The sensor was in a "fairly sad state of repair".
Methane levels measured in the main shaft on November 15 and 17 were 2.75 per cent, but closer to the coalface they would have been at least twice as high.
However, records from those days did not include this.
Mr Reece said it was apparent sensors were not being calibrated and maintained because two that were close together gave different readings.
He said Pike River staff reports and modelling after the disaster showed a "serious lack of ventilation" for the number of coalfaces being worked.
"You could see that deputies at times were having trouble just getting air to move in the direction it should."
Putting the main fan underground and in a high-hazard area was "highly unusual".
"Stoppings" in some parts of the mine, used to direct air around the tunnels, were simply "pogo sticks and brattice and plasterboard". Another crucial one had a cloth trapdoor. The pipeline meant to drain gas could not cope with the volume and was probably blocked.
Mr Reece said he thought the first explosion was probably fairly deep inside the mine. The two survivors did not recall feeling a heatwave and the temperature at the point of ignition would have been 700C.
He said that in Australia there was a push towards three entries into a mine. Pike River had just one, plus a ventilation shaft.
- GREYMOUTH STAR