e. below 50-60 C), and the
H2- O2 recombining action at the catalyst bed in the gas-recombiner system had
been a little erratic. (Electric heaters had been ordered for installation on
the catalyst chambers to remedy this difficulty, but they had not arrived). In
operations on this particular day, however, the reactor had performed quite satis-
factorily in every respect.

The reactivity was normal, the recombining action at the catalyst seemed
perfect, and every aspect of the reactor system seemed normal. The reactor was
closed down and placed in standby condition at the end of the day, with the core
pressure at thirteen inches of water below atmospheric ( normal is from 5 to 15
inches of water below atmospheric).

On Monday, May 9, it was noted that the pressure in the core was at atmos-
pheric. This would not have been expected, but was no cause for concern.
Removal of a small volume of gas from the core volume into the gas-disposal
holding tank quickly reduced the pressure again to the normal value. The res-
ponse in this procedure appeared normal.

Reactor operation was not scheduled for Monday, and observation was made of
the pressure behavior of the core system. During the day, the pressure increased
and Tuesday morning it was clear, from the pressure which had again risen to
atmospheric, that air was leaking into the system. Careful monitoring of radio-
activity levels indicated that the leak was in that portion of the system inside
the Safety Envelope. (A large inert atmosphere chamber enclosing the core,
control rods, etc, and a portion of the gas-circulating system). One original
purpose of the Envelope was to serve just the function encountered here:
containing radioactive gases in case a leak should occur in the core system.

No evidence could be found of liquid fuel leakage from the core.

Radioactive gas was found in rather large concentration in the Safety Envelope
volume. Pressures were equal or lower in the core, hence the gas in the envelope
was by diffusion. The amount present indicated that the leak was rather large.

Having ascertained that the leaking radioactive gas was contained in the
Safety Envelope, and that no hazard of loss or leak of fuel was involved, two
decisions were made as to procedures over the next following days:

• 1. Efforts would be made, by measurement and calculation, to ascertain
  the amount and identity of radioactive gas in the system which would have to
  be disposed of-before repairs could be undertaken, and

• 2. Efforts to repair would be deferred for a week or so. This course was
  chosen (a) in order to allow decay of the short lived activities, and
  (b) in order for repair work to be scheduled after the climax of school

  ---

  [page 3]
  
  

  activities then in progress; graduate examination, building dedication,
  commencement and terminal activities of the school year.

In the interim period, Dr. Wray calculated that a maximum of about 0.77
curies of I131 and 0.33 curies of Xe133 might possibly be in the system in
gaseous form as a result of recent prior operations of the reactor. Careful
measurements by Dr. Waltner established with certainty that the major component
of the gas was Xe133, with only a trace of I131 present. The remainder presum-
ably being contained in the fuel solution (as would be expected).

About May 22 a schedule of bleed-off of radioactive gas from the core and
Safety Envelope to the atmosphere through the building stack, was established.
The gas was diluted in the stack with 20,000 cfm of air. The activity of the
gas was continually measured, and at no time did rate of release permit maxinuum
permissible tolerance levels to be reached in the stock discharge, even if a11
the activity were I131. It was known of course, that most of the activity was
Xe133, which has a much higher tolerance level.

By June 4, the radioactive gases in the core and those which had leaked into
the Safety Envelope had been bled off to the atmosphere, to such a point that the
gases then being drawn from inside the core were only slightly above background.
The best measurements indicate that a total of less than l/2 curie, mostly Xenon,
had been released. Preparations were made to open the reactor assembly on Monday,
June 6 to seek the damaged position and make repair.

It was surmised that the leak might be located either in a plastic gasket in
the piping system, which would be relatively easy to repair, or in a welded joint
in the lid of the reactor core or in the piping system.

At this point ( June 3) another phenomenon entered the picture. A slight
increase was noted in the activity of the cooling water in the cooling system of

the reactor core. A sample withdrawn from the system revealed a very low alpha
count, in addition to the low but slightly increased beta and gamma activity
normally present. It was suspected that this night possibly be evidence of a
uranium solution fuel leak into the cooling water ( in the coils inside the
reactor core). The amount of activity present was not sufficient to give clear
indication whether this was the case or not. A variety of other measurements
were therefore made:

• (a) The level of liquid in the core was measured. It was normal.

• (b) Fifty-five pounds of pressure were placed on the water inside the
  cooling coil system, which was then valved off, and evidence of a
  leak was sought by observation of the pressure. No change in pressure
  occurred over several hours.

• (c) The reactor was brought up to critical, at a very low power level.
  Criticality was achieved at normal control rod position, indicating
  that no detectable ( by this method ) amount of fuel had been removed
  from the core.

It was concluded, therefore, that if any fuel leak were present it was an
exceedingly small one.

Two days later a sample of water drawn from the cooling coils indicated
considerably higher beta, gamma activity, which was positively identified with
fission products, and alpha activity was found which showed clearly that traces
of uranium were also present. Thus it become clear that a small, but definite
leak in the cooling coils inside the reactor had developed. The cooling coils
were drained of water which was collected and held for further analysis. Pressure
and vacuum alternately were later applied to the cooling coils ( after removal of
the fuel from the core) and, by observation of pressure changes of air in the

sealed off cooling coils, attempts were made to measure the extent of the leak.
It was most certainly very small. In the pressure testy fifty pounds was held
for twenty-four hours with only a small drop. In the vacuum test, 25 cm Hg
rose only a few millimeters in four hours due to air in leakage.

At the same time ( June 6) these phenomena which indicated a small leak in
the cooling coils wore observed, it was noted also that the liquid level of fuel
in the core had dropped slightly; about three or four mm below normal. ( As noted
above, on June 4 the level was normal). ( Four mm drop indicated a change in
volume of 250 cc.). When this was observed, an electrical resistance measurement
on an insulated probe in the button of the reactor envelope, just outside the
core cylinder was made and the presence of moisture outside the fuel cylinder
(but inside the safety envelope) was indicated. This clearly indicated that
fuel solution to the extent of some 250 cc had leaked out of the core, presumably
into the Safety Envelope where the presence of liquid was found for the first time.

Thus, to summarize,

• 1. On May 10 it was established that a leak had developed in that portion
  of the reactor core system contained inside the Safety Envelope. The
  leak seemed to involve gas only, and no evidence could be found that
  the rector fuel was involved. Over the next few weeks the radioactive
  gas inventory of core and Safety Envelope was purged slowly into the
  atmosphere, with appropriate dilution.

• 2. On June 6, after purge of the radioactive gas inventory had been com-
  pleted, evidence was obtained that a small leak had developed in the
  water-filled cooling coils of the reactor.

• 3. Also on June 6 evidence was obtained that the reactor fuel solution had
  leaked into the Safety Envelope enclosing the core.

It was decided at once that the reactor fuel should be removed from the core
and stored in the Emergency Storage Cylinder, of "always safe" geometry which
had been included in the reactor assembly for just such contingency as had now
developed:

Dr. Roth, of the A.E.C. at Oak Ridge, who had been apprised of these develop-
ments concurred in the wisdon of this move:

The fuel was moved from the reactor core to the Emergency Storage Cylinder
on June 7.

It is the intention now to complete the disassembly of the reactor components,
already begun, and to determine if possible the cause of the leaks ( apparently
there are at least two) which have developed in the core assembly.

Several observations may be worth recording at this point.

• 1. The cause of leaks in the reactor core is not clear, and can be
  determined only by further examination of the core. Conceivably a
  weld near the bottom of the core may have failed or corrosion through
  the metal container may have occurred. If the latter, the previously
  observed tendency of the fuel to precipitate should not have been as
  likely to have occurred.

• 2. The sequential development of the leak is surprising and somewhat con-
  fusing. At first it was only a gas leak, with no evidence to be found
  of liquid leak. Thirty days later there was slight evidence of a leak
  in the cooling coils. Confirmation of this two days later was accom-
  panied by clear evidence, sought in vain two days before, that some
  200 cc of fuel has leaked from the core into the Safety Envelope.

• 3. The Safety Envelope was incorporated into the design of the reactor
  "to contain radioactive gases and the reactor fuel solution in case
  a leak should occur." It worked precisely as intended.

• 4. The wisdom was demonstrated of providing in the reactor assembly an
  emergency storage vessel into which the highly radioactive fuel
  solution could be moved, in case of a leak in the core, by remote
  control manipulation of pneumatic pressure lines.

• 5. A small but detectable amount of fuel solution escaped through a leak
  in a cooling coil into the reactor cooling waters; Further measurenent,
  is being made to ascertain the total amount involved but apparently, as
  judged by the circumstances involved and the activities thus far measured,
  it is quite a small amount which escaped into the cooling water by a
  small leak in a cooling tube.