The Fire That Changed Everything - And What is Being Done

On Dec. 17, 2005, my department in Barre City, VT, responded to the worst fatal fire in our history. At a few minutes before 6 a.m., we were dispatched to a fire in a two-story wood-frame apartment building. The initial report was that there were people...


My Chief, Peter John, attended a conference in a neighboring town presented by Deputy Chief Joe Fleming of the Boston Fire Department. During this conference, Chief Fleming discussed the differences between ionization and photoelectric smoke detector technology, among other issues related to this topic. When Chief John returned from this conference, he handed me a large stack of information he picked up, and told me the detectors in the apartment didn't activate.

The Fish Tank Test
I had never heard of such a thing before, and spent most of that night researching on the Internet to find out what I could. It didn't take me long to come upon some disturbing information. The most interesting site was from a group out of Australia. Their site, www.theaquariumtest.com, shows a fish tank full of smoke and an ionization smoke detector inside that didn't make a sound. When a photoelectric smoke alarm was introduced to the smoke, it sounded almost immediately. That was to hard to believe. So we decided to try it ourselves.

Two days later, we had our own fish tank and created our own smoke outside the firehouse and we video taped it (see videos below). What happened was almost beyond words. We created thick, yellow/brown smoke that hurt to breath and placed an ionization smoke detector in it. It did not activate. Yet a photoelectric smoke detector activated after five seconds when exposed to the same smoke. The ionization smoke alarm did finally activate, but only after it was almost completely obscured by the smoke.

All of a sudden, the idea that a smoke detector was just a smoke detector was as foolish as saying a fire is just a fire. They are not the same. They are very different. And knowing the differences can save your life. Since this first test, we have conducted several tests in an abandoned house, testing both ionization and photoelectric smoke alarms in different types of smoke.

Ionization Smoke Alarms
An ionization smoke alarm has two plates that have a small voltage going across them. Added into this mix is a small amount of radioactive material called AMERICIUM-241. This material ionizes the oxygen and nitrogen in the air. Basically what this means is that an electron is taken off from each atom. These electrons are now considered free and have a negative charge to them. The atoms, on the other hand, now have a positive charge. The electrons, with the negative charge, are attracted to the positively charged plate. The atoms, with the positive charge, are attracted to the negatively charged plate. The amount of electrical current traveling between the two plates is monitored by the detector.

As small particles enter the chamber, they interrupt the electrical current between the plates causing the electrical current to drop. This is what causes the horn to activate.

Photoelectric Smoke Alarms
A photoelectric smoke alarm works all together different. A beam of light is projected inside the chamber so that it does not come in contact with a sensor called a photo-detector. When smoke enters the chamber, it scatters the light so that it reaches the sensor. The light reaching the sensor is what causes the horn to activate.

Just like a smoke detector is not just a smoke detector, smoke is not just smoke. Different stages of fire create different types of smoke. A flaming fire creates smoke particles that are 0.01 - 3 microns in size. Smoldering fires create smoke particles that are 0.3 - 10 microns in size. Smoke from a smoldering fire is referred to as "Cold Smoke". A flaming fire produces a large volume of smoke with particles that are 0.01 - 3 microns in size. These small particles fill an ionization chamber with sufficient quantity to disrupt the current between the metal plates. This disruption then causes the horn to activate. These small particles will also scatter the light in a photoelectric smoke alarm, causing this horn to activate as well. Generally, however, the photoelectric is slightly slower to activate. The difference in activation time during a flaming fire is measured in seconds.

A smoldering fire produces a large volume of smoke with particles that are 0.3 - 10 microns in size. These larger particles still enter the ionization chamber. However, because of the larger size, not enough can fit into the chamber to disrupt the current. The current is able to, essentially, filter between the particles without enough of a disruption to sound the horn. Because of this, the ionization alarm can be in an environment full of this cold smoke and yet it may not activate. The photoelectric alarm is much better at detecting this smoke. As the smoke enters the alarm, the smoke scatters the light causing the alarm to activate. In almost every test we have done, the photoelectric alarms have sounded before we could see any smoke in the room.