Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Radon counting statistics - a Monte Carlo investigation

Journal Article · · Health Physics
OSTI ID:394062
 [1]
  1. Radiation Protection Service, Weston, Ontario (Canada)
Radioactive decay is a Poisson process, and so the Coefficient of Variation (COV) of {open_quotes}n{close_quotes} counts of a single nuclide is usually estimated as 1/{radical}n. This is only true if the count duration is much shorter than the half-life of the nuclide. At longer count durations, the COV is smaller than the Poisson estimate. Most radon measurement methods count the alpha decays of {sup 222}Rn, plus the progeny {sup 218}Po and {sup 214}Po, and estimate the {sup 222}Rn activity from the sum of the counts. At long count durations, the chain decay of these nuclides means that every {sup 222}Rn decay must be followed by two other alpha decays. The total number of decays is {open_quotes}3N{close_quotes}, where N is the number of radon decays, and the true COV of the radon concentration estimate is 1/{radical}(N), {radical}3 larger than the Poisson total count estimate of 1/{radical}3N. Most count periods are comparable to the half lives of the progeny, so the relationship between COV and count time is complex. A Monte-Carlo estimate of the ratio of true COV to Poisson estimate was carried out for a range of count periods from 1 min to 16 h and three common radon measurement methods: liquid scintillation, scintillation cell, and electrostatic precipitation of progeny. The Poisson approximation underestimates COV by less than 20% for count durations of less than 60 min.
OSTI ID:
394062
Report Number(s):
CONF-9607135--
Journal Information:
Health Physics, Journal Name: Health Physics Journal Issue: Suppl.6 Vol. 70; ISSN HLTPAO; ISSN 0017-9078
Country of Publication:
United States
Language:
English