It’s entirely possible they had a shutoff device, but even if they did, the gas is under high pressure, so the now-open-to-atmosphere section of the pipe contains a lot of gas.
Suppose you have a 2-foot-diameter pipe running at 500 psi, and the shutoff valve is five miles away. There’s about 2.8 million standard cubic feet of natural gas that will come out of that pipe before it equalizes to atmospheric pressure.
Don’t know what the actual pressure was in the San Bruno pipe, but 500 psi doesn’t seem outlandish. You can insert your own estimates for pressure and shutoff-valve spacing and tweak the estimate accordingly.
In a background article in the paper this morning, it was said that even after the pipe was shut off the valve was far enough away from the site of the break that the gas would still burn for an hour.
It also said that the first responders were convinced for quite a while that a plane crash was the source of the fire, not totally crazy given the proximity of San Bruno to SFO. They eventually figured out that plane crashes make smoke, and this fire did not. That slowed the proper response a bit. Not surprising, given the chaos at the site, and the impossibility of getting close enough to see the cause.