The application of nuclear instrumentation demands an understanding of the behavior of a reactor. Because power density varies with position in the reactor, an average power measurement is needed. Out-of-core detectors are con­sidered to be spatially averaging and are discussed here from this viewpoint. Detectors for measuring spatial variations in nuclear fluxes are discussed in Chap. 3. Out-of-core detectors are reasonably good averaging devices for most reactors if their installation is properly designed, especially in regard to shadowing by movable objects in or around the reactor.

To a limited extent, nuclear instrumentation influences reactor design. For example, it may be necessary to adjust the location of control elements to avoid shadowing effects on radiation sensors. It may be desirable to introduce a window to cause streaming that will ensure an adequate level of radiation for reliable instrumentation response. Although it is always desirable to avoid reactor-vessel penetrations, penetrations are sometimes necessary to ensure an adequate signal. The minimum reactor power level must be determined to ensure measurements at all reactor levels. If the minimum is too low or uncertain, a neutron source must be provided to maintain the minimum level at a measurable value. There must be provisions for renewing or replacing the source.

All these requirements stem from the mandate that the state of the reactor must always be known. In other words, the reactor level and the rate at which the reactor level is changing must always be known and must always be under control. To ensure this knowledge, redundancy is always used to some degree (see Chap. 12). A common mode of redundancy is to make measurements with three separate detectors or channels, each with independent circuitry. The shutdown signal from any one channel must be in coincidence with another signal (i. e., a two-out-of-three coincidence) before shutdown is allowed.

Radiation detectors sample radiation intensity. Initially, the relationship between reactor power level and the sampled radiation intensity is based on design calculations alone. At power levels near full-power operation, the detectors must be calibrated. This is best accomplished by making heat-balance measurements. Subsequent cal­culations, using the calibration, then relate the detector response to the reactor power level. Periodic recalibration is required to take into account changing radiation patterns and spectra, fuel burnup, and changes in detector sen­sitivity.

The great range in reactor power (from watts to hundreds of megawatts) makes it impossible to use one set of detectors and circuits, despite the wide range of the detectors. Research has produced detector and circuit arrangements capable of measuring over a range of 10 decades. Signal-conditioning circuits are the key to success (see Chap. 5).

A single set of detectors can be used to measure only a part of the reactor range and must be complemented by additional sets of detectors. For safety and reliability, a part of the range of the detectors is sacrificed by having them duplicate the measurements of a part of the range of other detectors. This duplication, or overlapping, is needed for a smooth transfer of control and safety functions from one detector set to another. The amount of overlap is typically one to two decades.

The most common way of dividing the power-level range is to use three ranges: source, intermediate, and power ranges.6 This nomenclature is used in commercial

practice. Figure 2.2 shows a typical selection of neutron detectors to cover these ranges.

Each range has peculiarities that depend on the radia­tion levels corresponding to that range and on whether or

not a reactor has been operated. Special features, some­times temporary, must be incorporated into the instru­mentation design to ensure reliable performance during the initial period of a reactor when it is “clean and cold.” The same instrumentation must operate when the reactor has accumulated its full burden of radioactivity and at every condition in between.

In the source and intermediate ranges, the reactivity of the reactor is limited by controlling or limiting the rate (period) at which the power can be increased. In the power range, instrumentation must prevent the reactor from exceeding its rated or licensed operating limit.

In the fully shutdown condition, the neutron density to which the detectors are exposed is frequently quite low, in fact, so low that individual neutrons are counted in order to gain information about the reactor status. Counting is also the only way to detect neutrons in the relatively high gamma fields that may be present.

The limits of the source range (or counting range) are determined by permissible counting rates, expressed in counts per second. The low end of the source range is determined by the counting rate needed to achieve a safe condition, as specified in the safety review (see Chap. 12). This minimum counting rate is normally from 1 to 10 counts/sec. The counting rate is established during the preliminary design and is fixed by consideration of the statistical nature of the neutron population and the time interval needed to achieve a measurement of prescribed accuracy. The counter must be located where the flux density is sufficiently large to ensure that this counting rate is achieved. The magnitude of the neutron source is selected to attain (at the detector location) a neutron flux that results in at least the minimum counting rate at all times. The maximum or high end of the source range is determined by the ability of the counter and the associated electronic circuits to resolve the individual counts. If the counting rate is too high, the resolution loss produces a serious error in the signal. Typical maximum counting rates are 0.5 to 1 X 106 counts/sec, and the allowable resolution loss is less than 10%.

The source range presents an adverse situation for the detection of neutrons. The detector used must be carefully selected for its sensitivity to neutrons in the presence of a large gamma background. The condition of few neutrons and many gammas exists immediately following a scram from full power.

The intermediate range overlaps the source range, and its gamma background is not as severe. However, since the neutron flux is high, individual neutrons are no longer resolved, and the signal takes on a direct-current aspect, becoming indistinguishable from the gamma background (also a d-c signal). Here again, it is essential to know the gamma level at the low end of the intermediate range. Through sensor design the gamma contribution at the low end is normally kept below 10% of the neutron signal. Again, the worst condition exists during a start-up immedi­ately following full-power scram. The intermediate range usually extends into and completely overlaps the power range.

The power range covers from 1 to 150% of full power to provide some allowance for small power excursions. In the power range there is normally no great difficulty with interfering radiation. Neutron detectors that are not gamma compensated are satisfactory, but gamma-compensated sensors may be used for uniformity. These are similar to those used in the intermediate range.