By the time closure is paid there is no production left to depreciate it against. Pooled into capital under units of production, the deduction is stranded entirely.
Units-of-production depreciation spreads a cost over the tonnes or barrels that the cost helped produce. It is the right method for a processing plant. It is incoherent for closure.
Reclamation is paid after the last tonne is mined. There are no subsequent units for the deduction to attach to, so a units-of-production pool containing closure cost simply never depreciates. The tax shield that the model booked silently fails to arrive, and the error is invisible because the depreciation schedule still balances.
There is a second problem sitting underneath. Many models stop at the last year of production. Closure spending happens after that, so it falls off the end of the horizon entirely and the model values a mine that never gets cleaned up. The liability is real, it is often bonded, and somebody is going to pay it.
Check where the model horizon ends. If the last period is the last producing period, closure is not in the answer at all.
Find the closure line. Trace whether it is capital or operating cost, and if it is capital, find which pool it went into and what depreciates that pool.
Compare against the bond. Regulators require a bonded closure estimate. If the model's number is materially below the bond, the model is optimistic about something the regulator has already priced.
Treat closure as cash in the period it is spent. That is what it is. Its tax treatment is a separate question from its cash timing, and conflating them is what caused the problem.
Run the horizon past last production. Far enough to cover reclamation, monitoring and bond release, which can be many years for water treatment obligations.
Model the accrued liability separately from the cash. The balance sheet obligation and the cheque are different objects on different timelines, and lenders will ask about both.
An Asset Health Check finds this one on a producing asset in two weeks, fixed fee.
The Asset Health Check is the cheapest possible way to find out whether we are worth the larger engagement. Two weeks, fixed fee, and a written answer either way.