In the paper an error occurred in the calculation of the cumulative reactor years from the data of IAEA report RDS2/31 (2011 edition). It affects Figure 1 and Table 2 of the original paper. Minor implications in the following sections: Materials and Methods, Results, and Discussion, have been corrected below. The main results of the paper are derived with the correct value of (2010) = 14097 cumulative reactor years for the Fukushima accident. In the Acknowledgment section there, is also a change.
Materials and Methods. The curve of cumulative reactor years is corrected in the new Figure 1. Likewise, values for cumulative reactor years and the corresponding hazards of the TMI accident and of the Chornobyl accident in original Table 2 have to be replaced with those of the new Table 2.
Results. The first two sentences refer to the original Table 2 and are discarded. The last three sentences are replaced by “The empirical estimates for the reactor-specific hazard decrease from 1.2 × 10−5 RY−1 (1979) and 3.3 × 10−6 RY−1 (1986) to 4.8 × 10−7 RY−1 (2011). The value for 2011 is considered the most reliable because it is derived with the largest numbers of accidents and cumulative reactor years. It falls below the point estimate of RY−1 from the PSA study and stays within its 95% CI. Since the PSA considered accidents of INES level ≥4 and pertains to level ≥5, the results are compatible.”
Discussion. In the fourth sentence, the phrase “a factor of three” is replaced by “a factor of 25.”
Acknowledgment
The author would like to thank Dr. Yang Duanjie, Beijing, China, for carefully reading the original paper and for reporting the error.