Grilling: Gas vs. Charcoal

No, I am not writing this to further the debate of which is better (Clearly charcoal is). Sorry, this has to do with the general population not understanding how to use basic statistics, and going ahead anyway to use them to try to prove their point.

At work, our monthly safety bulletin was about grilling. Rather appropriate for the summer. The first error made was in the title "The Backyard BBQ." BBQ is a style of cooking and using an outdoor grill does not make it BBQ. It's called grilling - and can be just as tasty as BBQ I might add. But I digress, the title had nothing to do with the statistics. The following phrase however did:

"81% of the grills were fueled by propane/gas and 16% used charcoal or other solid fuel. Although gas grills are only used roughly 1.5 times as often as charcoal grills, they were involved in five times as many fires."

The emphasis was in the handout, I didn't add it. When this was read, I did a double take and had to look at the sheet to see if it really said that. It did. The point they are trying to get across is that even though gas grills are more popular, they are the cause of a disporportionate number of the fires. So let's look at the statements.

First, 81% of grills are fueled by propane/gas and 16% used charcoal or other solid fuel. I am assuming that the remaining 3% or grills are liquid fuels (although, I have never actually seen one of those, unless you consider a Coleman stove a grill). I have no objection to this, it makes perfect sense. Go down to the hardware store and the vast majority of grills are gas.

Next, gas grills are used 1.5 times as often as charcoal grills. OK, this can be taken one of two ways. A) (an absolute view) For every 2 uses of any charcoal grill, any gas grill is used 3 times or B) (a relative view) For every two times the average charcoal grill owner grills, the average gas griller grills three times. They may seem the same, and if there were equal numbers of gas and charcoal grills then they would be. But there are not.

So, let us do a little math. Assume for a moment there are 81 gas grills in the US and 16 charcoal grills in the US. We know how many grills there are so we want to calculate total uses. Using the absolute view (A) above, if every charcoal grill is used twice in one year, they account for 32 total uses. This would mean the gas grills account for 48 total uses (1.5 time 32). In other words, at least 33 of the gas grills were never fired up even once. Another way of looking at it, is that 16% of the grills (charcoal) accounted for 40% (32) of the total uses (32+48=80). Does that make sense?

No, it is like saying that people with charcoal grills are 3.8 times ((32/16)/(48/81)=3.78) as likely to grill as people with gas grills. Having owned both gas grills and charcoal grills, I know this can't be true. While I love the taste of charcoal flavored burgers, nothing beats the ease and convenience of turning the knob, pushing the button, and pop you have your flame. I grill more often with gas than I do with charcoal, because of the ease of use.

So let us look at the relative case (B) to see if it makes sense. Since each of these is an average griller, the percent ownership is irrelavant. So it is simply stating that if you own a gas grill you are 1.5 times as likely as a charcoal griller to use your grill. This is the point of what is being said. I hope you are still following along.

Now the last part, gas grills are responsible for five times the number of fires. This is said as the ominous warning. But is it really? Let us look at our first set of numbers, 81% vs 16%. Turns out there are approximately 5 times as many gas grills as charcoal grills. This statement can also be taken in an absolute or relative manner so let us examine both for the common sense test.

Absolute. Of the total number of fires, five times as many are caused by the gas grills. We can assume that a fire isn't going to happen unless the grill is used, so each use should count as a potential. We already know from above that 81 grills x 3 uses + 16 grills x 2 uses = 275 total uses. If we assume that 13.3% result in fires (no where near this high in real life but this is just for comparison purposes) then the 275 uses will result in 36 fires. Of that 36, 6 were from charcoal grills and 30 (5*6) were from gas grills. Or a rate of 0.1875 per use for charcoal and .1234 per use for gas. This would indicate that while gas is 5x more prevalent than charcoal, it is actually 51% less likely (.1875/.1234-1=0.51) to cause a fire than charcoal. Hence, no need for the ominous warning.

Relative. If 3% of charcoal uses cause a fire, then 15% (3%*5) of gas uses cause a fire. So, for 32 uses of charcoal grills, there would be 1 fire. For the 243 uses of gas grills there would be 36 fires. Now does this make sense, no not really. There isn't a mechanism that would make gas grills more fire prone than charcoal grills. Granted, if you leave the gas on for a while before lighting it, you could have a nice fireball. However, this pales in comparison to the safety hazard of squirting lighter fluid on your already lit coals. I'll admit, I like to do this! So the relative case to me seems like bunk.

Turns out, the NFPA has these statistics on their website. Of the 7900 fires, 6400 are from gas grills (so I assume that 1500 are from charcoal grills). This confirms my analysis that the relative case is not correct. However, there are still some problems with these numbers. 6400/1500 = 4.3 not 5 times. So this makes the absolute case even more peachy. Even though 4.3 times as many fires occur from gas grills as charcoal grills, since gas grills are 5 times as prevalent as charcoal grills and used 1.5 times as often, they are actually 57% (4.3/5/1.5=0.57) less likely to cause a fire (hence they are much safer).

So really, what the NFPA (and my work) should be saying is that charcoal grills are fire magnets! And without changing any other practice (like not grilling indoors), if we switched the 16% of charcoal grills to gas grills, we would decrease the number of fires by 11% (57%*1500/7900=0.108).

Now, if this doesn't bore you, I don't know what will. I find it rather exciting!