In 1975, the auto industry began to equip vehicles with catalytic converters to meet the emission limits of the Clean Air Act of 1970. Sitting unobtrusively between the engine and the muffler, the “cat” changes the noxious gases in automobile exhaust into harmless nitrogen, oxygen, carbon dioxide, and water. The result, according to the National Institutes of Occupational Health, was an 80 percent decline in the number of unintentional vehicle-related deaths caused by the most dangerous byproduct of combustion engines: carbon monoxide. 

But catalytic converters don’t entirely eliminate the CO, so their advent did not eliminate motor vehicle CO poisonings. This lingering safety issue resurfaced in 2009, when keyless ignition systems severed drivers from their relationship with traditional metal keys, making it easy to inadvertently leave their vehicle engines quietly running while CO levels built inside the garage – and then the house. Motor vehicle carbon monoxide poisonings became newsworthy again in 2017, when numerous police departments with fleets of newish Ford Explorer Interceptors (a police version of the Explorer) began to complain that their officers were being sickened in their squad cars. Just last week, WRC, NBC’s Washington D.C. affiliate, broadcast a story showing that the problem affects civilian Explorers, too. The story featured two owners who suffered from the symptoms of carbon monoxide poisoning.  

The truth is: carbon monoxide continues to kill hundreds and sicken thousands of people in the U.S. each year. 

Carbon monoxide is a colorless and odorless toxic gas produced as a by-product of incomplete combustion of carbon-based substances.  It is the most common lethal poison worldwide. When inhaled, CO is absorbed from the lungs into the bloodstream. CO binds with hemoglobin with an affinity of more than 200 times that of oxygen, forming a tight complex with hemoglobin, which impairs the oxygen-carrying capacity of blood. 

How many CO injuries and deaths are caused by motor vehicle exhaust is up for debate, because the data are hard to come by. In 2000, NHTSA’s National Center for Statistics and Analysis updated a 1996 study estimating the number of deaths caused by carbon monoxide poisoning by motor vehicle exhaust. It found that from 1995 to 1997, the total number of fatalities due to CO poisoning was 3,715, with the majority of those fatalities involving stationary motor vehicles; 665 of those fatalities were accidental. A 2002 analysis of U.S. motor vehicle CO deaths from 1968 to 1998 estimated that annual accidental CO deaths from all sources decreased from 1,417 in 1968 to 491 in 1998, with 48 percent of the 1998 deaths attributed to motor vehicles. 

Establishing a link between CO deaths and motor vehicle exhaust using the Centers for Disease Control’s WONDER database, an open access online search system using mortality data drawn from all filed and coded U.S. death certificates, has been stymied by a change in coding practices. In 1999, the International Classification of Diseases coding system dropped the identification of CO sources. 

Nonetheless, a 2004 NHTSA study on non-traffic, non-crash motor vehicle fatalities found “somewhere between 200 and 250 deaths a year that are not known to be suicides result from vehicle-generated carbon monoxide. These types of deaths occur more frequently than deaths from any of the other issues researched.” 

A more recent 2011 study of data gathered from news stories about residential CO poisoning in the United States published between March 2007 and September 2009 found 837 reported CO poisoning incidents, 59 of which were the result of a vehicle left running in the garage. The author, Dr. Neil Hampson, who has published extensively on the epidemiology of CO poisoning,  concluded that household CO poisoning from a motor vehicle left running in the garage “is relatively common.” 

The annual number of CO injuries in a much harder figure to ascertain. In 2011 the CDC, using data from the National Poison Data System (NPDS) to characterize reported unintentional, non- fire-related CO exposures, found 68,316 CO exposures reported to poison centers during 2000-2009. 

Translating exposures to injuries is tricky. The initial symptoms of CO poisoning are nonspecific and vary widely – anything from headaches to fatigue to dizziness and vomiting. Severe exposure can lead to confusion, loss of consciousness, seizure and cognitive difficulties.  Almost half of the people exposed to CO develop delayed neuropsychological responses that can be disabling and sometimes permanent. These effects can develop days or weeks after exposure. 

Petitioning for a Solution

Squishy data hasn’t stopped citizen advocates from asking the National Highway Traffic Safety Administration to require countermeasures to either warn occupants of the presence of CO in the occupant compartment, shut off the engine when it reaches pre-determined levels, or both. Since 1997, there have been three petitions for rulemaking.

In March 1997, Herb Denenberg, an attorney, former professor, journalist and consumer advocate, petitioned the agency to require carbon monoxide detectors to be installed in all motor vehicles, to mandate manufacturers to offer them as optional equipment and to alert consumers in owner’s manuals to the availability and value of installing a carbon monoxide detector.  Denenberg also requested that NHTSA issue press releases and consumer advisories regarding the availability of carbon monoxide detectors. Denenberg cited a December 1993 NHTSA Consumer Advisory showing 353 fatalities that year from accidental carbon monoxide poisoning. His petition asserted that these were preventable deaths, and that manufacturers could install CO detectors in vehicles for about $16 per unit.

Eight years later, toxicologist Albert Donnay, representing a non-profit that  focuses on Multiple Chemical Sensitivity disorders, filed a petition for rulemaking making similar requests: press releases on the dangers of vehicle carbon monoxide poisoning, more research on all CO vehicle-related fatalities, a requirement that manufacturers warn occupants about the dangers of carbon monoxide poisoning in owner’s manuals and install CO detectors in new vehicles, with the capability to cut-off the engine when carbon monoxide levels inside a stationary vehicle exceed a concentration of 200 parts per million.

In his petition, Donnay argued that there was ample evidence that the agency considered motor-vehicle CO poisoning a serious problem, but had abandoned the issue, failing to continue its research, or do more consumer education on the issue, or initiate rulemaking – even though the agency had influenced recalls related to CO exposure and had promulgated rules to address serious, but occasional problems, such as trunk entrapment.

The most recent petition was filed in March 2016 by Public Employees for Environmental Responsibility. Like its predecessors, PEER requested that the agency issue annual consumer advisories and recommend the use of onboard digital CO monitors; to track and report all CO-related fatalities; require manufactures to include information in new vehicle owner’s manuals about the health dangers of CO, and require all manufacturers to install CO detectors in the passenger compartment of all new motor vehicles; and require manufacturers to install engine shut-off technology. 

PEER echoed Donnay’s point about NHTSA’s failure to prevent continuing CO deaths:

Since NHTSA was first informed of the life-saving potential of CO detectors linked to engine cut-off switches, in excess of 20,000 North Americans have died needlessly from vehicular CO poisoning. At the same time, these CO detectors are reliable and very inexpensive. They are far less expensive than other measures that NHTSA has approved. In short, requiring these devices may one of the most significant and cost-effective vehicle safety measures since the seat belt.

NHTSA Wants Nothing to Do With It

The agency’s response to motor-vehicle CO poisoning issue has been, to put it charitably, contradictory. 

(Remember – the U.S. Environmental Protection Agency was the prime mover behind the widespread implementation of catalytic converters – not NHTSA. Prior to the addition of catalytic converters, “CO in motor-vehicle exhaust accounted for the most poisoning deaths in the United States caused by a single agent,” according to a 1996 CDC Morbidity and Mortality Weekly Report (MMWR). Out of 11,547 unintentional CO deaths during 1979-1988, 57 percent were caused by motor-vehicle exhaust; of these, 83 percent were associated with stationary vehicles.” Most of the deaths in garages occurred with the garage doors or windows open.) 

NHTSA has long regarded an unattended vehicle with the keys left in the ignition to be a safety hazard. In 1967, the impetus was auto theft, leading to police chases that often ended in fatal crashes. The agency’s first proposal for Federal Motor Vehicle Safety Standard 114 would have required cars to be equipped with devices to remind drivers to remove keys when leaving their vehicle. And the agency made it pretty clear that the solution should be based in vehicle engineering: 

It is, of course, the operator's responsibility to remove the key when the car is left unattended, and drivers should continue to be exhorted to take this elementary precaution. Nevertheless, many do not, and the interest of safety would be promoted by the existence of a visible or audible warning device on the car, reminding the driver when he has neglected his responsibility. This is an instance in which engineering of vehicles is more likely to have an immediate beneficial impact than a long-range process of mass education.

But, NHTSA has resisted regulating in any way an unattended vehicle with the keys left in the ignition and the engine running.

In the early 1990s, the agency appeared to have some interest, contracting the Carnegie Mellon Research Institute to evaluate its metal oxide semiconductor gas sensor technology for application as a low-cost carbon monoxide monitor in the automobile compartment. The study cited 500 accidental deaths each year, and, like the petitions that would follow it, observed that many “might have been prevented if the automobile passenger compartment were equipped with an appropriate CO monitor and alarm system.” The report concluded that the technology had was at the point “where a stable selective CO monitor is within reach,” and recommended that NHTSA undertake further research. 

But, The Safety Record couldn’t find evidence that NHTSA took this research any further. Instead, it knocked down most of the 1997 Denenberg petition, but grudgingly allowed that it might add some information about carbon monoxide poisonings to future advisories. NHTSA dismissed CO poisoning incidents as a cold weather phenomenon, most likely caused by people running their engines in an enclosed space to keep warm or failing to clear snow from the tail pipe area. “For this reason,” the agency wrote, “we do not think it is justifiable to require that all vehicles be equipped with these detectors. A large portion of the vehicles sold in this country will rarely, if ever, be driven in cold weather.” Besides, the agency said, mandatory installation of carbon monoxide detectors industry wide would cost at least $240 million, eventually consumers would be forced to replace them after six years. And, since the problem really affected the older models in the fleet, a regulation would not benefit the vehicles that needed it most.

In 2005, it denied the Donnay petition, claiming the data showed that the number of CO incidents was falling absent any regulation and because a mandate for in-vehicle carbon monoxide detectors would fail to address more than 70 percent of vehicle-related carbon monoxide deaths, because the victims are outside the vehicle. NHTSA argued that “a home CO detector would be substantially more effective than a vehicle CO detector at preventing these deaths because 92% of the fatalities occurred at the home.” The agency rejected the idea of an engine shut-off that would prevent CO accumulating to dangerous levels because it “could prove to be a hazard. For example, in a tunnel with congested traffic, the concentration of CO may cause the device to shut off the engine, resulting in further traffic congestion or even possible crashes.”

The agency hasn’t yet articulated a position on the PEER petition.

In 2011, for the first time the agency considered the unattended key-in-vehicle scenarios with the engine running, and the problem of CO poisoning. A FMVSS 114 Notice of Proposed Rulemaking – as yet unfinalized – attempted to deal with the proliferation of keyless ignition systems. The NPRM recognized that current keyless ignition systems had led to driver confusion, and that these designs allowed drivers to exit the vehicle without the transmission locked in Park, and sometimes without actually turning off the engine. The NPRM noted that the lack of standardization in combination with the lack of visual and tactile cues about the status of the vehicle engine has set the stage for the real world incidents in which drivers, mistaking the fob for the key, inadvertently leave a vehicle running and/or exit the vehicle without putting the transmission into “Park.” The proposal specifically deleted the door opening alert exclusion currently in FMVSS No. 114 for a running vehicle, but only for vehicles equipped with keyless ignition. The agency’s strategy for addressing the rollaway and carbon monoxide safety issues are for internal and external audible alerts, based on the Platform Lift standard, which is part of FMVSS 403.  

As part of the rulemaking, NHTSA said that it had considered “a requirement for an automatic shut-off feature applied to vehicles fitted with electronic key code systems.” But, it declined to propose such a feature because: “We have been unable to conclude that there is a specified period of time after which the propulsion system should be shut down to effectively address various scenarios mentioned in VOQs submitted to the agency.”

What, no mention of the simultaneous-shut-offs-in-a-tunnel scenario?

Motor Vehicle CO as a Function of Vehicle Design and Wear

In 1972, research sponsored by the Insurance Institute on Highway Safety noted that the available data suggested “that over 500 Americans die each year from carbon monoxide poisoning because their vehicles are defective due to deterioration, damage, or poor automotive design.”

That last bit is still the case, and the Ford Explorers currently sickening lots of people are good examples. In July 2016, NHTSA’s Office of Defects Investigation opened a probe into reports of occupants smelling exhaust odors in the occupant compartments of 2011-2015 Explorers. By the time it bumped the probe up to an Engineering Analysis, a year later, Ford had tallied 2,051 complaints, while NHTSA’s Vehicle Owner’s Questionnaire hotline received 791. The high-profile victims were police departments across the country, which were reporting that at least five officers lost consciousness, were hospitalized for CO exposure or crashed their SUVs, poisoned by the cabin air of their Interceptors.  You can read more about it here. 

Ford, which does a booming business with the law enforcement market, rushed to investigate the CO problem in those vehicles, and promised to cover the costs of any modifications. The automaker blamed the problem on unsealed spaces and wiring holes drilled in the course of implementing after-market features specific to police work, such as emergency lights and radios, and said that none of those problems affect civilian Explorers.

So why are so many Explorer owners complaining? In its August 2016 response to NHTSA’s Information Request, Ford argued that really not that many people complained. An analysis showed that “approximately 0.29 percent of all 2011 through 2015 model year Explorer owners have complained of some sort of exhaust odor in their vehicle,” and even if Ford included all of the ambiguous reports the percentage is still a piddling 0.38 percent of all vehicle owners. 

As for the cause, Ford isolated it to the unique combination of driving at “wide open throttle (WOT) events with the vehicles climate control system in the recirculation mode.” Ford said that “the fuel enrichment used for the exhaust catalyst protection strategy commonly used during wide open throttle events caused a more detectable odor being emitted from the tailpipe. Second, a negative cabin pressure was created from the vehicle climate control system being in recirculation mode. Ford notes that the vehicle drive cycle necessary to reproduce this condition is beyond what Ford would consider normal or typical customer usage.” In any event, the CO readings never topped 8 parts per million (ppm), and dissipated within 10 seconds. 

This does not seem to comport with real-world incidents. In preparing the WRC story, reporter Susan Hogan teamed up with toxicologist Albert Donnay, who measured the presence of CO in the occupant compartment of an Explorer whose owner had been complaining to Ford for months. Donnay gathered readings of CO with sophisticated detectors in the front and back seat. At low speeds, with the vehicle in re-circulation mode, there was no change in air quality, the report said. When the vehicle speed climbed to over 40 miles per hour, the CO level was 9 parts per million (ppm) in the front and 30 ppm in the back. You can read it here. 

Further, Hogan says that Donnay’s testing showed that the CO levels evened out at 15 ppm-17 ppm and stayed there for a good 10-15 minutes. The levels didn’t drop until they brought fresh air into the cabin.  

Another good example is General Motors’ 2015 “emission recall involving certain 2008 Chevrolet Avalanche, Silverado, Suburban, and Tahoe; and GMC Sierra, Yukon, and Yukon XL vehicles equipped with California Bin 4 emissions RPO NU5.” GM said that “the design of the fuel control system did not adequately control carbon monoxide emissions under certain operating conditions.” The remedy was a reflashing of the engine control module with a modified fuel control calibration.

Combatting Motor Vehicle Carbon Monoxide at the Source

There are several design solutions to the motor vehicle carbon monoxide problem, and for nearly forty years, individual inventors, suppliers such as Lear Corporation, and automakers have put forward ideas or implemented them. There are patents, dating back to at least as early as 1974, related to carbon monoxide detectors in motor vehicles that either warn the driver when the in-board air reaches a certain threshold and/or shut off the engine. 

Automakers have been using engine cut-off designs as a safeguard to remote start features for at least a decade. Two automakers have also added extended engine idle shut-offs to their keyless vehicles to prevent CO poisonings when drivers mistakenly leave their vehicles running. In 2013 Ford became the first automaker to add an engine idle cut-off to its keyless vehicles. This feature shuts off the engine after 30 minutes if there have been no inputs from the driver.

In March 2015, General Motors remedied 50,249 2011-2013 Chevrolet Volt vehicles with a software fix that would shut off the engine after an hour and a half, to prevent drivers from inadvertently leaving the vehicle running. GM was reportedly prompted to implement the recall after two injuries and push from NHTSA. The 2014 and beyond model years already have this feature. 

Another design includes the use of air classification or control module (ACM). This technology combines CO and nitrogen dioxide sensors to protect the cabin air.  Some automakers offer ACMs on vehicles with electronic climate control. While ACMs alone do not solve the problem of accumulating CO emanating from a running vehicle, they can be paired with auto shutoff mechanisms or alarms.  In 2007, AppliedSensor, chemical sensor components and modules manufacturer, and Sensata Technologies, a developer of industrial sensors and control solutions, announced that their Air Classification Module was being integrated into the 2007 BMW X5 Sport Activity Vehicle: “This ACM prevents the intake of harmful combustion fumes – such as carbon monoxide, nitrogen dioxide and volatile organic compounds: “The module includes a built-in sensor with two separate sensing elements to enable continuous detection of the presence of diesel and gasoline exhaust fumes faster and more reliably. Through integration of a corresponding module in the air intake duct of the vehicle’s heating, ventilation and air conditioning system (HVAC), the sensor can signal accurately timed, automatic activation of air circulation to the vehicle's controls.” 

More than 40 years after the advent of the catalytic converter, we still have a CO problem that endangers public health – and we have a menu of technologies that might all but eliminate the threat. The will to fix it seems to be the last obstacle.