My logic is:
1. You have no proof that excessive pressure is used.
2. A rusted tank may not be leaking prior to the water cut-off. But when the water pressure is relieved and then restored, the metal experiences a change of pressure, first a reduction then an increase. This might be the last straw that break the back of your rusty camel...
