Finance Gazette analysis of five UK-listed energy groups with significant North Sea exposure finds decommissioning provisions moving very differently as acquisitions, revised estimates, discounting, currencies and asset sales offset hundreds of millions in clean-up activity.
Hundreds of millions of dollars can be spent decommissioning oil and gas assets without producing anything close to an equivalent reduction in the liability sitting on a company's balance sheet. Finance Gazette analysis of five UK-listed energy groups with significant North Sea exposure shows how far the two can diverge.
Serica Energy's decommissioning provision increased 72.4% during 2025, while EnQuest's rose 23.5% and Ithaca Energy's 16.1%. Harbour Energy moved in the opposite direction — but only just: it used $374 million of its decommissioning provision during the year, while the closing liability fell by only $93 million.
The explanation lies beyond the amount companies physically spend. Acquisitions, revised cost estimates, discount unwinding, currency movements and asset disposals can all materially reshape the provision, sometimes overwhelming the effect of the cash spent during the year.
| Company | 2024 provision | 2025 provision | Change | 2025 utilisation/spend* |
|---|---|---|---|---|
| EnQuest | $741.6m | $915.6m | +23.5% | $38.5m utilisation / $56.8m cash spend |
| Ithaca Energy | $2.655bn | $3.082bn | +16.1% | $107.2m |
| Harbour Energy | $7.114bn | $7.021bn | -1.3% | $374m |
| Serica Energy | $145.7m | $251.1m | +72.4% | $1.1m |
| Centrica | £1.459bn | £1.302bn | -10.8% | £71m |
*Companies do not use identical definitions for provision utilisation and reported cash decommissioning expenditure. Finance Gazette has retained the companies' reported measures rather than treating them as interchangeable.
Source: Finance Gazette analysis of company annual reports, financial statements and full-year results. Percentage calculations are by Finance Gazette.
The closing balances tell only part of the story. A provision can increase because a company acquires assets carrying future clean-up obligations, because expected costs are revised, because the passage of time raises the present value of discounted liabilities, or because currency movements alter the reported balance. Equally, selling an asset can move an obligation elsewhere without any corresponding physical decommissioning taking place.
| Company | Major upward drivers | Major downward drivers | 2025 result |
|---|---|---|---|
| Serica | Acquisitions $56.5m; estimates $41.7m; discount unwind $6.6m | Payments $1.1m | +72.4% |
| EnQuest | Acquisition $89.1m; additions $46.7m; estimates $40.9m; discount unwind $35.9m | Utilisation $38.5m | +23.5% |
| Ithaca | Business combinations, estimate revisions and accretion | Decommissioning expenditure $107.2m | +16.1% |
| Harbour | Discount unwind $284m; FX $274m | $374m used; estimate reductions and asset transfers | -1.3% |
| Centrica | — | £71m decommissioning spend; position also affected by disposal-related liabilities | -10.8% |
Source: Finance Gazette analysis of company disclosures. The table highlights significant disclosed drivers rather than representing every accounting movement within each reconciliation.
Taken together, the five companies show that cash spent on decommissioning is only one of several forces determining the liability that investors eventually see on the balance sheet. The movement bridge is often more informative than the headline year-end number because it reveals what is actually changing the obligation.
Serica Energy provides the sharpest example. Its decommissioning provision increased from $145.7 million to $251.1 million during 2025, a rise of approximately 72.4%, while payments against the provision were just $1.1 million.
| Serica 2025 movement | $m |
|---|---|
| Opening provision | 145.7 |
| Acquisitions | +56.5 |
| Changes in estimates | +41.7 |
| Discount unwind | +6.6 |
| Payments | -1.1 |
| Currency movement | +1.7 |
| Closing provision | 251.1 |
Acquisitions and revised estimates added approximately $98.2 million between them, explaining most of the increase before discounting or currency movements are considered. Serica's closing liability was therefore shaped far more by changes in its asset base and expected future costs than by decommissioning payments made during the year.
Harbour Energy presents almost the reverse picture. Its decommissioning provision started 2025 at $7.114 billion, and $374 million was used during the year, yet the closing provision was still $7.021 billion — only $93 million below where it began.
Two movements explain much of that apparent contradiction. Unwinding the discount added $284 million, while foreign-exchange movements added another $274 million, meaning those two items alone increased the provision by $558 million.
Other movements went the other way, including changes in cost estimates and amounts associated with assets held for sale. Simply subtracting the amount used from the opening provision would therefore give a misleading picture of Harbour's year.
Decommissioning expenditure can occur decades into the future, so companies estimate the future cash cost and discount those payments to calculate the liability recognised today. Harbour expects its decommissioning expenditure to extend over approximately the next 40 years, with around half expected between 10 and 20 years from the reporting date.
Its sensitivity disclosures show how important those assumptions are. Holding everything else constant, a one percentage-point reduction in applied discount rates would increase Harbour's provision by approximately $312 million, while a 10% increase in estimated decommissioning costs would add approximately $740 million.
Those are substantial balance-sheet movements generated by changes in assumptions about future expenditure and its present value, rather than by the amount of physical decommissioning completed during the reporting year.
Ithaca Energy increased its provision from $2.655 billion to $3.082 billion during 2025 — approximately 16.1%. The liability represents the present value of costs expected over approximately 40 years, and calculating it requires assumptions covering cessation-of-production dates, future rig and vessel rates, post-production operating expenditure, inflation and discount rates.
Its scale is also notable alongside Ithaca's annual cash generation. The company reported 2025 free cash flow of $683.3 million, putting its year-end decommissioning provision at approximately 4.5 times that year's FCF.
That figure is not a leverage or solvency ratio because it compares a balance-sheet stock representing expenditure potentially stretching over several decades with cash generated during a single financial year. Used cautiously, however, it gives useful context for the scale of the long-term obligation.
EnQuest increased its provision from approximately $741.6 million to $915.6 million, or 23.5%, during 2025. The reconciliation included approximately $89.1 million from acquisition activity, $46.7 million of additions, $40.9 million from changes in estimates and $35.9 million from unwinding the discount.
Approximately $38.5 million was utilised against the provision, while EnQuest separately reported $56.8 million of cash decommissioning expenditure. Those two figures should not be treated as interchangeable because provision utilisation and cash expenditure can differ as accrual movements are recognised.
The larger point is that acquisition activity, additions and revised estimates outweighed the reduction associated with utilisation, leaving EnQuest with a substantially larger provision at year-end despite ongoing decommissioning activity.
Centrica saw its group provision fall from £1.459 billion to £1.302 billion during 2025, a decrease of approximately 10.8%. Around £961 million of the closing provision related to Spirit Energy and another £321 million to Rough.
Centrica reported £71 million of decommissioning expenditure, but spending was not the only factor affecting its exposure. Its year-end disclosures reflected £129 million of decommissioning liabilities associated with disposals: £85 million relating to the completed Cygnus disposal and a further £44 million connected with a disposal agreed in December and classified as held for sale.
A reduction in a company's provision therefore does not necessarily mean an equivalent amount of decommissioning work has been completed. Ownership changes can alter who carries the future obligation.
Centrica also provides useful visibility over when the liability is expected to turn into expenditure. Approximately £734 million of its £1.302 billion provision falls within 2026–2030, with another £549 million during 2031–2035.
The five companies show why the closing provision is only the starting point. Cash expenditure matters, but so does the reason the provision moved: Serica's 72% increase was dominated by acquisitions and revised estimates, Harbour's $374 million of utilisation was substantially counteracted by discounting and currency movements, and Centrica's falling liability was influenced by changes in asset ownership as well as decommissioning expenditure.
Timing adds another layer. A £1 billion provision expected to result in cash payments over several decades has different implications from the same amount concentrated into the next few years, while sensitivity to assumptions can be just as important. Harbour's disclosures show that relatively modest changes in discount rates or estimated costs can alter the recognised liability by hundreds of millions of dollars.
This is the same reason headline financial measures often need to be read alongside the underlying financing structure. Finance Gazette's recent analysis of supplier-finance arrangements at eight major UK-listed companies found that the balance-sheet amount alone did not reveal how extensively suppliers were actually using the programmes or how those arrangements related to cash generation.
For investors, the useful analysis is therefore not simply whether the provision increased or decreased. It is how much cash was spent, what else moved the provision, how sensitive the estimate is and when the underlying expenditure is expected to occur.
The difference between accounting provisions and eventual cash expenditure is particularly visible at Harbour. Its recognised decommissioning provision at the end of 2025 was $7.021 billion, while the company's estimated undiscounted decommissioning and restoration obligation was approximately $10.5 billion.
The roughly $3.5 billion difference does not represent clean-up work that disappears; much of the expenditure is expected years into the future, so its present value is lower than the nominal amount expected ultimately to be spent. The same principle runs through decommissioning accounting more broadly, with today's liability combining engineering estimates and assumptions about future costs, inflation, asset lives, field closure dates and discount rates.
As those inputs change, the balance sheet changes with them.
Finance Gazette reviewed the latest full-year financial disclosures of five UK-listed energy groups with significant exposure to North Sea oil and gas: EnQuest, Ithaca Energy, Harbour Energy, Serica Energy and Centrica. The analysis extracted reported decommissioning provisions and, where disclosed, reconciled movements including acquisitions, additions, revisions to estimates, discount unwinding, expenditure or utilisation, disposals and foreign-exchange movements.
Companies do not report every measure on an identical basis. In particular, provision utilisation and separately reported cash decommissioning expenditure can differ, so Finance Gazette has preserved the terminology used in the underlying financial disclosures rather than treating those figures as equivalent.
The comparison uses group-level provisions. Ithaca and Serica are heavily UK-focused, while groups including Harbour and EnQuest also carry decommissioning obligations relating to international operations. The figures should therefore be read as a comparison of UK-listed energy groups with significant North Sea exposure, rather than as a calculation of exclusively UK North Sea liabilities.
Where a decommissioning provision is compared with annual free cash flow, the calculation illustrates relative scale only. A year-end provision is a stock representing estimated future expenditure potentially extending over decades, while free cash flow is generated over one financial year. Finance Gazette has used that distinction elsewhere when analysing the relationship between cash generation and large corporate cash commitments.
This Finance Gazette analysis was constructed principally from company annual reports, audited financial statements and full-year results. Finance Gazette extracted and standardised the reported figures and calculated percentage changes and comparative metrics from those disclosures.
The five companies produce a strikingly different picture once the movements behind their headline provisions are separated. Serica's liability increased more than 70% while payments against its provision were little more than $1 million, Harbour used $374 million against its provision yet discount unwinding and currency movements added $558 million in the opposite direction, and Centrica's year-end position was also affected by liabilities associated with asset disposals.
These are not marginal accounting details. They can determine whether a multi-billion-dollar liability rises or falls despite the cash a company has spent addressing it.
For investors, that changes the question from simply “How large is the decommissioning provision?” to what caused it to move, how much cash is actually being spent, when the remaining expenditure will fall due and how sensitive today's estimate is to changes in costs and discount rates.
Across these five companies, those answers tell a substantially richer story than the closing liability alone.
