Back in June 2007, the Blue Ridge Environmental Defense League along with Joseph Mangano released a report on the supposed “Health Risks of Adding New Reactors to the Vogtle Nuclear Plant.”
The report analyzed the 11 counties closest to the existing Vogtle reactors for infant mortalities and cancer deaths. Based on their findings, they have concluded that the two existing reactors would cause “over 500 excess cancer deaths in Burke County” over the projected 40 year license period. The report also states that “adding two new reactors could potentially double the total.”
The mortality rates the report cites comes from the National Center for Health Statistics of the Centers for Disease Control. The website with the data can be found here.
For my rebuttal, I only looked at the numbers for Burke County, GA (the county housing the Vogtle reactors). After corroborating the data for this county, I found so many incongruities in the analysis that I felt I didn’t need to look at any other counties.
To start with, I pulled the infant mortality data off the CDC website from 1979 to 1998 which the report cites on page 16. Only 11 years of data were available but were good enough to match the results from the report. Below is a chart with the number of infant mortalities in Burke County (data not available for ’84, ’92, ’94).
Below is another chart on the airborne emissions from the Vogtle Nuclear Plant based on the data presented in the report on page 7. According to the report, the emissions are “Iodine-131 and particulates, which includes all isotopes with a half-life of more than eight days.”
Vogtle 1 began initial criticality in March 1987 and Vogtle 2 in March 1989. After looking at the two charts above, a correlation cannot be seen between the start-up and operation of the reactors, and an increase in infant mortalities. By using the correlation equation in Excel, a correlation of only 4.2% was found between 1987 and 1993 (0 percent is no correlation and 100 percent is a perfect positive correlation).
It’s convenient how the data in the report lines up to paint the picture of correlation. But that’s all it is: convenient.
Here’s a list of some of the causes of deaths for the infants off the CDC database.
Extreme immaturity (’87, ’89, ’90), respiratory problems (’85, ’86, ’87, ’88, ’89, ’90), down syndrome (’87), SIDS (’88), and meningitis due to unspecified bacterium (’89) to name a few.
It's impossible that these could be caused by the effluents of the Vogtle plant. For some reason, though, none of these were mentioned in the report. I wonder why.
Moving on to total cancer deaths.
If the reader goes to page 21 of the report, they will find selective data on the cancer deaths from all ages combined. Again, the report chooses time periods which suitably fits the perception that the Vogtle reactors cause cancer deaths. Check out my chart below on the cancer death rates from 1979-2003.
There is an obvious increase shown by the trendline in the cancer death rates for that specific period above. Let’s show only the years for the period displayed in the report (1987 – 1990).
The data clearly shows a declining trend from 1987 to 1990. See below for the yearly data for the period 1991 to 2003.
The chart also shows a declining trend for 1991 to 2003. What’s going on here? Cancer death rates for the selected periods are both declining, yet for the entire period (1979-2003), an increasing trend seems to be occurring. Well folks, that’s how you manipulate the outcome.
Page 21 of the report shows the combined rates of each period where the latter comes out to show there is an increase in the rate of cancer deaths. This data is correct. However, I’ve shown the same data broken out by year in a different light, and was able to show completely opposite conclusions.
Who is right then?
To find this answer, one needs to at least determine if there’s a correlation between total cancer death rates and total effluents from the reactors. By using Excel, the correlation between the years 1987 to 2003 comes out to 53.4 percent. This is a much higher correlation than the 4.2% found with infant mortality rates.
However, a 53.4 percent correlation is not enough. To prove there is some correlation between the reactors’ effluents and cancer deaths, the percent needs to be at least 80 and above. Some would argue it would need to be at least 90.
The report appears to have bypassed this initial screening, and moved on to data manipulation. The actual numbers are not wrong. As proved above, though, it just so happens that the authors were able to work the data to fit their presumptive conclusions.
The report analyzed the 11 counties closest to the existing Vogtle reactors for infant mortalities and cancer deaths. Based on their findings, they have concluded that the two existing reactors would cause “over 500 excess cancer deaths in Burke County” over the projected 40 year license period. The report also states that “adding two new reactors could potentially double the total.”
The mortality rates the report cites comes from the National Center for Health Statistics of the Centers for Disease Control. The website with the data can be found here.
For my rebuttal, I only looked at the numbers for Burke County, GA (the county housing the Vogtle reactors). After corroborating the data for this county, I found so many incongruities in the analysis that I felt I didn’t need to look at any other counties.
To start with, I pulled the infant mortality data off the CDC website from 1979 to 1998 which the report cites on page 16. Only 11 years of data were available but were good enough to match the results from the report. Below is a chart with the number of infant mortalities in Burke County (data not available for ’84, ’92, ’94).
Below is another chart on the airborne emissions from the Vogtle Nuclear Plant based on the data presented in the report on page 7. According to the report, the emissions are “Iodine-131 and particulates, which includes all isotopes with a half-life of more than eight days.”Vogtle 1 began initial criticality in March 1987 and Vogtle 2 in March 1989. After looking at the two charts above, a correlation cannot be seen between the start-up and operation of the reactors, and an increase in infant mortalities. By using the correlation equation in Excel, a correlation of only 4.2% was found between 1987 and 1993 (0 percent is no correlation and 100 percent is a perfect positive correlation).On page 16, the report shows the total number of infant mortalities from 1985-1987 (16), and from 1988-1990 (28). The change in the rate of infant deaths between the two periods comes out to 70.1 percent. But if you look at the chart and the rest of the data available above by year, the number of infant mortalities does not show much, if anything, of an increase after the start of the reactors.
It’s convenient how the data in the report lines up to paint the picture of correlation. But that’s all it is: convenient.
Here’s a list of some of the causes of deaths for the infants off the CDC database.
Extreme immaturity (’87, ’89, ’90), respiratory problems (’85, ’86, ’87, ’88, ’89, ’90), down syndrome (’87), SIDS (’88), and meningitis due to unspecified bacterium (’89) to name a few.
It's impossible that these could be caused by the effluents of the Vogtle plant. For some reason, though, none of these were mentioned in the report. I wonder why.
Moving on to total cancer deaths.
If the reader goes to page 21 of the report, they will find selective data on the cancer deaths from all ages combined. Again, the report chooses time periods which suitably fits the perception that the Vogtle reactors cause cancer deaths. Check out my chart below on the cancer death rates from 1979-2003.
There is an obvious increase shown by the trendline in the cancer death rates for that specific period above. Let’s show only the years for the period displayed in the report (1987 – 1990).The data clearly shows a declining trend from 1987 to 1990. See below for the yearly data for the period 1991 to 2003.
The chart also shows a declining trend for 1991 to 2003. What’s going on here? Cancer death rates for the selected periods are both declining, yet for the entire period (1979-2003), an increasing trend seems to be occurring. Well folks, that’s how you manipulate the outcome.Page 21 of the report shows the combined rates of each period where the latter comes out to show there is an increase in the rate of cancer deaths. This data is correct. However, I’ve shown the same data broken out by year in a different light, and was able to show completely opposite conclusions.
Who is right then?
To find this answer, one needs to at least determine if there’s a correlation between total cancer death rates and total effluents from the reactors. By using Excel, the correlation between the years 1987 to 2003 comes out to 53.4 percent. This is a much higher correlation than the 4.2% found with infant mortality rates.
However, a 53.4 percent correlation is not enough. To prove there is some correlation between the reactors’ effluents and cancer deaths, the percent needs to be at least 80 and above. Some would argue it would need to be at least 90.
The report appears to have bypassed this initial screening, and moved on to data manipulation. The actual numbers are not wrong. As proved above, though, it just so happens that the authors were able to work the data to fit their presumptive conclusions.
Comments
Your rebuttal is a nice piece of work. The next question is how can we get it equal airtime?
John
There's something else, unrelated, that I'm interested in, however.
Reading Southern's page about the Vogtle plant, that you linked to, they're talking about capacity factors exceeding 100% - 100.37%, 102.4%, etc. I don't suppose you'd be able to explain what a capacity factor in excess of 100% means? Does that mean that it's operating at a power output exceeding designed output? Interesting.
Capacity factors exceeding 100% are really a matter of which capacity numbers the plant refers to. I would bet that the capacities used for determining the capacity factors in the link are the summer capacities as opposed to nameplate capacity.
Summer capacities are always lower than the nameplate and are generally what's used in all of EIA's figures. I've never seen capacity factors higher than 100% using nameplate capacity. One could argue which is better to use and I'm sure it could go either way.
But since EIA uses summer capacities, I'm sure that's what Southern is doing as well.
Hope that helps.