The liability born from the asset.

A wind farm enters operation and, at the same instant, an obligation is born that will only fall due in twenty or thirty years: to dismantle the turbines, remove foundations, and restore the land at the end of its useful life. This obligation goes far beyond an environmental footnote. It is a long-term liability that must be recognized today, measured with method, and revised periodically, and whose counterpart does not go to profit and loss but to the cost of the asset itself. The energy transition and the densification of governance requirements have placed decommissioning obligations at the center of the balance sheets of those operating energy, mining, and infrastructure, and most companies still treat them as generic provisions, when in essence they are an asset valuation problem disguised as an accounting provision.

The confusion is understandable, because the decommissioning liability lives on the frontier between three standards that must speak to each other. It is a provision, governed by the provisions and contingent liabilities standard. It is born capitalized into property, plant and equipment by requirement of the PPE standard, which mandates including in the cost of the asset the estimate of dismantling and restoration costs. And its subsequent changes follow a specific interpretation, which disciplines exactly how to treat changes in that liability over time. Whoever sees only the provision misses the cascade effect on the asset, on depreciation, and on the recoverability test.

How the liability enters the balance sheet

At initial recognition, the future cost of dismantling and restoration is estimated, projected to the date when it will occur, and brought to present value by a discount rate reflecting the time value of money and risks specific to the liability. The present value calculated does two things at once: it recognizes the decommissioning liability and increases the cost of property, plant and equipment by the same amount. From there, the asset is depreciated by a higher amount, distributing the cost of future decommissioning over the useful life, and the liability grows each period by the unwinding of the discount, the so-called discount unwind, which flows through finance expense until reaching, on the decommissioning date, the estimated nominal value of the outflow.

The practical consequence tends to go unnoticed: the cost of removing an asset in the future is recognized while the asset generates benefit, not only when the check is signed. The result of each period carries its corresponding decommissioning portion, and the balance sheet stops hiding a certain obligation behind a distant date.

Why the discount rate moves everything at once

Here is the point that makes the topic technically rich and dangerous. The measurement of the decommissioning liability is dominated by two long-term estimates, future removal cost and discount rate, and both are revisited periodically. When the discount rate changes, and in a high-interest environment it changes materially, the present value of the liability changes, and the applicable interpretation mandates adjusting the counterpart in the carrying value of the asset, not directly through profit and loss.

The effect is chained. A rise in the discount rate reduces the liability and, by the same measure, reduces the carrying value of the asset. A reduction in the rate does the inverse, and may inflate the value of the asset to the point of requiring verification. This is why the standard contains a safeguard: the increase in the carrying value of the asset arising from the liability revision is an indicator that recoverability of that asset may need to be tested, bringing the impairment standard into the discussion. A single premise, the discount rate, simultaneously moves the liability, the asset, future depreciation, and potentially the impairment test. Treating that premise lightly is treating the entire long-term balance sheet of the operation lightly.

The cost estimate is an engineering problem

The most underestimated portion of the problem is the estimate of the future cost of removal and restoration. How much it costs, three decades from now, to dismantle turbines, remove reinforced concrete foundations, decontaminate soil, dispose of equipment, restore vegetation. Before being an accounting question, it is an engineering and asset valuation question, and the quality of the answer determines everything that follows, from the recognized liability to the expense of each period.

This work requires the vocabulary of machinery, equipment and installations valuation, and requires the signature of someone with technical engineering competence to size the removal effort. This is where fixed assets valuation and provisions accounting meet: the number that goes to the liability is born from an asset cost exercise conducted by a qualified engineer, and this origin is what gives it defensibility before the auditor.

The sector that made the issue urgent

Renewable energy has transformed decommissioning from exception to rule. Wind farms and solar plants have defined useful lives and occupy large areas requiring restoration, and sector regulation has come to treat decommissioning as a responsibility to be planned from the start, not at the project's closing lights. For those who operate, finance, or audit these assets, the decommissioning obligation has ceased to be a remote hypothesis and has become a relevant balance sheet line, with direct effect on results, on covenants, and on the fair value of the project.

Estimating the future cost: engineering, escalation and contingency

The number that enters the liability is born of an engineering question, and the way to answer it defines the quality of the entire measurement. The robust estimate is built from the bottom up, from the real physical scope of decommissioning: equipment dismantling, foundation removal, decontamination and land restoration, transport and disposal of materials. Each component is priced at current cost, based on productivity, labor, equipment mobilization, and waste disposal, and only then the set is projected to the expected disbursement date, applying a price-escalation expectation over the horizon. A generic percentage arbitrated on the initial investment may produce a number but does not produce an estimate, because it lacks anchoring in what will actually be done.

Contingency, far from being padding or pessimism, is the honest recognition that estimating an outflow that will occur twenty or thirty years from now involves genuine uncertainty about removal technology, future regulatory requirements, and asset condition at end of life. Handling that uncertainty with scenarios, rather than a single point, allows the valuer and the auditor to discuss the estimate in terms of range and sensitivity, without feigning cent-level precision over a distant event. Periodic revision of this estimate, as the date approaches and uncertainty diminishes, is part of the liability's normal life cycle, not a sign of original error.

The discount rate and the treatment of risk

The provisions standard requires that the liability be measured at present value of expected outflows, discounted at a rate reflecting the time value of money and risks specific to the liability, without double-counting. Here lies a technical care many models run over: either risk is adjusted in the expected cash flow, estimating outflows so as to already incorporate uncertainties, or risk is adjusted in the discount rate, but never both at once for the same risk. Counting the same risk twice inflates the liability without foundation and contaminates the entire chain that depends on it.

Sensitivity of the result to that rate is high, because the horizon is long and small variations in the rate produce relevant variations in present value. That is why the choice and documentation of the rate is not formality. There is also a question that divides standards and generates legitimate debate: whether the rate should embed the credit risk of the entity itself that will bear the outflow. The international provisions standard orients to reflect time value of money and risks specific to the liability, and tends to keep own credit risk out of the calculation, while the Brazilian present-value adjustment rule treats the matter differently. Declaring which premise was adopted, and why, matters more than defending either side. In a structurally high-interest environment like Brazil's, the rate applied to a long-term liability carries disproportionate weight on the value recognized today, and its periodic revision moves liability, asset, and expense in chain. In sectors where regulation requires financial guarantees or earmarked funds to secure decommissioning, this discipline ceases to be merely accounting and acquires cash effect, reinforcing that the estimate must be as solid as any other certain obligation on the balance sheet.

The principle

The decommissioning liability teaches a lesson that holds for all asset valuation: nothing that has a beginning has cost only at the beginning. The sooner decommissioning enters the calculation, the less it surprises when the time comes.