The reservoir is treated as one tank. Matched to history, the balance estimates hydrocarbons in place and which drive is doing the work: oil and gas expansion, a gas cap, or water moving in. The usual plots compare underground withdrawal with those expansion terms.
The method cannot see where the remaining oil sits. A map or a simulation is needed for that. It also fails when the reservoir is so compartmentalised that one average pressure does not describe it. A pressure used in the balance has to be a static reservoir pressure, not a flowing pressure and not a buildup that was still climbing.