The theory of checking statistical hypotheses is used to elaborate the probabilistic approach to establishing
the reference levels of emissions and discharges from nuclear power plants. Expressions are obtained and
a scheme is proposed for establishing these levels. The scheme is based on optimizing the indicator and
control functions and the monitoring sensitivity. For more effective monitoring, it is proposed that when the
reference levels are established the operating regimes of the power-generating units of a nuclear power
plant be divided into operation at power and a regime of planned preventative maintenance. Using as an
example the first unit of the Khmel’nitsk nuclear power plant, the characteristics of the statistical
distribution of the various components of real gas-aerosol emissions are determined and the corresponding
reference levels are calculated.
The establishment of standard levels of emissions and discharges of radionuclides is the most important measure for
ensuring the radiological safety of a nuclear power plant. Being distinctive threshold values, they perform an important indi-
cator function in the system that monitors the state of a very complex object – a nuclear power plant. In addition, they are
important control parameters of the radiation state in the system nuclear power plant + environment, playing a key role in
ensuring that the radiation effect of emissions and discharges from a nuclear power plant on the environment and the public
does not exceed the regulation level.
The effectiveness of implementing the indicator and control functions indicated above depends directly on the
optimality of their values. The best approach must take into account the real properties of the parameters monitored and
the characteristics of the measurement means used (sensitivity, accuracy, and masurement limits) and the checking proce-
dure (check points, periodicity, measurement time). The factors indicated can be taken into account in such a comprehen-
sive manner on the basis of an analysis of the data from real measurements, which are a convolution of the intrumental
function of the measurement system used and the function characterizing the physical quantity being monitered. Since the
processes resulting in the formation of emissions and discharges for a nuclear power plant are stochastic, the function char-
acterizing the physical quantity being monitored should be interpreted as a density distribution function. Indeed, because
they are affected by many factors which vary conitnually in time the emissions and discharges from a nuclear power plant
are stochastic quantitites, which makes the probabilistic approach better substantiated and more suited for establishing the
reference levels.
The need to use the results of a stochastic analysis of real emissions in order to establish the reference levels has
been pointed out in [1, 2]. It has been shown for radioactive inert gases that the distribution of the normalized emissions (pre-
sented per unit of energy generated) in the normal operating regime of a nuclear power plant is log-normal, and it has been
suggested that the maximum real emissions be used as the reference levels.
Atomic Energy, Vol. 100, No. 2, 2006
UDC 621.039.58:519.25
Insitute of Nuclear Research, National Academy of Sciences of Ukraine. Translated from Atomnaya Énergiya, Vol.
100, No. 2, pp. 135–146, February, 2006. Original article submitted August 2, 2005.
1063-4258/06/10002-0132
©
2006 Springer Science+Business Media, Inc.
132
A. N. Berlizov, M. A. Grigorenko,
and V. V. Trishin
PROBABILISTIC APPROACH TO ESTABLISHING
THE STANDARD LEVELS OF EMISSIONS AND
DISCHARGES FROM A NUCLEAR POWER PLANT