That was the reading on the handheld carbon monoxide detector brought along on a follow-up safety consultation visit to the service department of a large automobile dealership. The conditions were typical for the the early part of their service day – several vehicles running at once inside their 20,000 square foot service area, with one of the six large overhead doors fully open to the outside and the others closed. A discussion with the service manager revealed that this was typical in the morning, when cars would be brought in to be checked out and run briefly for diagnosis.
Exposure With Consequences
The spot reading of 400ppm of CO does not tell the whole story related to exposure in the garage. Just a few minutes before, the shop was closed and the reading was effectively zero. And a few minutes later the reading began dropping from the peak. But the 400ppm number was definitely cause for concern. That level is enough to approach short-term exposure issues (see the CDC data on that risk) with exposed workers, and if it is sustained for long it could cause serious issues.
Why Was This Occurring
Even though the facility was equipped with good means of ventilation and a dedicated exhaust removal system, neither of these controls were helping much. The early spring moderate temperatures meant that the ventilation system wasn’t running in the morning, and not one of the running vehicles were hooked up to the exhaust removal system. When asked about this, one of the mechanics said that they will use the system if running a car for a while. When he and others were queried further as to just how long that meant, the answer was far from definite. “More than five minutes.” “Any extended time.” “Ten minutes or more…” It was clear that this wasn’t well defined, and also was not based on any actual determination of what level was an issue. It was much more about convenience, an informal determination on each technician’s part about if it was worth the trouble of hooking a car or truck up.
Learning from Other Environments
Some of the work I’ve done recently that centers on parking lot and parking garage safety has highlighted the usefulness of carbon monoxide monitoring systems. Though I was focusing on pedestrian safety and vehicle-into-building crashes, many elements of parking safety were looked at to some degree. It became clear that the issue of carbon monoxide in parking structures had been given a lot of attention and had subsequently become the focus of regulatory activity. This set the stage for some great technological options for reduction, monitoring, and exhaust of carbon monoxide that could be applied. In the case of the auto dealership service department, it was monitoring that was a glaring omission from their arrangement.
A survey of some other auto repair environments showed that some were employing consumer/household style carbon monoxide detectors. The rise of UL standard 2034 and its inclusion in many building codes has led to good availability of simple consumer style detectors. Many more homes are protected today than just a few years ago, but the problem with using a UL2034 detector in a commercial setting such as a garage is that the standard for these residential detectors places a great deal of importance on avoiding false alarms, which means that these detectors are not nearly as sensitive as most commercial models. The right solution for a garage environment typically requires some analysis by an industrial hygienist, and would include an appropriate advanced detector along with both alerts and possibly automatic activation of ventilation.
What About Procedures?
Note that the engineered solution is the right place to begin controlling this hazard. The service manager’s suggestion that they begin with a procedural response, including a shorter timeframe to attach the exhaust removal system, and more overhead doors open was not a sufficient solution. The way the shop operated, the longstanding practices employed, and the difficulty in determining action triggers in the existing environment all pointed toward not accepting an administrative-only solution. Naturally, every situation is different, and this case does not indicate that all similar operations should employ the same controls, but it does show that the right factors need to be considered, and any controls employed should have a good expectation of actually being effective!