Obama Helped Make Cars More Efficient, but Now They Spew Black Carbon

Yves here. Cars as a case study of unintended consequences.

By Emily Pontecorvo. Originally published at Grist

If you’re getting more bang for your buck at the gas pump today than you were a decade ago, you can thank the Obama administration. Obama’s 2012 updates to the Corporate Average Fuel Economy standards, or CAFE standards, made new cars more fuel-efficient and reduced their carbon dioxide emissions. But a studypublished this week casts a dark shadow over that success story. Researchers found that the path we’ve chosen to better fuel economy could end up costing the U.S. hundreds of lives each year.

“I think this was a classic case of unintended consequences,” said Rawad Saleh, an assistant professor at the Air Quality and Climate Research Laboratory at the University of Georgia and a co-author of the study.

Here’s what happened. To get in line with the CAFE standards, automakers leapt at a technology called the gasoline direct injection, or GDI, engine. In 2008, GDI engines were in a mere 2.3 percent of new vehicles sold in the U.S., but by 2018, that number jumped to 51 percent. The EPA expected it to rise to 93 percent by 2025 under the rules set out under Obama, which the Trump administration is trying to roll back.

Clayton Aldern / Grist

Before GDIs came along, most cars had port fuel injection engines, or PFIs. With a PFI engine, gasoline mixes with oxygen from the air, forming a vapor, before reaching the engine. In a GDI engine, the liquid fuel is injected directly. GDI engines produce more power for every drop of fuel, but the fuel burns less uniformly. The result is a car that emits less CO2 per mile driven, but the incomplete combustion creates more particulate pollution, like black carbon.

Black carbon is one of many pollutants known as “particulate matter 2.5” or PM2.5 — tiny, inhalable, solid particles that are 2.5 micrometers and smaller and are linked to respiratory diseases, asthma, and premature death for people with heart or lung disease. Exposure to black carbon has also been associated with cancer and birth defects. Unlike CO2, a gas that hangs around in the atmosphere for hundreds of years, black carbon particles float around for a few days to a few weeks before falling back to earth. The dark surface of the particles directly absorbs radiation from the sun, adding heat to the atmosphere and contributing to climate change.

Clayton Aldern / Grist

Saleh said there has been a growing body of evidence documenting the dirty little secret behind GDI engines ever since they hit the market. His team’s new study puts the issue into perspective by calculating the public health and climate costs of switching the entire U.S. vehicle fleet over to these engines. The researchers found that such a change could almost double the number of premature deaths associated with PM2.5 vehicle emissions, up to 1599 per year from 855 in 2011. The health risks would be most severe in neighborhoods surrounded by major roadways, which are disproportionately communities of color that already suffer from poor air quality, like Oakland, California.

When it comes to the climate, the question is, will the benefit of reducing CO2 outweigh the cost of increased radiation from black carbon emissions? On a global scale, the data says yes. But since black carbon is short-lived in the atmosphere, and will be more concentrated over vehicle-dense areas, the researchers also calculated this tradeoff at the regional level. They found that in areas with lots of cars, the heat absorbed by black carbon will be greater than the heat averted by GDIs’ CO2 reductions.

That doesn’t mean cities like Oakland are going to be veiled in a shroud of warming due to these particles. Drew Shindell, an earth science professor at Duke University who was not involved with the study, said that just because the radiation is absorbed in a particular place, that doesn’t mean there will be a climate response in that place. His research group, along with NOAA and the National Center for Atmospheric Research, have run models on this question, and found that the effects would be smoothed out over the whole Northern Hemisphere. “They are consistent in finding that just because you change radiation in one small spot, you don’t get a response just in one spot,” he said.

There’s one caveat to all of this, which is that there is still a lot we don’t understand about how black carbon interacts with the climate. In addition to the solar radiation it absorbs, it can also have indirect effects on clouds. Black carbon particles are little surfaces that water can condense on, contributing to cloud formation, and those clouds also affect solar radiation. Shindell said some studies have shown this indirect effect could enhance the radiation, while others show it might cancel it out.

The GDI study illustrates the danger of rushing to deploy new technology to solve complex problems like climate change without fully understanding the risks. “It shows that if you don’t think of both public health and your CO2 budget at the same time, you’re not going to make very optimal decisions,” Shindell said.

On the bright side, the study scenario is pretty unrealistic. With electric vehicles on the rise, a wholesale switch to GDI is unlikely. However, more than 50 percent of new vehicles are blasting black carbon into the environment, and those vehicles could be on the road for another few decades.

Saleh observed that the unforeseen effects of GDI engines point to the pitfalls of piecemeal climate solutions.

“Doing half-measures — not fundamentally changing our energy system, but doing half measures of improving an engine a little bit to achieve certain gains — in a way, hinders real movement toward clean energy and really addressing the climate and the public health issues at the same time,” he said.



Print Friendly, PDF & Email


  1. rd

    A few questions:

    1. They are raising the specter of an increase in PM2.5. Are air pollution studies showing an actual increase in PM2.5 in the air in cities?

    2. Car tailpipes are usually within a foot of the ground. Wind velocities are much lower near the ground than up in the air. The standard height for rural wind observations is 10m off the ground but urban areas are more chaotic. https://www.wmo.int/pages/prog/www/IMOP/publications/IOM-81/IOM-81-UrbanMetObs.pdf So how much dispersion will there be from particulates from car exhaust inches above the ground compared to a diesel truck with a vertical exhaust 10+ feet off the ground or an industrial smokestack 100+ feet tall?

    3. If the car fuel efficiency is much higher, is there actually much more PM2.5 emitted if less fuel is burned? Are other gaseous pollutants that can disperse much more easily reduced thereby producing a net benefit?

    But still, it is just another reason to accelerate the move to hybrids and electric vehicles.

  2. California Bob

    GDI also causes carbon buildup in the intake system and the backside of intake valves–which PFI prevents–mostly from PCV gases being recycled back into the system to be burned in the cylinders. This results in expensive maintenance to clean the intake system (sometimes, by blasting with walnut shell media). My 2019 Mustang GT’s ‘Coyote’ 5 liter engine uses both PFI and GDI, and the PFI can presumably prevent the buildup, but I don’t know if the engine is mapped so that the PFI can reduce the black carbon emissions. I certainly hope so.

    1. Carolinian

      Right. I recently bought a car to replace my ancient jalopy and made a point of choosing one that uses more traditional port injection (same as said jalopy) due to the by now well known problem of carbon buildup on valves. This buildup may take tens of thousands of miles to become a problem but still. Newer cars also use something Continuously Variable Transmission or CVT for their automatic transmission. This too has proven to have drawbacks compared to the traditional type but does get better gas mileage. The car I bought gets 38 highway–better than the one it replaced–so there are alternate methods to improve gas mileage.

      Perhaps car companies see electric as the future and therefore aren’t that concerned about troublesome bridge technologies. But there should be more attention focused on things like GDI and CVT

      1. Bill Carson

        I hate CVT’s. My Subaru has one, and it is awful. I haven’t checked to see if the engine is a GDI. I notice that only 2% of Toyotas have GDIs, and I think I’ll buy one of those next. Still, the model I want to buy has a CVT. Argh..

    2. cnchal

      > . . . which PFI prevents–mostly from PCV gases being recycled back into the system to be burned in the cylinders.

      PFI (Port Fuel Injection) injects the fuel into the rushing air filling a cylinder from behind the intake valve, continually washing the back of the intake valve in gasoline vapors, preventing carbon buildup.

      The workaround is to have both systems, a direct injection system and a vestigal port injection system to compensate for having direct injection. Moar complexity and things to go wrong for less than no benefit. The miniscule savings in fuel is overwhelmed by maintenance bills from that complexity.

    3. mbc92028

      Direct injection is hard on motor oil too. DI can also create fuel wash, where some of the fuel sprayed into the injection chamber hits the cylinder walls and sticks and some gets passed the piston rings and contaminates the oil with fuel, breaking the oil down more quickly.
      Direct injections biggest benefits are too turbo engines. Its big gains are from, delivering a more exact fuel air mixture and the spraying cools down the cylinder tops which allows higher compression and more boost. I don’t see much upside with DI on a non-turbo motor.

      1. California Bob

        Yeah. If I had my druthers I’d pick straight port injection, but I wanted the car and that’s the engine that came with it. Many owners complain about a ‘barbeque (or typewriter) click,’ which is likely due to the GDI high pressure pump. I only drive about 8K miles/year, and it will be my last new car (at least, the last ICE car).

  3. voislav

    The article is a bit misleading. Increase in particulate emissions is not due to GDI per se, it’s because car makers have been running engines leaner to get better fuel economy.

    To explain a bit further, mixing of air and fuel in the engine can be done in different ratios. Stoichiometric ratio means that air and fuel are mixed in proportions necessary to combust fuel into CO2 completely. Usually this is 14.7 parts of air to 1 of fuel. Using more fuel than stoichiometric, ratio of less than 14.7, is called rich mixture. Using more air, ratio more than 14.7, is called lean mixture.

    Running engine at sotichiometric ratio is optimum for emissions, but not for fuel economy. Lean-burn is optimum for fuel economy, but produces excess NOx emissions and forms particulate matter. Diesel engines run lean in normal operation and they require special NOx aftertreatment and Diesel Particulate Filters (DPF’s) to control particulate emissions.

    Over the past 10 years manufacturers have been pushing their gasoline engines more and more to the lean burn side in order to comply with the fuel economy standards. They know exactly what they are doing and they are ready for the incoming regulation on the particulate matter emissions. Right now gasoline engines are not regulated for particulate matter, but this will change in the next couple of years. Europe is ahead of the curve I believe, they might have regulation already in place, set to come into effect in either 2022 or 2024.

    Once the regulation comes into effect the manufacturers will add the Gasoline Particulate Filter (GPF) into the system. A lot of the recent testing we’ve done with different manufacturers included GPF in the exhaust system, so they are ready to launch it.

    1. Altandmain

      Working in the auto industry, I would agree with this.

      Manufacturers are under a lot of pressure to maximize fuel economy at all costs, so yes, they will try things like running lean.

      Some of the problems with GDI can be solved by dual injection (particularly the carbon build-up in engines), which is starting to become more popular. But yes, filters are needed to catch the particulate matter.

      The filters can be quite reliable at catching the particulate matter.


      Unfortunately there are trade-offs to getting fuel economy. The automakers can’t win this one. Either they get blamed for fuel economy not improving as fast as many would like or they get blamed for this one. At least with particulate matter, there is the possibility of filters catching the matter.

      1. California Bob

        Is the black carbon from gunk in the intake system, or a combustion byproduct anomaly from the inside of the cylinders?

      2. Bill Carson

        I understand that automakers are also using greater tolerances* in the engine, which reduces friction and improves fuel economy, but these engines are more prone to burn oil, and I don’t know how the oil consumption affects emissions. Is that your understanding? (My car uses so much oil now.)

        * Not sure if I am using the correct terminology.

        1. California Bob

          You’re thinking of clearances, essentially the size of the gaps between parts; tolerances is how much variation is allowed in the clearances; i.e. the clearance may be 0.005 inches, the tolerances a ‘thou’–thousandth or an inch–or two.

          Ford–and Honda and probably others–uses almost exclusively 5W-20 motor oil. This is a relatively ‘thin’ oil that reduces friction; and increases mileage, but not by a whole lot, and likely gets used up (‘burned’) in the engine a bit more than a thicker oil (say, 10W-30 that was prevalent a few years ago). I installed a catch can in the PCV return circuit of my Mustang, it accumulates about a teaspoon of oil–which would otherwise get deposited in the intake system–every thousand miles or so. Ironically–or, more appropriately moronically–this catch can is ‘not approved for sale in California,’ not because it creates any sort of problem, but likely because getting CARB’s approval is not worth the trouble to the manufacturer.

  4. Peter

    It seems to confirm what I have been arguing. Technology is not a solution for environmental problems. Every technology has trade offs that solves one problem but creates a different one, from wind energy to nuclear to e-mobility, and the alternate problems created might be even worse than the ones solved.
    You know what is really environmentally friendly? Oxcarts or horse drawn vehicles made from wood…..

    1. Dan

      My thoughts exactly. And it goes beyond just technology. What seems like “progress” in other fields has a devastating impact on how we fundamentally perceive ourselves:

      “According to primitivism, then, humans are nearing the beginning of the last phase of the domestication process as we are now experimenting with direct genetic engineering, and are making dramatic and frightening advances in the fields of psychology, anthropology, and sociology. This thereby allows us to quantify and objectify ourselves, until we too become commodities of no greater or lesser fundamental import than any other asset.

      -from the Wikipedia entry on Anarcho-primitivism, 1/20/2007

      1. Synoia

        This thereby allows us to quantify and objectify ourselves, until we too become commodities of no greater or lesser fundamental import than any other asset…

        Does the term “peasant” or “slave” ring a bell? Twas always thus.

    2. Arizona Slim

      Or walking.

      Yeah, I know. Putting one foot in front of the other.

      Case in point: Last Wednesday, I was walking to the bus stop. Needed those SunTran wheels to get me across town.

      While I was hoofin’ it to my date with SunTran, I passed by a couple of university students who were standing by the curb. Why they weren’t walking was beyond me. The morning was sunny and the university’s only about a mile away.

      But no. They were waiting for an Uber.

    3. Math is Your Friend

      On the contrary, oxcarts and horse drawn vehicles would almost certainly be less environmentally friendly.

      There are at least two basic problems – what we don’t know yet, and what we don’t think about in the context of a particular problem.

      The first is handled by learning, trying things out, not changing too much all at once, exploring alternate solutions, and not hurrying too much.

      The second is the endemic problem of trying to change something without thinking about collateral effects. That happens all the time. The answers are generally about finding and using the right technologies appropriately.

      Unfortunately too many people want to rush ahead with their pet idea, without really thinking it through. If you hurry too much, you end up with more problems on a larger scale by the time you have the experience to evaluate the initially favoured solution.

      For example, certain ozone friendly refrigerants are also very powerful greenhouse gasses.

      Diesel engines improve fuel efficiency and reduce CO2, but tend to produce oxides of nitrogen. Ooops.

      Hydro-electric reservoirs emit large amounts of methane for decades, boosting global warming.

      Different details, same basic errors.

      That is part of the reason I distrust ‘simple, obvious solutions’ to major problems.

    4. d

      Hate to tell you this, but they weren’t pollution free. There were lots of issues with horses. Among them was that when they died, they were just where died, no matter where that was. Even if there was say a meat packing plant near by. And nobody cared at all, since nobody had any idea that this was dangerous. I suppose if we just all went back to walking, gave up electricity, and other technology we have ever created, we too could be like cave people. We humans always seem to think things were better in the past.

      They weren’t

      1. witters

        “They weren’t” – Really? Send me your refutations of James. C. Scott’s recent “Against the Grain” and Marshall Sahlins’ classic “Stone-age Economics.” Exceptionalism – especially of the easy and obvious kind – should always raise flags.

        1. JBird4049

          Things get better and then worse in a constant cycle. And what is better or worse also changes. Hunter-gatherers had better diets, but often more violent lives than of city dwellers. Also, there was a growing environmental crisis just before the automobile was introduced. Growing cities required increasing movement of food, goods, and people, which required evermore horses, which eventually lead to feet of manure sometimes being on the streets. Yes, there were efforts to clean up, especially after the true nature of disease was known, but were just overwhelmed. This is a reason why the car and trucks became popular.

          But mountains of manure compared to air pollution. Hmm.

          1. Peter

            To remove horsemanure etc. is an organizational problem, and the bigger the pile of manure in front of your barn was in Germany a sign of wealth. It indicated the mount of animals you kept but also the potential you had to improve your soil.
            In Frankfurt i.e. it was the local population of gardeners who worked the fields along the banks of the Main River who did collect the horse manure.
            To somehow maintain that animal and human power and the construction of vehicles from wood is actually less environmentally friendly is somewhat – lets call it bullshit?

            Of course, it is all also connected to population growth, but the industrial mode of production that includes farming methods that at present destroy the health of soils and lead to the demise of large amounts of beneficial insects madea steep increase possible, although now the increase due to a more secure live and the better survival rate of progeny actually slow the increase down.

            1. Math is Your Friend

              And on certain Pacific islands, a big pile of yams was a sign of extreme wealth. That doesn’t make it true in all times and circumstances.

              Removing horse manure from a large city is not an easily solved organizational problem. it is a physics/math/technology/economic problem.

              If it is a four day trip with a horse and cart to a disposal site from the central region of the city, how many additional horses are needed to move the manure? How much additional manure do they produce? How many horses, carts, and teamsters are needed to move the original manure, and how do you raise, train, pay for them and feed them or build them (as appropriate)? How far do you have to go to get the raw materials, food, etc? What is the effect of this drain on resources and labour have on the rest of the society’s activities? What is the effect of manure cart traffic jams on things like bringing food into the city?

              The claim that a pile of manure in front of a barn is not a problem, so manure in a large city is not a problem is rather like concluding that since being in the middle of a bath tub is not a problem, being in the middle of the ocean is not a problem either.

  5. Grumpy Engineer

    This description could be confusing for people: “With a PFI engine, gasoline mixes with oxygen from the air, forming a vapor, before reaching the engine. In a GDI engine, the liquid fuel is injected directly.

    A better description would be as follows: “With a PFI engine, gasoline mixes with oxygen from the air in the intake manifold, forming a vapor, before reaching the cylinders and pistons. In a GDI engine, the liquid fuel is injected directly into the cylinders.

    As for the trade-off between efficiency and pollution, that’s been a long-standing problem in the automotive industry. During the early days of emissions controls, auto manufacturers would retard the spark timing to reduce pollution. But this reduced efficiency and performance. Some owners would re-advance the timing to get some performance back, and better gas mileage to boot. And diesel manufacturers have been battling the NOx vs. efficiency trade-off for as long as there have been NOx regulations.

    1. California Bob

      re: “A better description would be as follows: “With a PFI engine, gasoline mixes with oxygen from the air in the intake manifold, forming a vapor, before reaching the cylinders and pistons. In a GDI engine, the liquid fuel is injected directly into the cylinders.”

      Still not exactly correct. PFI engines squirt fuel into the intake ports, which are in the head(s), not the manifold. Hence Port Fuel Injection. There is a type of fuel injection that injects fuel into the throttle body before the manifold, called, surprisingly, ‘Throttle Body Fuel Injection,’ but it’s not much used anymore..

      1. Grumpy Engineer

        Aye. I focused on the “inside cylinder” vs “outside cylinder” aspects and forgot about the differences between TBFI and PFI. Thanks for the additional clarification.


    Minor tangential quibble, but it’s odd to see turbochargers listed as an emerging technology. Those have been used in commercial applications going back nearly a hundred years. F1 and Rally racing used it extensively in the 80’s as well as today, and the whole sport compact movement that started in the 90’s and served as the inspiration for the Fast and Furious movie series was built around turbocharged inline 4 and V6 engines.

    1. Synoia

      Catalytic converters make the reaction

      2CO + 02 -> 2 CO2

      happen at a lower temperature.

      C + 02 -> CO2 is a different reaction, and not assisted by catalytic converters.

  7. cnchal

    > However, more than 50 percent of new vehicles are blasting black carbon into the environment, and those vehicles could be on the road for another few decades.

    Doubtful. I can see it now. It needs tires, brakes, a few non critical warning lights are on, back up camera not working, and the service advisor mentions a $2000 walnut blasting jawb to get rid of the rough idle and gutless top end you are complaining about. How much are the payments on something new, and just as crappy?

  8. Bill Carson

    Funny, I haven’t seen anybody promoting that “amazing technology” that “Detroit doesn’t want you to know about,” and then show me how to install a system to inject water into the carburetor.

  9. Daniel F.

    These sould like those unexpected consequences Germany’s dealing with right now.
    According to this NBER paper, maintaining an acceptable baseload while moving from nuclear energy to renewables meant they had to burn more coal, which has an approx. 12 billion dollars per year social cost. Also, the increase in pollution resulted in excess mortality, some 1100 people per year.

    1. vlade

      The German nuclear move was idiotic populistic gesture.

      I can see how you want to move from nuclear – but you need to plan for it, and it’s not easy (because of the baseload you mention).

      TBH, Germany is nothing compared to Austria, which screams all the time about its neighbors nuclear power stations, when it imports 2/3rds of its electricity needs, a lot of it from the nuclear stations. “We don’t want you nuclear stations, but we want your baseload electricity!”

  10. jackiebaass

    I live through the whole process of making cars more fuel efficient. It wasn’t a smooth journey without bumps in the road. In the beginning people actually paid to have pollution devices removed. My personal opinion is this produced more efficient better automobiles. I remember driving cars from the 60’s. They are nice to look at but I wouldn’t want to go back to what they were. Cars today are more efficient, reliable, safer , and better built. They don’t rust out in couple years and if maintained will last a lot longer than the old cars did. What I’m saying is the faults will be worked out if given time. That’s why I believe rolling back standards is a big mistake and a bad idea. We don’t need to return to the days of polluted air and water like it was in the past for the sake of someone making a buck on the backs of peoples health.

    1. California Bob

      “We don’t need to return to the days of polluted air and water like it was in the past for the sake of someone making a buck on the backs of peoples health.”

      Tell it to the GOP.

      1. drumlin woodchuckles

        If California can win its court cases to keep its own Waiver to set its own automotive mileage and etc. standards, then in theory California would not have to tell it to the GOP. California could keep doing its own controls its own way.

        And Oregon and Washington could adopt the same standards if they want to.

        And Oregon and Washington, but esPECially California, are semi-sealed off from the states to their East by big powerful mountain ranges. At least California is. So California at least could raise gas taxes and use the money to pay for anti-gas transport improvements with very little fear of massive leakage of “runaway customers ” driving over the Sierra to buy their gas in Nevada and Arizona.

        Of course this would require conservation-minded California liberadicals to crush, smash and “exterminate” the Pelosi-style limousine-liberal virtue signallers from out of California politics and public life. It would also require the change of mindset needed to spend money on rational train travel from somewhere to somewhere . . . . to be paid for in part by using the money NOT spent on that symbolic High Speed Rail line planned to destroy some of California’s best farmland.

  11. Titus

    “When it comes to the climate, the question is, will the benefit of reducing CO2 outweigh the cost of increased radiation from black carbon emissions? On a global scale, the data says yes.” Ah, these models and the data I don’t recall nor can I find any results being put up for peer review. And the the conclusion seems off to me. I run models as well, but key properties of black carbon are missing. Given the amount of bogus science floating around and the severe consequences for getting things wrong (the new 2100 is now 2070), I can think of better ways to design an ICE engine then That what exists. Many engineers have. But it’s moot, ICE engines have no place beyond 2030.

  12. drumlin woodchuckles

    I have to rush back to work. So for the moment I will just say about black carbon locally causing more heatup locally . . . if the rest of the atmoshphere all around the locally heated black-carbon-heatup localities has less greenhouse gas in it, then the greater heat in the local black carbon localities will flow down the heat gradient into the rest of the atmosphere, from which it will then re-radiate back out into space better than it would have in a “higher” greenhouse gas situation. So the black carbon will not add ANYthing to the OVERall heat-retention picture at Earth’s surfacesphere.

Comments are closed.