SNT and the Fukushima Personnel, 2nd Try
This piece assumes some familiarity with the The Green Table and SNT. It is a replacement for the first version which made the elementary blunder of assuming the inhaled dose is incurred when inhaled. If you downloaded the first version, please trash it.
Figure 1. Fukushima flooded area. Reactor sites had been excavated down to about 10 m above mean sea level. That area was flooded, 2-3 m deep.
The Green Table is based on the premise that if you want to test a radiation harm model you must look at situations where people have received very large doses and compare the model's predictions with reality. Radiation is such a weak carcinogen that the difference in predictions in low dose cases is for practical purposes untestable. The Green Table lists some 20 such high dose groups. The table makes the point that the cumulative dose is far less important than how rapidly that dose was incurred. Given our remarkable ability to repair radiation damage, this should have been expected.1 But there is one glaring omission in the Green Table: the Fukushima plant workers. Unfortunately, like the Chernobyl liquidators, what can be learned from this group is much less than what might have been learned.
Fukushima Daichi had a system in which controlled areas of the plant within which significant exposure might occur were fenced off.2 If you needed to go into these areas, you were issued an APD (Alarm Pocket Dosimeter) and required to wear protective clothing. After finishing the job, the APD was returned to its docking station, and the dose the individual received automatically uploaded to a computer. As long as the APD's were read at least daily, this system would have given us the individual daily dose rate profiles SNT needs to estimate the resulting cancer incidence.
Figure 2. This photo taken well after the tsunami shows the concreted embankment behind the reactors.
Unfortunately, almost all the APD's were stored on the ground floor of the Service Buildings which were flooded by the tsunami.3 Of the 5000 APD's on site, only 320 were working after the tsunami. The data collection and analysis computer went down with the grid and was unavailable. A system for manually recording the doses was apparently not established until April, well after the worst of the exposures was over. About that time, everybody was finally issued an APD, but they were a variety of brands and types, so the doses still had to be recorded manually, which was subject to a range of errors and omissions.
The other poorly anticipated issue was inhaled dose. While personnel on jobs outside the control rooms donned protective gear including full face masks, this was not the case in the control rooms. The team had to communicate, had to eat and drink, so they counted on the building's filtration system, which were largely inoperable.4 As a result the personnel with the largest individual doses were mostly internal from breathing in contaminated air.5 This must have occurred over about a five day period very early on. Once the releases stopped, airborne contaminants would have cleared out or fallen out in a matter or hours.
A final problem is misplaced and possibly feigned concern for worker privacy. Radiation high priests have consistently invoked worker privacy to avoid releasing individual dose rate profiles. This goes back to the atom bomb survivors, and has continued through INWORKS. NIH has refused to release the data they have on the Chernobyl thyroid cases. This has not stopped these groups from using such data to make claims such as their data supports LNT. Refusing to release the raw data in such situations violates a basic rule of scientific integrity.
Fukushima is no exception. UNSCEAR has refused to release the worker dose rate information they have on the basis of privacy. This is completely unnecessary. We don't need to know whose profile it is. It's easy to disidentify the data. Privacy is not the real motive.6
Despite all this, it is possible to come up with at least tentative dose-rate profiles for the most exposed personnel. Thanks to UNSCEAR interviews, we have estimates of the cumulative internal and external doses for the 41 highest dose people.\cite{unscear-2013}[Tables D3, D6, D7]. We know they received essentially all of that dose in the first 20 days of the release up to April 1. We know nobody got more than 50 mSv per month after that.\cite{unscear-2013}[Table D3] That's below the 2 mSv/day tolerance rate, if it were received evenly. While the follow on month doses would not have been incurred evenly, it's clear that the great bulk of any harm will result from the first month.7
With respect to the internal doses, UNSCEAR says
D40. Work history information for 10 out of 12 of the workers with the higher reported internal exposures (i.e. committed effective doses greater than 100 mSv) indicated that there was reasonably good evidence that intakes commenced on 12 March 2011. However, the available evidence for this group of workers did not allow a clear conclusion to be drawn about whether intakes effectively ceased at some time during 12 March, or whether they continued for a few days. For all of the workers in this group for whom intakes probably commenced on 12 March, it was likely that significant intakes had ceased by 15 March.\cite{unscear-2013}[p 231]
The release took place in a series of spikes over about 5 days, Figure 3, with a lull on the 13th and 14th. The UNSCEAR guess of 3 days is not unreasonable.
Figure 3. Fukushima plant boundary dose rate measurements. Spikes drop off by about a factor of 10 within six hours. All permanent monitors went down with the grid. These measurements were taken by cars on the western perimeter of the site. The winds were light westerly on the 12th and 13th before becoming onshore. This figure grossly underestimates the Unit 1 explosion spike on the afternoon of the 12th.
To concoct an internal dose rate profile, we must
1) Estimate the time history of the intake of radioactive material
2) For that intake history, estimate the dose rate incurred as the material decays and is eliminated from the body in the natural course of events.
In theory, we need to do that for each inhaled isotope. But in the Fukushima case, we can be sure that the inhaled dose was dominated by Iodine-131. Table 1 shows the release itself was mostly Iodine. Moreover, given the rapid decay, the iodine dose will be incurred over a shorter period (higher dose rates) than the cesium. So as a first cut, I'm going to assume all the internal dose was I-131. The big error here is the Te-132/I-132 which was important in the first few days.
We think the worst dose rates were incurred in the control rooms. If we combine UNSCEAR's estimate that the internal dose was inhaled over a three day period, with the fact that they tried to rotate out the control room personnel on an 8 hour shift basis, we can guess that the I-131 dose was inhaled in three 8 hour periods. Combining that with the effective half-live of 7.3 days, we end up with a dose-rate profile that is shaped like Figure 4.
Figure 4. Concocted Internal I-131 Profile for 200 mSv Control Room Crew. Dose rate (blue) read left. Daily doses (red) read right.
Given this simple construction, the cumulative dose is linear in the assumed control room dose rate. We can easily adjust the control room dose rate to match any target cumulative dose. Combining this with the assumption that the external dose was received evenly over the first 20 days, gets us to a total dose profile for which we can compute the SNT cancer incidence. Of course, this procedure involves a string of assumptions, none of which is correct; but any job worth doing, is worth doing poorly, so let's see where this takes us.
If we use this dose profile shape for the internal dose and assume --- optimistically --- that the external dose was received evenly over the first 20 days to April 1, and ignore doses received after that, we have concocted dose rate profiles for these people. I used the same reasoning to gin up dose rate profiles for the 399 people who received more than 50 mSv in the first month.
Once you have the dose rate profiles, the SNT process is completely mechanical. Figure 5 shows the resulting Lost Life Expectancies for these people.
Figure 5. Top 399 Fukushima Personnel SNT Lost Life Expectancies.
This figure is strongly top end weighted. The person with the worst dose rate profile has a Lost Life Expectancy (LLE) of 47 days. This falls off very rapidly. The 10th worst profile has an LLE of less than 4 days. The bottom sub-group in the top 400 has an LLE of 0.22 days. Their bar is barely visible on the figure. Notice also that the internal dose dominates the external for at least the top 10.
The total LLE for this group is 223 days. If they were compensated according to the US Dialysis Standard ($350/day), their compensation would be $75,000, with $16,500 going to the worst profile person. This number is very different from the UCert public compensation which is based on the maximal dose rate profile. The maximal dose rate profile for someone standing unmasked, 24/7, outside the Unit 1/2 Service Building throughout would result in UCert compensation around a million dollars.
Figure 5 is almost certainly optimistic mainly because the dose rate profiles were spikier than my simple model. But even so, from the point of view of the Green Table, the number of high dose rate and high dose individuals is almost certainly too low to be a strong test of harm models. However, the Fukushima workers are being tracked. As far as I can tell, so far the only negative finding has been a non-significant decrease in the fraction of high dose individuals whose thyroid was judged to be normal relative to a control group.\cite{sobue-2014} This was matched by a non-significant increase in the fraction of the control group for whom further examination was recommended relative to the high dose group.
A key lesson here is the importance of the inhaled dose during plume passage. During plume passage, the unmasked, outdoors, inhaled dose will be 100 to 50 times the external dose as measured by a dosimeter, depending on the plume composition, Table 2. This applies to the public as well as plant personnel and responders. The decay process will tend to spread the inhaled dose out; but without protection it will usually dominate during plume passage.
Table 2. ICRP Plume dose coefficients}
In most releases, the radioactive material will be released in one or more spikes, as it was at Fukushima. Near the plant, the plume passage time for a spike will usually be no more than a couple of hours. A closed building with an air exchange rate of around 0.25 per hour will offer considerable protection for a few hours. But the real protection is a properly fitted N95 mask, which can cut the dose rate by a factor of 100. But masking up is rarely if ever mentioned in release response instructions. I find that perplexing.
But by far the most important takeaway is the lack of detectable harm. Here we have thousands of people who worked within a few hundred meters of multiple,:x fully breached, melted down reactors, in a situation in which just about everything went wrong; hundreds of whom suffered dose rate profiles far in excess of what would be incurred by the public in a release, and we can't reliably detect any harm. That should tell us all we need to know about the horrors of a nuclear power plant release.
LNT takes the opposite view and claims the only thing that counts is cumulative dose. As a result, it does a horrible job of predicting cancer incidence when a dose is received more or less evenly over long periods. Under any coherent logic, LNT must be rejected.
I've pieced this description together from TEPCO and UNSCEAR sources.\cite{tepco-2012, unscear-2013} Some of the details are probably wrong.
Each of the three Service Buildings served two reactors. These buildings were sandwiched between their pair of reactors. The plant site was excavated down to about 10 m above mean sea level by digging into the seaside bluff, Figures 1 and 2. The tsunami peak was 12 to 13 m. The three control rooms were on the second floor of the Service Buildings and escaped flooding.
When the earthquake occurred, at least 2400 people were working in the excavated area, mainly on the Unit 4 shroud replacement.\cite{tepco-2012}[p 163] Thanks to a remarkably orderly evacuation to higher ground behind the reactors in the 40 minutes after the earthquake, only two people were killed by the tsunami.
Some doors were damaged by the hydrogen explosions. Others were bypassed to get emergency cables to the control room.
Internal dose rates must be measured by a Whole Body Counter. The plant's own WBC was unusable, in part due to the high on-site levels. A mobile WBC was installed at a staging area 20 km south of the plant.
If it's a legal issue, then nuclear workers should be required to sign releases as a condition of employment
It seems that, whenever a team was sent to do a high dose task, at least one member was given an APD set to alarm at 20 mSv accumulated dose. The team was instructed to abandon the job if the device alarmed. The alarm was ignored at least once on March 25th. However, it is likely few if anybody received more than 20 mSv in a day after April 1.
Re: "A key lesson here is the importance of the inhaled dose during plume passage. During plume passage, the unmasked, outdoors, inhaled dose will be 100 to 50 times the external dose as measured by a dosimeter"
I infer this is because the dosimeter measures counts from decaying radiative atoms for short times as they pass by, but human lungs trap the radioactive atoms for their longer effective half-life.
In your Table 2 I'd think the beta (electron) component of the external exposure should be nil because the electrons don't penetrate epidermis, while lung tissue has no such barrier.
I've been hosting a debate on LNT in the FaceBook forum Renewable vs Nuclear Debate. One of the anti-nukers commented that inhaled radiation was very different than external, sparsely ionizing radiation, and would probably follow LNT due to the particles that get stuck in the lung. I haven't seen any counter argument, but perhaps this data on Fukushima workers would do it.
Do we have any information on the nature of these particles? Are they just isolated iodine atoms floating in the air, or more like an aerosol with thousands of atoms in a small cluster? The concentration of radiation from one particle in a small volume of tissue could overwhelm any hormetic response.