March 14, 2008

Deaths per TWh for all energy sources: Rooftop solar power is actually more dangerous than Chernobyl

Comparing deaths/TWh for all energy sources

Energy Source              Death Rate (deaths per TWh)

Coal – world average               161 (26% of world energy, 50% of electricity)
Coal – China                       278
Coal – USA                         15
Oil                                36  (36% of world energy)
Natural Gas                         4  (21% of world energy)
Biofuel/Biomass                    12
Peat                               12
Solar (rooftop)                     0.44 (less than 0.1% of world energy)
Wind                                0.15 (less than 1% of world energy)
Hydro                               0.10 (europe death rate, 2.2% of world energy)
Hydro - world including Banqiao)    1.4 (about 2500 TWh/yr and 171,000 Banqiao dead)
Nuclear                             0.04 (5.9% of world energy)

Update: I have written a fairly comprehensive article about steps to lower deaths per terawatt hour. Primary focus on pollution mitigation. Air pollution causes 3.1 million deaths per year

A superior form of solar power would be the Coolearth concentrated solar power system which would be installed on the ground or wires over a ground installation.

Rooftop solar is several times more dangerous than nuclear power and wind power. It is still much safer than coal and oil, because those have a lot of air pollution deaths.

Rooftop solar can be safer [0.44 up to 0.83 death per twh each year). If the rooftop solar is part of the shingle so you do not put the roof up more than once and do not increase maintenance then that is ok too. Or if you had a robotic system of installation.

World average for coal is about 161 deaths per TWh.
In the USA about 30,000 deaths/year from coal pollution from 2000 TWh.
15 deaths per TWh.
In China about 500,000 deaths/year from coal pollution from 1800 TWh.
278 deaths per TWh.

Wind power proponent and author Paul Gipe estimated in Wind Energy Comes of Age that the mortality rate for wind power from 1980–1994 was 0.4 deaths per terawatt-hour. Paul Gipe's estimate as of end 2000 was 0.15 deaths per TWh, a decline attributed to greater total cumulative generation.

Hydroelectric power was found to to have a fatality rate of 0.10 per TWh (883 fatalities for every TW·yr) in the period 1969–1996

Nuclear power is about 0.04 deaths/TWh.

The ExternE calculation of death/TWh from different energy sources (not including global warming effects and is the average for European nations).

This draws on data from 4290 energy-related accidents, 1943 of them classified as severe, and compares different energy sources. It considers over 15,000 fatalities related to oil, over 8000 related to coal and 5000 from hydro.

Deaths statistics from the fuel chain for coal and nuclear

Higher level of deaths from coal in public health would be related to the increased deaths from particulates. The deaths totals are more from coal occupation are mining.

The World Health Organization and other sources attribute about 1 million deaths/year to coal air pollution. Coal generates about 6200 TWh out of the world total of 15500 TWh of electricity. This would be 161 deaths per TWh.
In the USA about 30,000 deaths/year from coal pollution from 2000 TWh. 15 deaths per TWh.
In China about 500,000 deaths/year from coal pollution from 1800 TWh. 278 deaths per TWh.

The construction of existing 1970-vintage U.S. nuclear power plants required 40 metric tons (MT) of steel and 190 cubic meters (m3) of concrete per average megawatt of electricity (MW(e)) generating capacity. For comparison, a typical wind energy system operating with 6.5 meters-per-second average wind speed requires construction inputs of 460 MT of steel and 870 m**3 of concrete per average MW(e). Coal uses 98 MT of steel and 160 m**3 of concrete per average MW(e); & natural-gas combined cycle plants use 3.3 MT steel and 27 m**3 concrete.

Wind power generation was 95 GW at the end of 2007.
1 MW produces 3,066 MWh if 35% efficient.
20 GW in Germany generated 30 TWh in 2006.
95GW would be generating about 150TWh.
95000GW would have taken 43.7 million tons of steel and 82.7 million tons of concrete. 3% of one year of global steel production. 4% of one year of the world’s concrete production. Half of one year’s production in the US for steel. About 15 deaths if corresponded to half of one years metal/nonmetal mining fatalities. 0.1 deaths per TWh. If the metal and concrete had come from China about 2700 metal/nonmetal mining deaths per year for 5 times the amount of steel. 270 deaths to get the metal for the wind turbines. 1.9 deaths per TWh. These construction related deaths are amortized over the life of the wind turbines of 30 years. Other wind power deaths need to factor in dangers associated with working with very tall structures (50 stories tall) and with deep water work associated with building and anchoring offshore.

Wind power proponent and author Paul Gipe estimated in Wind Energy Comes of Age that the mortality rate for wind power from 1980–1994 was 0.4 deaths per terawatt-hour. Paul Gipe's estimate as of end 2000 was 0.15 deaths per TWh, a decline attributed to greater total cumulative generation. By comparison, hydroelectric power was found to to have a fatality rate of 0.10 per TWh (883 fatalities for every TW·yr) in the period 1969–1996. This includes the Banqiao Dam collapse in 1975 that killed thousands.

Metal/Nonmetal fatalities in the USA (iron and concrete components mainly)

(3.1 GWp generated 2TWh in Germany for solar)

Coal and fossil fuel deaths usually do not include deaths caused during transportation. The more trucking and rail transport is used then the more deaths there are. The transportation deaths are a larger component of the deaths in the USA than direct industry deaths. Moving 1.2 billion tons of coal takes up 40% of the freight rail traffic and a few percent of the trucking in the USA.

Uranium mining is a lot safer because insitu leaching (the main method of uranium mining) involves flushing acid down pipes. No workers are digging underground anymore. Only about 60,000 tons of uranium are needed each year so that is 200 times less material being moved than for coal plants.

But what about Chernobyl ?
The World Health Organization study in 2005 indicated that 50 people died to that point as a direct result of Chernobyl. 4000 people may eventually die earlier as a result of Chernobyl, but those deaths would be more than 20 years after the fact and the cause and effect becomes more tenuous.

He explains that there have been 4000 cases of thyroid cancer, mainly in children, but that except for nine deaths, all of them have recovered. "Otherwise, the team of international experts found no evidence for any increases in the incidence of leukemia and cancer among affected residents."

Averaging about 2100 TWh from 1985-2005 or a total of 42,000 TWh. So those 50 deaths would be 0.0012 deaths/TWh. If those possible 4000 deaths occur over the next 25 years, then with 2800 TWh being assumed average for 2005 through 2030, then it would be 4000 deaths over 112,000 TWh generated over 45 years or 0.037 deaths/TWh. There are no reactors in existence that are as unsafe as the Chernobyl reactor was. Even the eight of that type that exist have containment domes and operate with lower void co-efficients.

The safety issues with Rooftop solar installations
Those who talk about PV solar power (millions of roofs) need to consider roof worker safety. About 1000 construction fatalities per year in the US alone. 33% from working at heights.

Falls are the leading cause of fatalities in the construction industry. An average of 362 fatal falls occurred each year from 1995 to 1999, with the trend on the increase. 269 deaths (combined falls from ladders and roofs in 2002). UPDATE: Based on a more detailed analysis of the fatal fall statistic reports I would now estimate the fatal falls that would match the solar panel roof installations as 100-150. Only 30-40 are classified as being a professional roofer but deaths for laborer or general construction worker or a private individual count as deaths.

Roofing is the 6th most dangerous job. Roofers had a fatality rate in 2002 of 37 per 100,000 workers.

In 2001, there were 107 million homes in the United States; of those, 73.7 million were single-family homes. Roughly 5 million new homes are built each year and old roofs need to significant work or replacement every 20 years. So 9-10 million roofing jobs in the US alone. In 2007, Solar power was at 12.4 GW or about 12.6 TWh. The 2006 figure for Germany PV was only 1TWh from about 1.5GW from $4 billion/yr. The German rate of solar power generation would mean 12.4GW would generate 8TWh. 2.8GW generates 2 TWh for Germany, assuming other places are 50% sunnier on average, then the 9.6GW would generate 10.6 TWh.

$4 billion is about the cost of one of the new 1.5 GW nuclear power plants, which would generate 12 TWh/year. Nuclear power plants (104) rated at a total 100GW generated 800 Twh in 2007.

The world total was from about 1.5 million solar roofed homes. 30% of the solar power was from roof installed units. 1/6th of the 9 million roofing job accidents would be about 50 deaths from installing 1.5 million roofs if other countries had similar to US safety. The amount of roof installations is increasing as a percentage. 4 TWh from roofs PV. So 12.5 deaths per TWh from solar roof installations. Assuming 15 years as the average functional life or time until major maintenance or upgrade is required. The average yearly deaths from rooftop solar is 0.83/TWh. Those who want a lower bound estimate can double the life of the solar panels (0.44deaths/TWh). This is worse than the occupational safety issues associated with coal and nuclear power. (see table below). 12 to 25 times less safe than the projected upper bound end effect of Chernobyl (from WHO figures). The fifty actual deaths from roof installation accidents for 1.5 million roof installations is equal to the actual deaths experienced so far from Chernobyl. If all 80 million residential roofs in the USA had solar power installed then one would expect 9 times the annual roofing deaths of 300 people or 2700 people (roofers to die). This would generate about 240 TWh of power each year. (30% of the power generated from nuclear power in the USA). 90 people per year over an optimistic life of 30 years for the panels not including maintenance or any electrical shock incidents.

Maintenance and Functional life of solar panels

[Q26. Do they require any maintenance?
A26: Only an occasional wipe to ensure optimal performance of the solar panel.]

15. How long will the panels last?
Generally, systems last 20-30 years since the waterproof seals on the panels tend to deteriorate over time.
16. If I move home, can I take the solar panels with me?
You could take your solar power system down and re-install it at your new house provided the roof of the new house is suitable. Or, you could include it in the selling price of your house. If your house is in a remote area and the solar power system is the sole source of power, the purchaser of your house would be wise to make sure the solar power system is included in the price, or they’ll be left without electricity.
[Generally hail resistant but a storm big enough to damage a regular roof would also damage a rooftop solar panel system.]

The 10 most dangerous jobs
Occupation     Fatalities per 100,000 
Timber cutters              117.8
Fishers                       71.1
Pilots and navigators       69.8
Structural metal workers      58.2
Drivers-sales workers       37.9
Roofers                       37
Electrical power installers   32.5  [also, solar power related]
Farm occupations       28
Construction laborers       27.7
Truck drivers               25

Source: Bureau of Labor Statistics; survey of occupations with minimum 30 fatalities and 45,000 workers in 2002

Nothing is perfectly safe. Chasing perfection can cause us to ignore just improving and trading worse for a lot better. Non-roof installations of solar is safer than roof installation. Nuclear, wind, non-roof solar and hydro are a lot safer than coal and oil. Natural gas is safer but not as much as nuclear and those others. The focus needs to be on getting rid of the most dangerous energy sources which are coal and oil first. Then after that decades long project is done to look at the other energy sources. Safety and improvements for all energy sources should be made as we go.

Rooftop solar is still a hundred times safer than coal and oil power because of air pollution deaths. Other ways to make solar power safer:
1. Increase safety for all rooftop work (can reduce deaths by half or more)
2. Rooftop solar tiles installed on new buildings might not have any more incremental deaths as opposed to panels that are separate from the roof tiles or systems installed that replace roof tiles before they would normally be replaced.
3. Create some new installation system where people stay on the ground using some forklift or crane to raise and place a solar power system onto a roof. Have to ensure that the heavy machinery system is safer than the roofing process being replaced.

Some responders online are in denial that people who work on a roof can fall off regardless of the reason they went up there. If I go up there to replace roofing tiles or go up there to install solar panels, the risk of falling is pretty much the same especially when the number of times being compared heads to large numbers like millions of times for each. As I noted in the comments, statistics show that 70% of fatal construction falls occur at height of 3 stories or less.

Some have also claimed that someone who went up onto a roof to install a solar panel but then fell is not a death associated with solar power. Similarly then if someone is killed in a coal mine then that is not a coal power death because the coal was not in the power plant yet or they might have some other reason for being underground and would have been crushed anyway.

189 page pdf from the 1997 Externe analysis of energy sources and fuel cycles.

Canada is increasing the planned number of nuclear reactors in Alberta to 4 plants generating 4 GW. The plan is to complete them by 2017.

Southern California Edison (SCE) plans to spend $875 million over the next five years putting solar panels onto commercial roofs to generate 250 megawatts of solar capacity. The panels will be on 65 million square feet of roof.

San Jose has a 15 year green vision to install 100,000 solar power roofs.

San Jose was chosen a Solar America City by the U.S. Department of Energy and will share $2.4 million in funding with 11 other cities. Other cities designated as Solar America Cities include Sacramento, Santa Rosa, Seattle, Wash.; Houston, Texas; Knoxville, Tenn.; Milwaukee, Wis.; Minneapolis & St. Paul, Minn.; Orlando, Fla.; Philadelphia, Penn.; and San Antonio, Texas.

Severin Borenstein, director of the U.C. Energy Institute and a professor at the University of California, Berkeley's business school, called existing technology "a loser" in a research paper. "We are throwing money away by installing the current solar PV technology," he said.

Borenstein calls for more state and federal money to be spent on research into better technology, rather than on subsidies for residential solar power systems. In his analysis, Borenstein found that a typical PV system costs between $86,000 and $91,000 to install, while the value of its power over its lifetime ranges from $19,000 to $51,000. Even assuming a 5 percent annual increase in electric costs and a 1 percent interest rate, the cost of a PV system is 80 percent greater than the value of the electricity it will produce. In his paper, Borenstein also factored in the value of greenhouse gas reductions into his calculations, and found that at current prices the PV technology still doesn't deliver.

California's Million Solar Roofs Plan, signed into law in 2006, which will provide 3,000 megawatts of additional clean energy and reduce the output of greenhouse gases by 3 million tons. The 2.9-billion-dollar incentive plan for homeowners and building owners who install solar electric systems will lead to 1 million solar roofs in California by the year 2018.

Sample solar power installation instructions

More rooftop solar panel installation instructions

Solar thermal panels for hot water heating are typically 36-75kg in weight per panel.

Solar PV panels are currently about 40-60 pounds (20-30kg).

US energy use by source


Trading Futures
Nano Technology
Netbook     Technology News
Computer Software
Future Predictions


Lobo7922 said...

I mujst say this is somehow funny XD

Those poor "roofers" should be paid better XD

Anonymous said...

Your analysis is dubious at best. Correlating random statistics of "roof worker deaths" and "electrician deaths" to solar panel installations is highly questionable. It goes without, saying that not all roofing work and electrical installation jobs are created equal (ie, high rises, high voltage lines, etc). Unless there is some kind of detailed analysis of the kind of environments in which these worker deaths occurred, it is quite a logical jump to assume that the probability of worker deaths in all roofing and electrical installations occur in the same proportions as deaths that occur during solar panel particular installations. Especially if you consider that most solar power installations occur at "low-altitude" (1-3 story buildings), and the voltages involved are relatively low (except perhaps at the interface to the actual grid).

bw said...

I did not add any deaths to the solar roof installation total for electrical deaths.

Analysis of roofer deaths

Falls from roof edges accounted for half of the fall deaths or three-fourths of the fall deaths from roofs. For roofers in residential construction, falls from roof edges accounted for 70% of work-related-fall deaths and 90% of roof fall deaths.

Solar panels are also being installed on non-residential buildings.

Distribution of heights of Washington state fatal falls

Distance of falls Deaths
1-9 feet 5
10-19 feet 16
20-29 feet 12
30-39 feet 4
over 39 feet 2
Unknown 1

So for Washington state it was mostly low altitude falls.

Roof edges and Ladders are where the accidents occur.

Workplace falls study from OSHA 1997 by Timothy Webster.
15% 10 feet or less
53% 11-30 feet
17% 31-50 feet
15% 51+ feet


bw said...

Electrocution deaths

Standing in water or having equipment such as trouble lights and extension cords touching water was a contributing factor in at least 51 electrocutions.

Low voltage (600 volts and under) was involved in the electrocutions of at least one-third of electrical workers and one-quarter of non-electrical workers. At least 14% of all electrocutions involved 120/220 volts (household current). (The voltage was not known in more than one-quarter of electrocutions.)

42 (12%) of the electrocutions of electrical workers and 88 (32%) of those of other construction workers that involved electrical wiring and equipment involved these contacts. The most common contacts were with metal ladders, metal pipes, metal wires that were deliberately cut or stripped or were accidentally cut by electric drills or other tools, wires that were energized by contact with live wires, and energized trucks and other vehicles.

Older panels usually made either 12 volts or 24 volts, and you'd use multiple panels in series for 48 volt home power systems. Now that 48 volt systems are the norm and 12 volt systems are becoming very rare, high-voltage panels are commonly available, running at 65 volts or more.

Maximum Power Point Tracking (MPPT) controllers. These new controllers can increase the output of any solar panel by 15-20 percent by trading volts for amps (and vice versa) to adjust the array output to changing conditions (clouds, snow on the ground, the angle of the sun). They also accept high voltage input directly, so we are frequently seeing 100-135 volt solar arrays.

Watson solar house
On our roof are 48 solar modules, each producing 110 watts DC for a total of 5.28 kilowatts. Power comes down from the panels on the roof as two ~400 volt DC circuits (rated at 600v). That's the sort of voltage that can kill you.

Cyril R. said...

BW, you still haven't actually answered the question anonymous posted.

Unless you can actually provide statistics of deaths caused explicitly by installing PV or maintaining/cleaning the system, this is just bad science.

Besides, you don't have to go up the roof for cleaning the panels, one of those long cleaning probes that cleaning companies use works just fine in most cases.

Oh, and there's this state of the art technological development that's called a ROPE.

Otherwise it's an excellent analysis and an interesting read.

bw said...

The accidents are primariily occuring as people climb up and down from roofs. At the high end of the ladder and at the edge of the roof.

To say that there are no accidents for PV installs is to say that when you install PVs you do not have to climb up the last part of the ladder or cross over the edge of the roof. Did they have some method for transporting PV installers up and securely attaching them to roofs. Are the small fraction 37 in 100,000 roofers who have fatal problems..none of them had crossed over into the million or so global PV installs ? PV installers do not have the same sloppiness/accident distribution as everything else in the world. When you get up into millions of events then you get that small fraction of the time when people did not follow safety or screwed and had the freak aor stupid accident. As yet there is not much tracking of PV specific accidents. Only 100,000 installs or less in the US. I have not seen stats for Germany and Japan where they have had more roof installs.

Also, if one believes that all maintenance is done from the ground then use the 30 year life instead of 15 year life for panels. I believe people can use a high pressure hose from the ground to clean the panels. (but again I bet that there are some people heading up on the roof to clean etc...). The people who are having the accident are taking shortcuts and not thinking about the dangers.
So, I believe I addressed the anonymous question.

50% of car accidents are at intersections, would it be the case that cars that are carrying solar panels are immune to collision in intersections ?

bw said...

Obviously accidents especially falls can be avoided. Just like it is possible to not spill cups of water or milk or beer. Yet sometimes people get sloppy and that is when the accident happens. So one can tie ropes and do things to prevent the accidents, but not if someone is in a hurry or they decide screw it even though it rained or is raining I am going up on that roof. I don't want to come back next weekend. I got to hurry up and get this done so I can do the next job or get to the game. Whatever safety practices there are this kind of stuff happens. A bigger system can be designed to tolerate the split second stupidity, but small cheap and frequent things are not engineered like that. Thus airplanes can be safer than cars which can be safer than bathtubs in the home.

Cyril R. said...

No sorry but you haven't actually answered the question at all.

Unless you have documented cases of PV related deaths it's still just bad science.

No empiric evidence, no proof of correlation. I have Googled for it intensively, but could not find any cases documented.

Do I really have to explain why generalising from correlations is really, REALLY bad science?

bw said...

You can choose to believe that going up a ladder and onto a roof is somehow different if you are working on roof shingles or working on solar panels. You apparently have a scientific reason for this difference. Anti-gravity or special solar panel air bags. I do not see the reason for the special case.

bw said...

I guess the same generalizing correlations would apply to denying that toyota prius and electric cars get into the same ratio of accidents as other cars until someone turned up the specific fatal accident rates.

different kinds of cars do have slightly different accident and fatality rates but with high volume cars the accident rates are in a close range. a safe class of driver or a particular safe car might be ten times safer than a particularly unsafe type.
Because of self selection in terms of the drivers as well as characteristics of the cars. but note that no cars with significant volume have no death rate.

you are just assuming that there is some massive difference in the behavior of the same installers of roofs and roof top solar. the houses are not different. the ladders are not different. There is significant overlap in the people who do the work. So on the days they are installing solar they are immune to falls, they magically only can screw up on roof tile or replacement days.

Cyril R. said...

Thank you for not answering my question.

I am not asserting no one has died from installing PV, just wanted to point out a methodological flaw in your argument. You haven't proven anything yet. You have presented an hypothesis about how many you think have died from installing and maintaining PV. What science is all about is not generalising from one level of correlations to another, but to prove the hypothesis through empiricity.

If you don't complete that last step, then it's not science. It's you who's doing the assuming here!

Analogues are not useful when trying to determine empirical facts. Statistics are. Where are the statistics?

I apologise if this is a bit nitpicking (deaths from PV and nuclear are both very low), but I'm just curious how many have actually died.

bw said...

I have not found direct statistics for tracking this. For the USA, it probably is not a separate category from other roofing deaths. US has too few 100,000 or so solar rooftop installs over 2 decades. Probably only 0 to 2 incidents per year.

Germany and Japan should have more but probably not over 5-10 per year maximum. I have not found any sources that break out even the German or Japanese construction deaths (in english language online sources). Probably a search in the native language would be needed.

However, since rooftop installations and solar is growing strongly 30-40% per year. Then accident numbers will become more significant.

In terms of science, this is a relatively detailed blog post and not a submission for peer review. That being said, I am confident of my logic and conclusions. I think the variance in actual numbers will be plus or minus 50%. I could easily see solar roof installations having higher accident rates than regular roofs. The panels are smooth/low friction, they are usually raised above the roof a few inches and could be a tripping hazard as they are installed, extra wires and frames for catching and tripping, people getting up a learning curve learning how to install them safely etc...

Cyril R. said...

However, since rooftop installations and solar is growing strongly 30-40% per year. Then accident numbers will become more significant.

Maybe not. Wind is on a lower number of deaths per energy generated trend as the total installed base increases. It would be presumptious to state that the same couldn't apply for solar.

When someone dies as a result of a company accident, companies will investigate to mitigate the risk in the future. Of course, there will always be people who want to install the panels themselves, and some of them may be reckless.

Darwin would say that if it were a significant cause for deaths, evolution might take care of it automatically.

clee said...

You estimated there are 9 million US roofing jobs per year, and that 1.5 million PV roofs should have 1/6 of the accidents that 9 million roofing jobs have. You said 1/6 is about 50 deaths, which implies 300 roofing job deaths per year in the US. And yet there are only 51 roofer deaths per year according to
1/6 of that would be 9, not 50.

If you are going to count 300 construction fatalities per year from working at heights, then you have to count more than 9 million roof jobs. You have to count non-roofing jobs where people die falling off of ladders.

bw said...

I was estimating 1.5 million PV roofs worldwide total ever.

Maybe 500,000 PV roofs in the world last year.

So the worldwide number of deaths from falling off roofs worldwide lastyear would be about 17.

The US number would be 0-4. I think the US might have installed 10,000-50,000 PV roofs last year.

clee said...

Maybe I wasn't clear. 1.5 million roofs worldwide total ever is 1/6 of 9 million roofs. 9 million roofs generated only 51 roofer accidents. So 1.5 million PV roofs worldwide total ever should generate 51/6 = 9 PV roofer accidents worldwide total ever.
The analogy is
9 million : 51 roofer deaths :: 1.5 million : 9 roofer deaths
or to continue with last year only
500,000 : 3 roofer deaths (world wide)
50,000 : 0.3 roofer deaths
10,000 : 0.06 roofer deaths (US)

bw said...

Previous research has identified falls as a leading cause of fatalities in the construction industry. Falls
accounted for 25% of the construction deaths identified by NTOF for 1980 through 1989,7 32% reported
by the 1996 CFOI,3 and 33% of construction fatalities investigated by the Occupational Safety and Health
Administration (OSHA) between 1985 and 1989.6 Falls comprised 29% of construction fatalities in
Washington State between 1973 and 1983,8 and 46% in New Jersey between 1983 and 1989.9
Occupational groups identified with high frequencies of fatal falls include roofers, painters, ironworkers,
carpenters, construction laborers, and tree trimmers.

The Census of Fatal Occupational Injuries (CFOI) of the Bureau of Labor Statistics (BLS)
reported that falls to a lower level comprised 9.9% of fatalities in 1996.3 The CFOI data identified roofs,
ladders, and scaffolds as the most common fall locations. Other studies have noted falls associated with
steel erection, falls from equipment or materials, and falls through floor openings. 540 deaths per year, from 1980-1994. Construction (4044) 49.9%. Half of those were roof, ladder, scaffold or unspecified.

40% by people in a new line of work less than 6 months.

So there are more deaths per year when it is not just the person officially classified as a roofer. If the person is not a card carrying member of a roofing guild, such as a laborer or a general construction worker or a private individual who is performing a solar power installation then I would consider it a solar power installation death.

There are also electrocutions and dropping something onto someone below.

Of the work related 8102 fatal falls (1980-1994)

1230 were from a roof
1055 were from a scaffold
994 were from a ladder (20% real estate related, 200 real estate ladder)
1297 unspecified location

From 1990-1994 for SIC code categorized (fewer years and where a SIC code was specified which is fewer)
147 roof SIC code 1761
63 General contractor single family
59 electrical work
56 building cleaning and maintenance
71 Painting
79 Carpentry

I would say that it is still at least 100/year that would fit into the solar power installation related categories and could be as high as 150/year. Plus if it being done by the homeowner or a laborer then the solar power installations would skew higher. (40% by first timers, amateurs who do not follow proper safety and have no protective gear etc...) Plus there is not enough calc on electrocutions or the guy up top dropping something heavy onto someone below.

1handclapping said...

Unfortunately, all of your comments neglect the real basis of danger associated with nuclear plants.

Chernobyl, though an awful accident was only a mild example of how bad a nuclear accident could become.

Nowhere do you factor the complete risk/reward profile for nuclear into your argument, so (other issues in your argument aside) the implication that nuclear is safer than rooftop PV rests on your assumption that Chernobyl is as bad as it gets --- sorry, but that's just not the case!

bw said...

Nuclear reactors have containment domes. Chernobyl's reactor did not. Safety has been improved at nuclear reactors with better designs and operation.

Radiation treatments appear likely to be radically improved very soon

It is possible to make nuclear bombs 10 to 1000 times less fatal with better construction and radiation treatments and prevention

I have seen some of the scenarios of "worse than Chernobyl accidents". Only some reactors would have this possibility be being located near richer areas like New York. Not that it would happen but if you write off all of a poorer state then you do not get to that high a price tag.

I do not find those scenarios credible, plus I have indicated that the deployment of new radiation treatments would make any reactor accident almost harmless death wise.

1handclapping said...

Your response to my argument is based on so much that is speculative for its justification, that I suspect you're just *DETERMINED* to believe that nuclear is safer, but where is your analysis of that "poorer state" you're writing off?

How many thousands of deaths from *ONE* incident does it take wipe out 20 years of PV installation accidental deaths? You claim to be making a rational case, but it's far from logical.

bw said...

Where is your data 1handclapping ? where is your proof ? You make assertions but have shown no substance.

In the article there is the links to the historical data and peer reviewed reports which shows that nuclear power has been safer.

You claim there will be worse accidents with nuclear power than Chernobyl. Where is the analysis to justify this claim based on the existing nuclear reactors and based on new build ?

What are the scenarios where containment domes are breached ?

I have presented data and analysis you have presented squat.

My "speculation" was to consider safety is coming soon that will even make nuclear bomb detonations a lot less fatal. There is no way that a nuclear power plant can go off like even a small nuclear bomb because nuclear bomb material is enriched to 90+% and nuclear plant material to about 5%.

1handclapping said...

I did not claim that there would be worse accidents only that there *could* be worse accidents.

I am well aware that a nuclear reactor is not a nuclear bomb. I am also well aware that a nuclear reactor's core in meltdown is quite capable of acting like a thermite bomb on steroids.

With the kind of heat this molten core could reach, it has (and I *have* read the research papers) only been conjectured just exactly what could contain such a core -- to the best of my knowledge there is no actual experimental data proving containment under these conditions. And yes, I've read the papers that argue that such a meltdown must be inherently self-limiting, but those are not proof, just arguments.

We do have the example of Fermi's first tiny reactor though, in case there's any doubt that meltdown is a possibility.

Scenarios proposed by nuclear power opponents suggest that if molten core like that were to hit a water table, the resulting radioactive vapor release could easily lead to the deaths of thousands of people.

I used to read heavily in this area, but it's been awhile and I have no idea of how much of this stuff has made it to the Web -- my past experience in almost any field has been that most research older than about a decade is invisible to the Web and most of the stuff I'm recalling is at least 20 years old, but it seemed to be common knowledge then.

Frankly, I'm pretty surprised that you're questioning whether a nuclear reactor failure could produce a major human catastrophe. I was certainly not under the impression that anyone in the nuclear community really doubts that this is possible. They may believe it's highly unlikely but that's not the same as impossible.

Johan said...


you seem to be the one speculating, TMI did demonstrated what a meltdown is like. Sure it was a partial meltdown but a full meltdown isnt expected to be much worse. In addition the phebus facility has given plenty of real data on severe accident. How the core material behaves, how the filter behaves etc.

A LWR meltdown cant eat its way through the containment down into the water table. Thats china syndrome fantasy and not reality.

The scenarios required for a light water reactor to cause a disaster even close to chernobyl are so unlikely that they might as well be considered impossible. Using them as argument against nuclear power is no more reasonable than arguing against airplanes due to the slim risk that two fully fueled 747 might crash over a packed sports stadium killing everyone.

If we expand the conversation to more advanced reactors like pebble bed reactors then the risk of disaster is more or less eliminated since a meltdown isnt even physically possible.

bw said...

Here is an 88 page study from 2005 about a severe meltdown analysis Proper design even in a full meltdown case does not get through more than 8cm of the concrete.

There has been plenty of analysis of maintaining containment integrity

Severe accident analysis for five nuclear plant models

Tom Craver said...

Based on the work death numbers presented, it appears that both society and at least some workers find up to 1/1000 worker deaths per year tolerable, given the benefits.

So the issue is probably moot to the question of whether society should/will adopt widespread use of solar roofs or nuclear power.

bw said...

Yes, society finds the historical distribution of 55 million deaths per year worldwide overall and worker deaths of 1 in 1000 tolerable so long as those deaths are good old fashioned coal mining, sweetly familiar traffic accidents, air pollution from coal, oil and natural gas and industrial accidents that are not radiation leaks from a nuclear plant, radiation from coal burning and natural gas are okay.

Falls from roofs are deemed to be the fault of sloppiness from the installer or roofer or worker.

People should consider at least better options in the category of solar power. Concentrated solar power and stringing metallic balloons over cropland as cheaper, faster and safer (like Coolearth).

1handclapping said...

A few responses to your comments, BW:

“Here is an 88 page study from 2005 about a severe meltdown analysis Proper design even in a full meltdown case does not get through more than 8cm of the concrete.”

And here is a relevant excerpt from that very same report:

“Despite the uncertainties in the heat transfer rates, the maximum possible ablation depth
can be calculated by assuming that all the heat from the melt goes to the melting of the
concrete. However, this method fails if spalling, or cracking of the concrete surface,
takes place. Spalling has been observed in fire tests of concrete, when the water in the
concrete pores evaporates rapidly, increasing the pressure inside the concrete and
generating stresses that exceed the strength of the concrete. Spalling of concrete under
the pouring of molten material has not been studied systematically. In this work, an
experimental facility to examine this phenomenon has been designed.”

Note that this possible (and perhaps likely, based on observations cited in the excerpt itself) failure mode is yet to be studied, in fact a significant part of the report is devoted to a proposal for a research facility and methodology for such study.

And here is an excerpt from the NRC report you cite regarding:

“Severe accident analysis for five nuclear plant models”

which you seem to imply supports your argument that these accidents cover the gamut of worst case possibilities:

“4.6 Containment Performance
The lack of information about many of the physical phenomena that determine the performance of a containment system in a severe accident situation is such that only educated guesses can be made for some sequences that might make significant contributions to risk. Although the large number of event trees developed in the containment analyses is indicative of what was hypothesized by the analysts, the amount and quality of information concerning a number of key phenomena that determine behavior at branch points are low. The difficulty of arriving at a result with significant confidence is illustrated by two examples. In the analysis of the performance of the Mark I containment used in early BWRs, the experts in the original study predicted a large conditional probability of early failure. In the second study a different group of experts produced a bimodal distribution because part of the panel concluded that the probability of early failure was high, and part considered it low. A second example is the calculation of risk produced by postulated direct containment heating (DCH). In the first study, the calculated risk due to DCH for PWRs with large dry containments was a major contributor to the total risk. In the second version, its contribution was significantly less. In neither case had there been a major change in the information about relevant physical phenomena available at the time of the first study. Further, we find no consideration of the impact of ex-vessel steam explosions on early containment failure. There is little unambiguous guidance here for a licensee performing an IPE.”

Finally, there is the September 2004 study, “Chernobyl on the Hudson? The Health and Economic Impacts of a Terrorist Attack at the Indian Point Nuclear Plant,” by Dr. Ed Lyman of the Union of Concerned Scientists, which, using the NRC’s own analysis method, found that a worst-case accident or attack at the Indian Point nuclear plant 35 miles north of New York City could cause up to 43,700 immediate fatalities and up to 518,000 long-term cancer deaths. Such a release could cost up to $2.1 trillion, and would force the permanent relocation of 11.1 million people.

That’s a pretty big risk to offset even if its likelihood is low, and that’s been my point all along --

Low probability uncertainties with big consequences should not simply be ignored.

bw said...

Yes, I have seen the 2004 study.

It assumes super-terrorists making a successful attack on Indian Point and then assuming that optimal weather and optimal everything else for maximum casualties.

I do not believe the starting point of the scenario. Successful hijacking of a plane [Since 9-11 pilots do not come out of the cockpit no matter how many passengers or crew are killed. If you are worried about terrorists doing this then do not secure all of the rich targets in the USA or other places but kill terrorists like Al Qaeda which has been done. thus no major terrorist operations] and running it into a reactor without being shotdown.:
The report considered an attack on
the Biblis B PWR by a small jet (Airbus A320) or medium-sized jet (Airbus A300) travelling at speeds from 225 to 394 miles per hour, where the peak speed of 394 mph
was determined through the use of simulators. GRS concluded that for an event in which the jet did not penetrate the containment, but the resulting vibrations caused a primary coolant leak, and the control room was destroyed by debris and fire (a condition similar to a station blackout), then control of the sequence of events would be “ uncertain.”
Biblis B was designed for protection against the crash of a 1960s-era Starfighter jet and
as a result is equipped, like most German reactors, with a double containment. In contrast, Indian Point 2 and 3, while of the same 1970s vintage as Biblis B, were not
designed to be resistant to airplane crashes, and do not have double containments.

I do not believe the radiations deaths figures that are quoted or the resultant cancer deaths.

the $2 trillion is based on the 99.9 percentile case of a BS scenario. 95% was $1 trillion.

Nuclear plant security is adequate

Anti-Radiation drugs far (5000 times ) better than potassium iodide are being developed.

Spalling concrete with material going deeper into the floor concrete and the ground does not create the bad scenarios that you are postulating.

Indian Points is safe

Independent safety evaluation came of Indian Point conclusion was it is safe

Indian points will stay open

The full independent indian point safety report is here

Pg127-133 are the relevant part for the airplane strike scenario.

Visual inspection of the concrete for signs of cracking and spalling are required by the
American Society of Mechanical Engineers (ASME) Section 11 IWL Code and are
performed regularly. The structural integrity of the buildings and its leak tightness are
regularly verified by testing. The buildings were pressurized to 54 pounds per square inch
(psi) (115 percent of the accident rating) – Unit 2 was tested in March 1971 and Unit 3 in
January 1975. Integrated leak rate tests are performed periodically to pressurize the
containment buildings to 47 psig and measure total leakage. During these tests, the
building expands several inches and the concrete is therefore expected to experience
minor cracking due to the physical growth of the structure. Visual inspections are
performed during and following the test to observe for unexpected cracking or spalling of
the concrete. The last tests were performed on Unit 2 in 2006 and on Unit 3 in 2005 with
no structural concerns identified. More details on this issue are contained in Appendix 3.

Numerous tests and analyses by research organizations conclude that the
large commercial aircraft and turbojet engines in use today would not
penetrate a containment structure like that at IPEC, even on a direct hit at 350 mph. Structural integrity and leak tightness of the buildings would be maintained. Furthermore, the energy of impact would be absorbed by the structure, causing only minor movement, and would not dislodge or damage equipment on the interior of the building.

Johan said...

1handclapping, you wrote

"That’s a pretty big risk to offset even if its likelihood is low, and that’s been my point all along --
Low probability uncertainties with big consequences should not simply be ignored."

But why restrict ourself to nuclear in that case. The probability that two fully fueled 747 could crash over a packed sports stadium killing everyone is non zero. Should we then decide to stop airplaines alltogheter, even though they during regular operation is the safest mode of transportation?

What about CDC and USAMRIID where they research extremely dangerous diseases, the probability that ebola could get out of one of those facilities is also non zero. Shud we then shut down all research on dangerous diseases?

What about terrorist highjacking and blowing up a LNG tanker in new york harbor? Stop natural gas alltogheter?

What about chemical plants, oil refineries or what about the energy source that has caused the most devestating disasters of them all, hydropower?

Is it only nuclear that has to be perfectly safe? Or do you want to ban any industrial activity that has a risk of large disasters no matter how small the probability?

You can not find any serious nuclear engineer that would agree with the statement that chernobyl wasnt the absolute worst case plausible scenario.

Tom Craver said...

If a risk is obvious, personally avoidable by taking care, and taken on voluntarily, it is considered ethically acceptable, even if statistically some level of deaths is pretty much certain.

So solar roofing risks, despite being objectively much higher, are much more acceptable than nuclear risks for most people.

Coal is a special case - we've lived with it's risks for centuries, and pollution from it has already been cut substantially - so people assume the risks are acceptable.

Justin said...

Did you count any of the accident related deaths of construction workers who died building nuclear plants?

vfx said...

What about Hydroelectric dams bursting?

Michael Rule said...

It is perfectly fine to cobble together an expected estimate of deaths from solar installation using existing statistics from other roofing and electrical jobs. Of course, this would be a prediction that could be tested against the actual data on the hazards of roof-top solar, once more data become available.

Iridos said...

Hm, seems you might have to revise your assessment in some time now.

I still see a couple of flaws there... there is a very high uncertainty on a lot of those numbers - i.e. "Chernobyl: Consequences of the Catastrophe for People and the Environment" comes to nearly a million deaths caused by Chernobyl -- a number vastly different from the 4000 predicted. Another thing is, that the half-life of Cs-137 is about 30 years - it's a long time to cause an unknown amount of deaths, that will look just like cancer or just like even something completely else.

There's another metholodical problem. Plutonium has a half-life time of 24000 years, there are fission products with even longer half-life times. It gets a bit philosophical to ask if humanity will live long enough to see those decay, but as we have seen with the most recent disaster (or will see, as may be), the stored used fuel rods turned out to be a major problem, and freed radiation can stay a problem for the next couple of ten thousand years to come.

So how did you account for the deaths caused by this over the period of the next 10000 years?

Alan said...

Ok so roof top solar may or may not be as dangerous as Chernobyl. Who cares if they are even close? What does that matter, when neither source supplies that much percent of power to us.
How about topping this danger with any of your nuclear power worst scenario events.
The World Health Organization estimates that urban air pollution causes 800,000 premature deaths each year. Fossil fuels burned by motor vehicles contribute 90 percent of urban air pollution, including lead, carbon monoxide, ozone and suspended particulate matter. The World Health Organization reports that 3 million people now die each year from the effects of air pollution.
This does not include mining deaths, radon from coal ashes, or the increased radiation from the sun as these things deplete the ozone.
It serves no one to compare nuclear to roof-top solar other then to say it is not necessarily much more dangerous as things stand today.
It is better to compare it to the other mass energy production used today. Which is obviously dangerous to everybody's health on a daily basis. No earth quake, no tsunami, no terrorist attack needed. Just start up your car and drive it for a few minutes!!

Alan said...

I am not sure, don't have time to look right now either. But I was lead to believe that batteries are an integral part of solar systems. If so what are they made of? What are the dangers with their manufacturing, handling by owner, and disposal? What are the possible long term affects of the their disposal?

pablo said...

Michael Rule brings up the logical point that you are comparing a number of deaths caused by an operating (or dysfunctional) nuclear plant with the installation costs of solar power.
Once a pv system is installed, the probability of it causing death is extraordinarily small. On the other hand, Fukushima and Cherynobyl show us that the danger of a Nuclear power plant continues indefinitely after construction, in fact the potential for leaks and other failures increase as the nuclear facility ages. Aging equipment and reduced inspections due to cost-cutting by utilities may combine to produce disastrous results, as many plants are approaching or have exceeded their 40 year design-life. Aging solar installations don't really present much danger.