A NEW APPROACH ON ENGINEERING HOT CHANNEL AND HOT SPOT STATISTICAL ANALYSIS
Technical Report
·
OSTI ID:4650118
Several attempts have been done by various authors to introduce the methods of statistical analysis in the calculation of correction factors due to fabrication tolerances (engineering hot channel factors); a generalized statistical approach is described. A nuclear core is physically defined by a certain number of independent variables (x/sub i/), which define core geometry and material characteristics: for example, fuel density and enrichment, fuel radius, cladding internal and external radius. These x/sub i/'s variables are subjected to statistical variations from point to point in the core. On designing a reactor, one is faced with the problem of determining certain thermal and hydraulic dependent variables Y = f (x/sub i/,... x/sub i/,...). In the present work, account is taken of the fact that the thermal and hydraulic variables Y's depend: a) on local values of the x/sub i/'s; b) on neutron flux spatial distribution and on the entire reactor thermal and hydraulic conditions. To do this a geometrical core model was developed. This model, called spot model,'' considers each fuel channel as divided in small regions: in each of these small regions the x/sub i/'s may be assumed constant and their actual values are chosen at random (independently one from the other, and in an independent way from region to region) according to their experimental distribution curves due to fabrication tolerances. Having represented in this way an actual'' core, it is possible, with the aid of a detailed thermal and hydraulic calculation routine, to evaluate the Y variables relative to such a core. This procedure is repeated according to a Monte Carlo technique, choosing at random as many random cores'' as necessary. In this way actual probability and distribution curves for the Y are obtained. By intersection of the distribution curves with a certain confidence level the expected Y/sup +/ values are obtained. This procedure was programmed on an IBM-7090 computer. The code, named HYTHEST, evaluates the correction factors relative to a PWR or a BWR reactor core. The first results obtained can be summarized: a) the most probable values of the Y's do not always correspond to the nominal values: the deviations are stronger for highly non-linearly dependent variahles (e.g. maximum fuel temperature); b) the distribution curves of the Y's depend from the assumed neutron flux shapes: so, for instance, the exit enthalpy variations are larger, if the axial flux shape is more peaked; c) the correction factors obtained with the spot model'' statistical approach are generally smaller by an order of magnitude than that given by the corresponding conservative method: however, this rule does not hold for some non-linearly dependent variables. A detailed comparison between the conservative method and the spot'' model was made. (auth)
- Research Organization:
- FIAT. Sezione Energia Nucleare, Turin
- NSA Number:
- NSA-17-035094
- OSTI ID:
- 4650118
- Report Number(s):
- EURAEC-702; FN-E-35
- Country of Publication:
- United States
- Language:
- English
Similar Records
STATISTICAL METHODS FOR HOT CHANNEL AND HOT SPOT CALCULATION. Quarterly Progress Report No. 3, July 1, 1961 to September 30, 1961
STATISTICAL METHODS FOR HOT CHANNEL AND HOT SPOT CALCULATION. Quarterly Progress Report, January 1, 1963 to March 31, 1963
Approximate tolerance intervals, based on maximum-likelihood estimates
Technical Report
·
Mon Oct 30 23:00:00 EST 1961
·
OSTI ID:4818813
STATISTICAL METHODS FOR HOT CHANNEL AND HOT SPOT CALCULATION. Quarterly Progress Report, January 1, 1963 to March 31, 1963
Technical Report
·
Tue Apr 30 00:00:00 EDT 1963
·
OSTI ID:4656734
Approximate tolerance intervals, based on maximum-likelihood estimates
Conference
·
Thu Dec 31 23:00:00 EST 1981
·
OSTI ID:6857967