Arthur E. Sowers is a resident of Harbeson.
Snowstorm Hernando messed up a lot of people recently. My wife and I lost power, starting Feb. 22, and it was not restored until two nights later. For heat, we used a single, unvented, portable space heater with propane. It uses small camp lantern propane bottles that need changing every six to eight hours. This held our room to about 55-60 degrees. Layers of blankets made life tolerable. I had a stock of bottles that outlasted the outage. In the public’s interest, I share the most important part of my experience — a carbon monoxide symptom — and what I learned from it. This may help other people with planning and to understand more about the safety and health aspects of survival during a power outage in the winter.
Whenever carbon burns by oxidation, the product is mostly carbon dioxide, but a minor part is dangerous carbon monoxide, and the main thing to know is that it can kill you. And propane-powered devices often have conspicuous warnings about this fact printed right on the devices, as well as in the instruction manuals.
The predominant carbon monoxide-poisoning symptoms mentioned in most articles are nausea, headaches and sleepiness. After the first day with the space heater running, however, I noticed that my hearing was noticeably weaker (as if you turned down the volume on a radio). Even a day after power was restored, voices of some actors in a movie we watched were distorted in a weird way. Otherwise, I never felt any headache, nausea or sleepiness. My wife reported no symptoms at all at any time. By the day after power was restored, and we were off the propane, my hearing started to return to what it was before the outage. But it was still not fully normal until after a second day. I searched the internet both during and after the power outage to better understand what happened.
An internet search on the keyword string “reversible hearing loss carbon monoxide” brought me to many scientific articles documenting both reversible and irreversible hearing loss, plus other reversible and irreversible symptoms and medical conditions, from exposure to carbon monoxide. It turns out that there are many other and less common symptoms of such poisoning. Most studies are carried out on animals because, since 1948, medical ethics forbids testing lethal substances on humans. Thus, it is always a valid question whether animal data can apply to humans. The few studies on humans are derived from, for example, various industrial accidents, where cause variables were not fully known or controlled. This is why there is no solid data for humans. Standards are based on committees taking votes to estimate safe numbers. I found one standard that said 9 parts per million was the carbon monoxide limit for 24/7 exposure. Another said 200 ppm will cause symptoms in “hours.” The Occupational Safety and Health Administration says 50 ppm. Another said death in minutes from 1,200 ppm. This is a large range in concentrations (and exposure times). Add to that the fact that biological variability means that, under some circumstances, one person might show symptoms, but another person might not. How do you deal with all this?
My story is as follows. The internet gave me this data: It is estimated that 5 million to 9 million people in this country use unvented CO-generating space heaters. One source said some 400 people died, annually, from poisoning by the substance. It did not seem to be a big problem. But, days later, I did a different internet search based on hospital emergency room visits for nonfire carbon monoxide poisoning, and, annually, it was about 100,000. That indicates that a lot of people are getting symptoms by exposure to too much CO but survive. It could be concluded that many people are less aware about this poisoning than they should be.
Also, my “brand X” digital carbon monoxide meter, just when I needed it, was not working, and when I checked the fine print in the instructions, it said the life of the sensor is three years, and I bought the thing many years before three years ago. When I checked on the internet with the keyword string, “complaints brand X,” I learned that the company was selling a rip-off product. Some users reported that the sensor lasted much less than three years.
At the beginning of the outage, calculations based on my room size, propane consumption and nominal assumptions — carried out by two free online artificial intelligences — showed that my wife and I were pretty close to OK to use our unvented propane space heater. Also, I recalled that, when I lived in Texas, a lot of older homes had unvented gas space heaters (even built permanently into walls), and nobody ever worried or talked about carbon monoxide. So, after that, we went ahead with continuous propane heating with some confidence.
The key factor — I more fully realized later — is that older homes are more “air leaky,” and fresh, cold and CO-free air is always flowing into the house to constantly dilute the warm, CO-loaded air. This was even mentioned in some articles I found during my more extensive and deeper internet searches after the storm.
I think my noticeable and reversible hearing loss was plausibly due to some level of carbon monoxide that built up in the house. My greater sensitivity to it, compared to my wife’s lesser sensitivity, might be related to my obvious and recent slow hearing loss and the presence of my lifelong tinnitus. My wife’s hearing was always normal throughout the outage. Other articles indicated that other and simultaneous hearing impairment conditions will cause an increase in sensitivity to CO poisoning.
In the future, I will arrange an increase in fresh-air flow from outside to dilute inside carbon monoxide concentrations. Refreshing the air should keep the substance down, while still keeping us warm most of the time. Lastly, I will get two new, but cheaper, digital monitoring devices, from different manufacturers. Why two? A deep dive with an artificial intelligence found high-quality industrial specifications and test results; many CO detectors/alarms are inaccurate and exhibit age-induced sensor drift. Even units of the same model can be off by 30%, and some new units were found to even fail before the end of the warranty. If you have two three-decimal-place digital meters, and they both give approximately (e.g., within 30% of) the same number, that is good. If you get numbers way off — by, say, 50% or more — then something is wrong, and it will be your call to decide what to do.
You can do an internet search on “best CO monitoring devices.” Also, do a search on “complaints brand Z.” Try to get downloadable copies of the instructions. Then, decide which to buy. I would urge digital devices, so you can continually monitor buildup, rather than ones that sound an alarm at some threshold.
Reader reactions, pro or con, are welcomed at civiltalk@iniusa.org.
