Then switch it off before you go into the filling station.
And how close to the tanks could one be without switching off?
How about passing the station on the street, possibly with the tank vent
pipes between the street (i.e. yourself) and the pumps?

The major problem you'd have is static electricity. RF will pose no problem in
the normal amounts emitted by a standard mobile rig (i.e., 5-50 wts). The
ignition method most commonly seen is static buildup, person does not discharge
against the car body before touching the nozzle (grounded to the pump via the
hose...it has a wire mesh and/or grounding wire built into it). Nice spark
results, as does ignition of the vapors. One thing that was noted....women
are more likely to have this happen...best guess is due to two reasons. One,
material in clothing may be higher in static-generating capability. Second, on
cold days, women tend to return to the car and wait for the tank to fill. When
they get out, they're a nice static source looking for a ground. The nozzle
provides that.

As for RF, we used to use an old Clegg FM27B to test the electronics (faraday
chamber)....never had any fuel ignition problems, and the lockup problems were
mostly due to ground loop situations which were resonant on some more commonly
used frequency, such as 400 mhz +/-. Drove one prototype unit nuts until the
chief engineer (also a ham) happened to key his 70cm rig when sitting by the
unit. Turns out the local police liked the donuts sold at the station......

A
(ex-R&D/Sales at a pump company)