It was interesting that Sarah Phillips choose to accuse me of “Cherry Picking”, for those of you who do not know what “Cherry Picking” is it is when a person uses a very small amount of data either while deliberately ignoring a large amount of data or while being too lazy to obtain more data.
The problem with a cherry picker is if their argument is reasonable and based on something which truly exists then their use of a far too small amount of data weakens their own argument. For example if I count the blue tits in my garden over one minutes then I might see if I am lucky in summer one blue tit.
If we assume that the area of garden outside my window where I am gazing has two states (free of blue tits or contains a blue tit) then we can use the same ideas as are used for counting radioactive events. The estimated standard deviation (no that is not some sort of sexual matter) of the number of blue tits counted is the square root of the number counted.
So if I look out of the window and see during my observation of the garden that one blue tit is present during the one minute observation time then the estimated standard deviation (ESD) based on the number of observations of the blue tits is 1. This means we have an ESD which is the same size as the average number of blue tits seen per minute.
If during my one minute observation of my garden I see one blue tit and if during a similar observation of the next door’s garden I see no blue tits then it is unlikely that a real difference exists between our gardens.
On the other hand if I observe my garden for one whole hour and see 60 sightings of blue tits while in the garden next door I only see 6 sightings. Then the ESD on my garden is 7.75 blue tits while the ESD on the observation of next door’s garden is 2.44.
The difference between the numbers of blue tit sightings in the two gardens is 56 which is 5.5 times larger than the sum of the ESDs of the two measurements. Thus the likelihood that the gardens are different is far more than 99%.
While bluetits are important to me, their appearance in the garden is a source of joy to me the significance of their numbers is less important than the incidence of adverse health effects in the human population. Now unless you have been hiding in a hole in the ground and making a point of never reading a newspaper, blog or even “yahoo news” you should be aware of the fact that a great debate exists regarding the question of “should we use nuclear power to make our electricity ?”.
After the Fukushima debacle the question of how likely is a nuclear power accident to harm a member of the general public has become important to many members of the public. The problem is that no matter what energy system we choose to make our electricity a finite risk exists that someone will be harmed.
For your information a series of adverse events associated with energy systems are listed below, to qualify as a death it must occur within one year of the event. A worker is a person working at the energy production site during the accident.
|Energy system||Location||Worker deaths||Non worker deaths|
|Coal||London (Great Smog, 1952)||0||12000|
|Hydropower||Vajont Dam (1963)||0||Circa 2000|
|Coal||Courrières mine (France)||1099||0|
|Natural gas||Ufa train disaster (USSR)||0||575|
|Natural gas||New London School (Texas)||0||Circa 300 children|
|Hydropower||Dale Dike Reservoir (1864)||0||244|
|Natural gas||Piper Alpha (1988)||167||0|
|Coal||Aberfan (1966)||0||28 adults 116 children|
|Biomass||Boston Molasses Flood (USA)||0||21|
|Nuclear||SL-1 (Idaho, USA)||3||0|
Now we should be able to see that in terms of short term human deaths that both Fukushima and Chernobyl are small compared with many of the other energy accidents which have occurred in the 20th century. While it was a small scale accident for pure gross out the Boston molasses accident is one which is etched into my mind, it was an event which condemned people to a truly sticky end.
We should also be aware of the fact that in England (London) and Japan the burning of fossil fuels has lead to clear adverse health effects, even if others are rushing to ban one energy system (nuclear) we should not ignore the fact the many energy systems do have the potential during accidents or in normal operation to harm people.
While I strongly hold the view that we should improve health and safety standards in the nuclear sector, it is clear that great room for improvement exists in the non nuclear sector. I was at a recent meeting on nuclear safety during which one of the speakers did consider the Bhopal event, I held (and expressed) the view that the international scale for radiological accidents should be expanded to consider non nuclear events. The Bhopal event is at least as serious as Chernobyl, the Chernobyl accident was not able to wipe out a whole town in the same was as the MIC gas was able to gas to death a vast number of members of the general public.
One way to put it is “why should other sectors of industry be held to a lower standard than the nuclear industry ?” which I think is a better attitude than “why can not the nuclear industry be held to the same lower standard as other sectors of industry are held to”.
While some people might think it is monstrous we need to ask ourselves the questions of.
- Which electricity generation method kills the least people per kilowatt hour ?
- Which energy system does the least amount of environmental damage per kilowatt hour ?
- Where can society make the largest reduction in death and human suffering for a given expenditure ?
Then based on the answers to these questions we should choose which energy systems to use rather on some philosophical or emotion based reason such as “coal is the devil’s fuel as it is found underground”, “biomass is good because it comes from nature” or “nuclear power is evil as some of the technology was invented during weapons research”.
I know that a lot of people choose their preferred energy system for emotional reasons rather than rational reasons, I would like to know how many of them would choose a health care provider based on emotional reasons such as
“that doctor must be good as he comes from my home town”,
“that doctor is no good as he comes from my home town”,
“that doctor is (good / no good) because the doctor is (Tall / Short / Male / Female / Irish / English / Swedish / An arab / Jewish / Buddhist / Muslim / Hindu / Black / White / insert some other type if you are not happy with the selection offered)”
“That doctor is good he always prescribes pills with nice colours”
I am more interested in questions such as
“Is the doctor sober when on duty”
“Does the doctor pay attention to my health issue or not”
“Does this doctor’s care give good outcomes for the patient”
“Does this doctor’s bedside manner rub me up the wrong way”
While a single incidence of finding the doctor drunk in his surgery, smoking weed beside the hospital entrance or snorting drugs off his desk or a single case of a lazy medic would make me go elsewhere the mere fact that one of the doctors patients had an adverse outcome (maybe he was unable to cure a person) or maybe the doctor seemed to annoy me once (properly because he/she failed to fix me as quickly as I wanted to be fixed with zero pain or effort on my part) does not mean a doctor is no good.
Even the best doctor has cases where he/she fails to cure a person, the doctor who only takes on easy to cure cases is deeply unethical. Also the doctor who I do not instantly feel is my best friend can still be a perfectly good doctor.
Rather than using a single observation of the doctor it is better to look at what has happened to a large number of patients. For example in the case of Harold he might have seemed like the perfect family doctor who was willing to visit the frail old lady if you made a single observation of him. But for those of you who do not know of this disgusting felon the local undertakers were aware that a large fraction of his patients were dying in an unusual manner. Harold Shipman the GP from hell was poisoning his patients for his own depraved amusement; an examination of a large dataset was needed to identify this nefarious fiend to enable the evidence which sent him to jail to be collected.
In the same way as you should choose a doctor based on rational reasons, you should choose your energy supply system on rational reasons AND using a data which can be trusted.
Now after seeing the value of using a large pool of data, I have to ask why does the anthropologist use such small samples of people to prove her points in her papers about Fukushima and Chernobyl ?
Regarding Chernobyl Dr Phillips (lets get her name and title right) uses a single case of a man who had brittle bones and died of lung cancer as evidence that radiation exposure is exceptionally bad for your health.
While I hold the view that radiation is bad for my health we need to consider the degree of risk of an adverse outcome, also we need to consider if some effects require a dose above a threshold to exert a malign effect on our bodies. Sadly she has failed to gather or otherwise obtain the evidence required.
She wrote about the alleged radiation induced ill health of a man who worked as a construction worker building housing in a contaminated area (1986 to 1987), this man also was said to have been catching fish from lakes in the restricted areas. While it impossible for me to estimate his internal exposure from the fish, I was able to look in a UN report and see what external exposure level construction workers who were classed as liquidators were subject to. On average a construction worker in 1986 would have had 86 mSv while in 1987 the average dose was 25 mSv. The total average dose would have been 111 mSv which a dose which will have a 0.56 % chance of inducing a cancer, while this is a non trivial dose it is not one which is able to cause the acute radiation syndrome (radiation sickness).
The problem is that the man died of lung cancer and was known to be a smoker, already because the nature of his cancer we have a very reasonable and likely alternative cause of the cancer. Also the sample size was very small. Furthermore a link exists between smoking and the bone disease (Osteoporosis) which further undermines her argument that the Chernobyl event was the cause of the man’s ill health. The fact that ICRP118 indicates that at least 50 Gy is needed to cause rib fractures, also undermines Dr Phillips report of broken bones being associated with Chernobyl. I strongly hold the view that the man would never have been able to get a whole body dose of 50 Gy and live for years after the exposure.
If we look at the UN report on the effects of Chernobyl then we can see some interesting things. Many people were considered as “accident witnesses” who were on the Chernobyl site at the moment when the accident occurred and the emergency workers who were on site at 8 AM the next day then you have a population of 820 people where almost everyone had a dose of at least 500 mSv. The vast majority of these people survived the accident. At first one might reason that surely if in any population brittle bones and cancer would be observed then this population might seem to be the best place to start looking for these effects. If we assume that all the 820 people had a dose of 500 mSv then we would be looking at a collective dose of 410 manSv. This would be expected to induce 21 extra cancers (We assume a 5% additional risk of cancer per Sv for a single person and one cancer case per 20 manSv).
But the problem with this group is that it is small, quite a few who had the higher doses died off early and their exposure is very varied.
I think a better group to look at are the military men who worked in 1986, their average external dose was 110 mSv and the population was 61762 people. This gave a collective dose of 6800 manSv, I would expect this to induce a total of 340 cancers. The great problem is that the natural background for cancers in the human population is high. If we assume that 20 % of the population die of cancer and that 80 % of cancers are fatal then in this population of military personal then 15441 will have get cancer over their lifetime regardless of their exposure at Chernobyl.
So we would be looking at a population with a total of 15781 cases of cancer, the ESD on this will be 125. If we were to take a population of 61762 unexposed men then we can expect 15441 cases of cancer (ESD on this number is 124). Now the sum of the two ESDs is about 249 which is smaller than the expected difference between the two populations, so I would conclude that it might be possible over their lifetimes to see a difference in the cancer rates between the military who worked at Chernobyl in 1986 and members of the military who were stationed elsewhere.
It will be important to use a control group which is well matched to the exposed group in terms of smoking habits, diet, social back ground, drinking habits, sexual habits and other things. If you were to expose a bunch of hard drinking smoking womanizing soldiers to radiation then it would be wrong to compare them with a group of teetotal non smoking monks as a control group.
Already a large group exists in which a health effect caused by Chernobyl can be seen, this is thyroid cancer in children. If we consider two areas of Belarus (Gomel and Mogilev) then we can see a difference. In Gomel the thyroid doses of children which occurred back in 1986 were estimated to be larger than they were in Mogilev. From the UN report I have seen data for 27463 children from Gomel and 4548 from the less exposed Mogilev area.
In the years 1986 to 1989 then Gomel had nine cases of childhood thyroid cancer (the ESD on this is 3, and it was about 6 cases per year per million children. While the Mogilev area had no cases of this disease. However over the time 1991 to 1994 there were 138 cases in Gomel (ESD 11.7 and it was about 140 cases per year per million children). So it is clear that the condition has become more common.
However the data for Mogilev for 1991 to 1994 suggests that only 16 cases occur per year per million children which shows that the higher the thyroid dose the greater the health effect is. This dose response dose makes the hypothesis that the radioactive iodine from Chernobyl causes thyroid cancer to be more convincing.
These reliable statistics do allow an intelligent discussion of how safe an energy system is, these are far better than using a single case in which it is not clear what was the cause of the observed effect. I have not included data for all energy systems, to do so would require me to work for weeks on end on a single blog entry. One adverse outcome from the use of coal / oil and gas which is being considered as length is climatic change.
I think that the Chernobyl accident is at the top of the scale of what can happen, I think that the passive and active features of better reactor designs will limit the consequences of reactor accident to a smaller accident, I am involved in work which is designed to reduce further the consequences of a reactor accident. I would also say that with better designs than a RBMK1000 the likelihood of serious core damage is lower which also makes a reactor safer.
In the case of the thyroid cancer this effect can be designed out of a plant, it is noteworthy that the release of iodine to the outside world can be suppressed by the use of a sodium thiosulfate filled scrubber, this safety feature is fitted to all power reactors in Sweden. Like a Volvo a Swedish nuclear plant is designed to be crash worthy. One of these days I may well blog about the scrubber of a light water reactor.
It is important to judge a technology based on the modern product or system which is being offered rather than making your choice based on yesteryear’s model, I note that the Fukushima boiling water reactors were very old models (The Ford model T of the nuclear world) so it is not reasonable to judge the latest power reactor designs based on the old designs. I would also say that a need exists to retrofit old units or replace them to improve their crash worthiness, the first major modification I would add would be a wet scrubber between the containment and the stack.
I would have also included passive hydrogen recombiners in the containments and reactor buildings at Fukushima. While these do not slow down the core damage or the radioactivity release into the containment, they do improve worker safety during the reactor accident. The likelihood of a hydrogen explosion or fire is reduced which eliminates one threat to the workers and also helps to keep buildings in a good condition.