The NTSB’s preliminary report on the San Bruno Fire has led some politicians to ask why it took PG&E so long to turn the gas off.

Jackie Speier. . . expressed bewilderment that PG&E failed to send anyone to turn off the ruptured pipe until "33 minutes after 100-foot-high flames . . .were clearly visible from (Highway) 101, more than 10 miles away.

Not a bad question.  But that’s not really the issue on which we should focus. Just for the moment, forget about where PG&E was after the fire erupted. Where was PG&E before?San Bruno Fire and Explosion

The Gas Line Was Leaking Before the Events of September 9

It’s unlikely that the pipe was just fine on September 8, and suddenly and without warning ripped apart on September 9.  Rather, in the days leading to September 9, the pipe likely had a small leak, sometimes called a “pencil leak.”  The pencil leak allowed gas to permeate the neighborhood. Gas found its way into the wall spaces of the nearby houses.  The gas met with an ignition source – perhaps a small spark from someone turning on a light switch. A flash fire began in the house.  The fire traveled back to the source of the gas – the pipe in the street. Only then did the pipe explode.

How do we know that?  Two reasons: Basic fire chemistry and residents’ reports of the smell of gas well before the explosion.

Fire Chemistry

Pure gas does not burn. Rather, for gas to ignite, it has to be mixed with air. Only when the gas-to-air mixture is 5-15 percent gas and 95-85% air will it burn. The gas-to-air mixture at the site of a high pressure pipe rupture is almost pure gas and so is generally too "rich" to burn.  A combustible mixture usually can be found only at some distance from the source of the leak.

Not only must the gas-to-air mixture be just right for it to be combustible, but that mixture needs to meet up with an ignition source.  Seldom is there an ignition source right at the site of the rupture. That’s another reason gas fires typically begin some distance away from the rupture and then work their way back to the leaking pipe.

The NTSB should quit hoping to find some flaw in the pipe that made it all come apart at once. It won’t find one. It should instead look for some small flaw that allowed a slow “pencil leak.”

Neighbors Smelled Gas

Next, we know that residents smelled gas in the days leading to the fire. Those residents include Chris Torres who lived at the epicenter, and others as reported here and here. The smell of gas in the days before the fire support the “pencil leak” theory.  But the preliminary report doesn’t talk about that at all. How can the NTSB ignore what the residents have to say?  

So, where exactly was PG&E in the days leading to the fire?  Unfortunately, the NTSB doesn’t seem to be interested in that part of the story.  Neither are the politicians.