The purpose of this post is to address an error in a recently published review of current release estimates from the Fukushima Dai-ichi nuclear power plant disaster that began in March 2011. The post is part of an ongoing effort to communicate results of scientific studies aimed at understanding the impact of Fukushima on the health of the North Pacific and residents of the west coast of North America. In a recent review paper published in Progress in Nuclear Energy by Koo and colleagues this July, compiled estimates of atmospheric and ocean releases from Fukushima were presented. Due to an error in interpretation they suggest that direct ocean releases were a factor of 4 greater than atmospheric releases of radiologically significant isotopes like 131-I (~8 day half life) and 137-Cs (~30 year half life). This error inflates release estimates and has been reported on to suggest Fukushima releases exceed Chernobyl’s. Accurate estimates of releases from Fukushima suggest that they are about an order of magnitude less than those from Chernobyl in 1986.
The study of Koo and others (link to a ResearchGate upload) estimated atmospheric releases of 131-I, 137-Cs and the noble gas 133-Xe (half life ~ 5 days) from the Fukushima Dai-ichii nuclear power plant. Their estimates compared with previously published estimates are reported in the following table (Table 2 from paper):
Similar to previous work, for example, they estimate the atmospheric release of 137-Cs from the plant to be 10-50 PBq or somewhere between 3 and 17 kg of the isotope. Given the core inventories of reactors 1-3 this release represents about 4% of the inventory at the time of the meltdowns in March 2011.
The authors make a significant error when they begin their estimate of direct releases from Fukushima to the ocean when they state the following in section 2.2.2. Release from the primary system into the sea:
It is reported that, of the total radioactivity released from the units 1–3 into the environment, more than 80% of it flowed into the sea (Hoeve and Jacobson, 2012 and Christoudias and Lelieveld, 2013), implying that 4 times more radioactivity was released to the sea than to the atmosphere.
Bolds are mine. In stating that 80% of the total radioisotope releases flowed into the sea they fundamentally misinterpret the studies they cite. What the study of Christoudias and Lelieveld (2013), and other studies not referenced here in the diary, actually show and establish is that (quoting from the Christoudias and Lelieveld work):
We calculated that about 80% of the radioactivity from Fukushima which was released to the atmosphere deposited into the Pacific Ocean.
This is a fundamentally different than the interpretation Koo and colleagues use in their study. By wrongly interpreting that atmospheric releases represent 20% of the total release they assume that direct ocean releases are 4 fold greater than the 4% of core inventories (10-50 PBq) or 16% of core inventories of 137-Cs in March 2011. This error greatly increases the estimated total releases from the plant (atmosphere + direct ocean).
It is very likely that this incorrect approach will lead others to conclude that total releases from Fukushima are greater than those from Chernobyl. For example a back of the envelope calculation assuming the 4% of the total core inventory of 137-Cs (760-820 PBq according to the table above) was released to the atmosphere and 16% to the ocean would lead to a total release of ~152-164 PBq. Such a calculation was done by a popular news aggregator and editorial site that has a history of misinterpreting and misinformation the public about Fukushima. This estimate, not surprisingly, is at great odds with existing estimates based on measurements and modeling.
Best estimates to date suggest that:
1. atmospheric releases of 137-Cs were 19.4 +- 3.0 PBq through the end of March 2011
2. direct ocean discharge of 137-Cs to the Pacific in addition to atmospheric deposition are in the range 2.3 to 26.9 PBq
3. About 19.5 +- 5% of releases were deposited to land while about 80% ended up in the Pacific Ocean
A report reviewing the most recent peer reviewed studies which reaches these conclusions was summarized in a post here.
Releases of isotopes that represent potential radiological health threats given their respective total activities and/or their significant half lives (e.g. 131-I and 137-Cs) were about an order of magnitude (factor of 10 times) lower than the releases from the Chernobyl disaster in 1986 (see reports here and here for example). More and more observations are being made globally by the international scientific community which will help to improve source term and release estimates. I will report on these studies as the data becomes available.
I have contacted the authors to bring their attention to this problem with their study.