Slashdot Mirror
38,000 People a Year Die Early Because of Diesel Emissions Testing Failures (theverge.com)
An anonymous reader quotes a report from The Verge: Diesel cars, trucks, and other vehicles in more than 10 countries around the world produce 50 percent more nitrogen oxide emissions than lab tests show, according to a new study. The extra pollution is thought to have contributed to about 38,000 premature deaths in 2015 globally. In the study, published today in Nature, researchers compared emissions from diesel tailpipes on the road with the results of lab tests for nitrogen oxides (NOx). The countries where diesel vehicles were tested are Australia, Brazil, Canada, China, the European Union, India, Japan, Mexico, Russia, South Korea, and the U.S., where more than 80 percent of new diesel vehicle sales occurred in 2015. The researchers found that 5 million more tons of NOx were emitted than the lab-based 9.4 million tons, according to the Associated Press. Nitrogen oxides are released into the air from motor vehicle exhaust or the burning of coal and fossil fuels, producing tiny soot particles and smog. Breathing in all this is linked to heart and lung diseases, including lung cancer, according to the International Council on Clean Transportation, which took part in the research. Governments routinely test new diesel vehicles to check whether they meet pollution limits. The problem is that these tests fail to mimic real-life driving situations, and so they underestimate actual pollution levels. The researchers estimate that the extra pollution is linked to about 38,000 premature deaths worldwide in 2015 -- mostly in the European Union, China, and India. (The U.S. saw an estimated 1,100 deaths from excess NOx.)
Considering that most Americans have less than $10,000 saved, dying early might actually be a good thing.
Seven puppies were harmed during the making of this post.
38,000 out of nearly 8 billion is hardly worth worrying about.
This to a point. I hate to say it, but dying a few days or weeks early is not a good measurement. How many healthy people are affected by this is a better measurement. My grandmother was considered a victim of mesothelioma because she worked in a plant that made asbestos shingles, but she died at 93. We should measure the effect in years of loss of expected life rather than in just pure numbers.
https://www.nasa.gov/topics/ea...
According to a new paper by Ott and Pickering in the Journal of Geophysical Research, each flash of lightning on average in the several mid-latitude and subtropical thunderstorms studied turned 7 kilograms (15.4 pounds) of nitrogen into chemically reactive NOx. "In other words, you could drive a new car across the United States more than 50 times and still produce less than half as much NOx as an average lightning flash," Ott estimated. The results were published July.
When the researchers multiplied the number of lightning strokes worldwide by 7 kilograms, they found that the total amount of NOx produced by lightning per year is 8.6 terragrams, or 8.6 million metric tons. "That's somewhat high compared to previous estimates," said Pickering.
https://en.wikipedia.org/wiki/...
One estimate is that 24,000 people are killed by lightning strikes around the world each year and about 240,000 are injured. Another estimate is that the annual global death toll is 6,000.
love is just extroverted narcissism
Further, there are multiple oxides of Nitrogen.
Nitric oxide, NO, is used by the body as a regeneration signal. An acquaintance has a job designing NO generators for battlefield use, either as part of an O2 delivery system (to help injured lungs) or to be directly applied to burned skin (to help with regeneration). He's associated with MGH in Boston. They will probably be seeing use in civilian hospitals in a couple of years.
The other oxides dissolve in water to form acids, but NO is actually beneficial.
I wonder if the NOx studies take that into account.
Here is how this works.
You didn't read the paper. I did.
The basic idea in the method is that they ran a number of simulations of where diesel vehicles are estimated to be driving in the 11 jurisdictions (based on real-world measurements), where emissions would end up being carried (and dissipated) by the atmosphere, and matching that up with population estimates. This was then applied to the current best estimates of the exposure-response curves to get an estimate of the number of deaths.
They ran with measured emissions to validate that it matches up with measured death rates. Then they reran with "theoretical" emissions (i.e. if the vehicle emissions actually were what they were supposed to be) and subtracted the two. Actually, they probably ran the simulation multiple times to get a confidence interval; more on this later.
NOx, ozone, and PM2.5 ("soot") were accounted for separately. This is clear from the paper, but not clear from the
The model was validated against real-world studies to make sure that it matched what we find in the field.
Now, if you're thinking that's a lot of assumptions, you're right. The 95% confidence interval is 23,000–47,000, which is quite a wide margin of error. Again, that is not clear from the writeup, but it's clear from the paper.
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
And in the same period, how many people died as a result of pollution from ordinary gasoline automotive emissions?
The question you should be asking is: How accurate are the emissions tests for "ordinary" automobiles?
Because that's what the paper is trying to estimate. Not how many people are killed by vehicle emissions, but how many people are killed by inaccurate vehicle emission testing.
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
But the right wing goes even further. Once they give a person heart disease or cancer from burning coal they also don't want them to have medical care. They excuse all this nonsense as a monetary issue. But nobody counts the costs associated with heart or lung cripples and the long term disabilities that eat up the national budget. One sick person can run up millions in public expenses.
That's exactly why we can't afford to have universal healthcare! We gotta drill, baby, drill and you want death panels! #Palin2012! ;)
Anons need not reply. Questions end with a question mark.
I am a big diesel fan, but your head is up your ass. Diesels ALWAYS generated NOx. ALL of them. Hell, EVERY internal combustion air-breathing engine generates plenty. As long as 4/5 of the atmospheric air they use for combustion is NITROGEN, there is no way in hell you can prevent some of the nitrogen forming oxides.
It's time to just accept that emissions testing is never going to work right and will always be cheated,
As I read the article (not having read the referred-to paper):
This wasn't about cheating. This was about the government-prescribed test cycles not correctly modelling the actual average driving cycle of the world's fleet of vehicles, the academic's model of the actual fleet emissions being somewhat off, the error happening to be on the low side, and recent measurements updating the model.
The mandated emissions testing cycle was KNOWN FROM THE START to be only roughly representative of overall vehicle usage. (That's why "your mileage may vary". Though on-the-road emissions are hard to measure, mileage is easy, and mileage during the test cycle falls out of a simple calculation of carbon emission as CO, CO2, unburned hydrocarbons, and exhaust volume. And it is - that's where sticker mileage comes from. When they systematically mismatch you know the test cycle mismatches average usage.)
And this was FINE. As long as the testing was REPEATABLE, ROUGHLY PROPORTIONAL to the actual cycle emissions, and covered all the common driving modes, it was an effective tool to drive emission-reduction engineering and legislation. The engineering might tend to improve the test results slightly more than fleet average (because that's where the selection pressure of the evolutionary algorithm happens to be). But it still drove a DRASTIC than fleet average (because that's where the selection pressure drop in emissions and a resulting drastic, and measurable, improvement in air quality.
Cheating consists of having the engine control recognize whether the vehicle is running the test cycle and changing its behavior - being good on emissions during the test, bad on emissions but better on performance and/or mileage when not on the test.
Voice of experience here: I wrote emissions analysis and engine control software for two of the Big Three during the 70s and 80s.
We all knew that it would be possible to write cheats into the code. And we all knew not to do it - because it would not only be bad for the environment, but it would also eventually be caught, and cause the company massive trouble. So the engineers were very careful not to do it, the managers were careful not to give the engineers an incentive to do it, and upper management put in drastic organizational controls to both prevent and detect it.
Fast forward a couple generations of executives and you find:
- A German automaker with executives who decided that cheating was the way to go.
- An American automaker (the former big-three company I didn't work for
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way