Criticality Accident Alarm Systems

[Rob Tayloe <tayloe@ibm.estd.battelle.org>]

On Fri, 17 Jan 1997 12:22:06 -0500 (EST),
  Chul Lee <chullee@umich.edu> asks about
  Criticality monitors at power reactors.

Alas, I can not give you specific advice on power
reactor alarm systems.  But I will add the following
snipped from 10CFR70.24 with some other comments:

   § 70.24  Criticality accident requirements. 

    Each licensee authorized to possess special nuclear material
    <...snip...> shall maintain in each area in which such licensed 
    special nuclear material is handled, used, or stored
    a monitoring system <...snip...> using gamma- or neutron-sensitive 
    radiation detectors which will energize clearly audible alarm 
    signals if accidental criticality occurs. <snip>

       (1) The monitoring system shall be capable of detecting a 
    criticality that produces an absorbed dose in soft tissue of 
    20 rads of combined neutron and gamma radiation at an unshielded 
    distance of 2 meters from the reacting material within one minute. 
    Coverage of all areas shall be provided by two detectors.

It is interesting to note that the above is loosely drawn from
ANSI/ANS-8.3-1986 "American National Standard Criticality Accident
Alarm System".  In general, the guidance in the ANS-8 series of
standards is directed to prevention of criticality accidents
for operations involving fissionable materials outside nuclear
reactors.  

In the Scope of ANS-8.3 it is stated that "this standard does not
require separate additional instrumentation when the operating 
instrumentation of facilities, such as nuclear reactors or critical
experiments, meets the requirements of this standard".

Another interesting, and significant, difference between ANS-8.3
and 10CFR70.24 is in the detection criterion.  In ANS-8.3
section 5.6 "Detection Criterion" the criticality alarm system is 
required to detect an accident delivering the equivalent
of an absorbed dose in free air of 20 rads per minute at a distance
of 2 meters from the reacting material.

For neutron detection systems, there is quite a difference in power
for a critical system delivering 20 rads per minute in free air versus
20 rads per minute in soft tissue.  My understanding is that the
origin of the ANS-8.3 standard was directed to gamma detection
systems.  In such cases 1 R is about 1 rad for tissue.

It is a non-trivial exercise to demonstrate that one has satisfied
the requirements for coverage of an area by a neutron detection
system.

Rob Tayloe <tayloe@battelle.org>
Columbus, OH