That is the problem with the report. Intrinsically safe barriers will not transmit high current to the sensor even if the wiring on the other side is shorted to power buses. That is the purpose of “intrinsically safe”. Wire separation, not an issue. Short circuit, not an issue. Deposits inside the tank, not an issue. The current in the sensor cannot reach levels that ignite anything in the tank. That is simply the design criteria of the barrier. The barrier itself is encapsulated.
What the NTSB found was that the wiring and components in the tank were not intrinsically safe:
1)Exposed conductors on FQIS wiring (caused by either mechanical damage or cold-flow) within a fuel tank could provide a mechanism that would lead to arcing inside the tank, which in turn could ignite the flammable fuel/air vapor. (This was one of the suspected ignition mechanisms for the 1990 Philippine Airlines 737 CWT explosion and the 1972 Navy C-130 fuel tank explosion.) Very little (about 4 feet) of the CWT FQIS wiring from the accident airplane was recovered, and, therefore, the degree to which the wiring in the tank might have been damaged before the accident could not be assessed. However, investigators found preaccident damage, including exposed conductors, on some of the recovered FQIS wiring from inside TWA flight 800’s wing tanks, and damaged FQIS wiring was found inside the CWTs of several of the other 747 airplanes examined by the Safety Board. In addition, the presence of a conductive material, such as metal drill shavings or safety wire (both of which were found in the fuel tanks of other airplanes), could have provided a mechanism that would lead to arcing of FQIS components. Although no clear evidence of arcing was found inside TWA flight 800’s CWT, fire damage along the route of the FQIS wiring was severe enough that it likely would have obscured any such evidence.
2)Another potential source of ignition energy is resistance heating, which could have resulted from a thin filament being heated through contact with a wire, probe, or compensator exposed to excess voltage. Although no clear evidence of a filament ignition was found inside TWA flight 800’s CWT, such evidence could also have been physically lost or obscured by fire damage.
3) Research and testing conducted during this investigation found that silver-coated copper parts inside fuel tanks, such as those used in the FQIS, can develop silver-sulfide deposits that are semiconductive and, therefore, can reduce the resistance between electrical connections and permit arcing. Such deposits can become a potential ignition mechanism inside a fuel tank.