Keep the useful parts together

Field Note 01 / The evidence behind a defensible closure estimate

END-FN-01 · Rev B · 14 September 2026 · Technical note

1. Start with the decision and the boundary

An estimate prepared to compare removal methods answers a different question from an estimate supporting a financial provision. State the purpose, asset boundary, end state, pricing date, execution window and exclusions before presenting the total. Otherwise, two apparently conflicting numbers may simply describe different work.

Begin with a physical inventory that can be reconciled to the scope. A platform, pipeline, well and onshore receiving facility can each introduce separate work packages. Include the interfaces: isolation, cleaning, temporary support, transport, waste acceptance, surveys and evidence required after execution. An exclusion needs both a reason and an identified destination; “by others” is incomplete unless someone owns that work in another package.

2. Connect quantities to a practicable method

A quantity multiplied by a rate is only useful when the method behind the rate remains feasible. Explain what must be true for the selected vessel, lifting sequence, cutting approach or waste route to work. Record whether quantities come from reconciled drawings, survey, inspection, contractor information or an assumption. Keep the source revision beside the input rather than in an unstructured attachment folder.

Consider a module with an uncertain weight. The important question is not merely how much steel to add to the quantity. It is whether the uncertainty could change the lifting arrangement, crane selection, offshore sequence or transport configuration. That is a method uncertainty with several cost consequences. Treating it as a small quantity allowance can hide the actual decision.

3. Make uncertainty reviewable

Separate missing scope, variable quantities, uncertain productivity, market rates and discrete events. Record how each has been treated so that the same exposure is not counted in a duration allowance, a rate premium and contingency. A range should have an explanation of what changes across it. A confidence label unsupported by a defined method does not make the estimate more reliable.

For financial reporting, AASB 137 addresses recognition and measurement of provisions, including the best estimate, uncertainty and discounting where material. Those are accounting judgments. Engineering should supply the scope, methods, timing and evidence needed to inform them, with a clear reconciliation to the finance team's calculation. AASB 137, paragraphs 14 and 36 to 47 ↗ (opens in a new tab).

4. Build the review file alongside the estimate

ASA 540 focuses the auditor's work on matters including methods, assumptions, data and estimation uncertainty. ASA 500 addresses the sufficiency and appropriateness of audit evidence. Neither standard provides a label that an engineering consultant can award to an estimate in advance. ASA 540 ↗ (opens in a new tab), ASA 500 ↗ (opens in a new tab).

A practical review file contains a scope-to-cost map, an assumption register, source records, method narrative, sensitivity results and changes since the preceding estimate. Record unresolved contradictions explicitly. The key test is whether a reviewer can select a material cost line and trace it back to an identifiable obligation, asset, quantity and execution assumption without relying on the estimator's memory.

Field Note 02 / The temporary state can govern the lift

END-FN-02 · Rev B · 14 September 2026 · Technical note

1. Define the states before checking the crane

A heavy lift is a sequence of structural configurations. The object may begin on its original supports, transfer onto temporary supports, become partly disconnected, transfer load into rigging and finally land on a transport support arrangement. Each state has a different load path. The original operating design does not establish that these temporary configurations are adequate.

Create a state register with a drawing for each meaningful change. Identify what carries the load, what restrains movement, what has been cut, which connections remain and what observation confirms the transition is complete. Include pauses in the work, because a configuration intended to last briefly may need to remain stable when weather or access interrupts the sequence.

2. Trace the complete load path

Follow load from the object through its local members, lifting points, attachments, rigging, crane and supporting vessel or ground. Review the landing path with the same discipline. A lifting point check alone says little about the surrounding structure if corrosion, modifications or removed bracing have changed its capacity.

For inherited equipment, establish the weight and centre-of-gravity basis and the uncertainty around them. Identify trapped liquids, residual contents, undocumented additions and temporary equipment. Decide which uncertainties can be resolved by inspection or measurement and which need to remain as design cases. The objective is a justified operating envelope, not a single optimistic weight.

UK HSE lifting guidance emphasises competent planning proportionate to complexity, with actions and responsibilities defined for each step. Its temporary works guidance includes permanent works used for temporary purposes and describes coordination and design checking. These are useful technical references; their legal language belongs to the UK context. HSE lifting operations ↗ (opens in a new tab), HSE temporary works ↗ (opens in a new tab).

3. Review the transition, not only the endpoints

The installed and suspended states can both be acceptable while the transition between them is poorly understood. During cutting, an attachment may still carry load. During initial hoisting, restraint can remain on one side. During landing, an uneven contact sequence can concentrate load before all supports engage. The engineering review should ask how those conditions are detected and managed.

Define hold points around irreversible steps. The required evidence could include an inspection release, confirmation of isolation, verification of rigging configuration or measured response within an approved envelope. Do not turn this reading note into an execution procedure: acceptance limits and recovery actions must come from the project-specific engineered method.

4. Make the interfaces visible

Allocate responsibility for the object, temporary steelwork, rigging, crane, vessel response and landing arrangement. Agree what information crosses each boundary and who accepts it. A contractor drawing can be internally correct while using a load, geometry or sequence that differs from another contractor's basis.

Safe Work Australia's crane guide provides Australian general guidance on planning, risk management and coordination among those involved. The applicable offshore or onshore statutory regime still needs to be established for the actual operation. General guide to cranes ↗ (opens in a new tab).

The review is ready when the team can explain what supports and restrains the object at every stage, what could invalidate that explanation and what prevents progression when the evidence is incomplete.

Field Note 03 / Inherited asset records: find the evidence that changes the decision

END-FN-03 · Rev B · 14 September 2026 · Technical note

1. Organise around decisions

An archive organised only by document type is difficult to use for closure planning. Start with the decisions in front of the owner: define the inventory, establish isolation boundaries, select a removal method, identify a waste route or demonstrate the final condition. Then identify the evidence each decision needs.

This makes prioritisation more specific. A missing drawing for a redundant instrument may have little influence on an early option comparison. An unresolved discrepancy in a structural connection or a well barrier record can materially change the method or the next investigation. The same document gap has different significance at different stages of the project.

2. Separate identity, currency and adequacy

For each important record, establish what asset it describes, which revision it represents and what it actually demonstrates. “Latest available” does not necessarily mean as-built. “As-built” does not establish present condition. A certificate, drawing, inspection report and operating log answer different questions and should retain those distinctions in a digital system.

Keep conflicting records visible until they are resolved. A useful discrepancy register records the competing statements, the decision affected, the interim assumption, the accountable discipline and the evidence needed to close the issue. Replacing an inconvenient record with a newer file can erase the very uncertainty the project needs to understand.

3. Verify the claims that carry consequence

Use targeted verification to test assumptions that affect the work. Examples include matching an equipment tag to the physical item, confirming a support arrangement, investigating an unrecorded modification or reconciling a cleaning record with the receiving facility's information needs. Survey and inspection should have a defined question and acceptance basis rather than a generic instruction to collect more data.

The evidence system should preserve original files and distinguish extraction from interpretation. A search result or AI-generated summary is a route into the record, not proof that the source is current, complete or technically applicable. Record the source location for extracted values and make engineering interpretations visibly attributable to a review step.

ASA 500's treatment of the relevance and reliability of information is useful comparative reading for this discipline of evidence. It is an auditing standard and does not prescribe a universal engineering document-control system. ASA 500 ↗ (opens in a new tab).

4. Connect the record to the asset's changing condition

An inherited record describes the asset at a time. Closure planning also needs an account of what has changed since then. Maintenance status, deterioration, modifications, isolation changes and access constraints can invalidate an earlier removal assumption without changing the original drawing.

NOPSEMA's March 2026 property policy discusses maintenance that enables safe removal. Its proactive-planning paper also connects late-life maintenance decisions with future decommissioning capability. These are regulatory sources for Commonwealth offshore context, not a substitute for asset-specific assessment. Section 572 policy ↓ (opens in a new tab), Planning for proactive decommissioning ↓ (opens in a new tab).

The deliverable is a decision-ready evidence register: supported claims, unresolved conflicts, consequential assumptions and the next verification actions. Counting files can describe an archive. It cannot establish whether an asset is ready for a closure decision.

Field Note 04 / How to read a NOPSEMA record without overreading it

END-FN-04 · Rev B · 14 September 2026 · Technical note

1. Identify the record before summarising it

Start by recording the titleholder, named activity, document type, submission or issue date, revision and displayed status. Save the date on which the public record was checked. Those details form the boundary of the claim: a decision relating to one activity or revision should not silently become a statement about an entire field.

NOPSEMA's environment-plan portal distinguishes proposals open for comment, submissions under assessment and approved projects and activities. It explicitly states that plans on the open-for-comment page have not undergone NOPSEMA assessment. Public availability therefore does not itself establish regulatory acceptance. Environment-plan portal ↗ (opens in a new tab).

2. Read status together with scope

An accepted plan should be read for the activity it covers, including its boundaries, commitments and exclusions. A decommissioning program can involve different work at different times. Cleaning, well abandonment, topsides removal and a final end state should not be assumed to sit inside the same accepted scope simply because they concern the same asset.

Use a simple record map. Put the question in one column, the relevant public record beside it and the conclusion that the record supports in a third. Add a fourth column for what remains unestablished. This makes the analysis useful for a board or project team without suggesting that a public portal contains every document the regulator has considered.

3. Preserve the meaning of enforcement terminology

NOPSEMA publishes directions, improvement notices and prohibition notices, but explains that the publication provisions do not cover every type of letter or enforcement action. Its register also explains that a notice marked closed means the duty holder has notified NOPSEMA of compliance. A revoked direction is cancelled from the specified date. These terms should be reproduced with that meaning rather than expanded into a broader assurance statement. Published directions and notices ↗ (opens in a new tab).

Read the underlying notice and related documents when they are available. Record exactly who must do what, by when and in relation to which facility or activity. Do not infer the current condition of an asset from the title of a historic notice alone. Equally, an absence of a result in a keyword search is not evidence that no relevant regulatory action exists.

4. Separate activity permission from title surrender

NOPSEMA's March 2026 Section 270 policy describes its advice to the Joint Authority on title surrender. The Joint Authority makes the surrender decision. That is a different decision from acceptance of an activity's environment plan. Section 270 policy ↓ (opens in a new tab).

For a practical review, distinguish three statements: what the titleholder proposed; what the regulator's published record says; and what engineering inference is being made. Place a source and checked date beside the first two. Mark the third as interpretation and state what evidence could change it.

The resulting reading note should allow someone else to reproduce the search and challenge the conclusion. It should not rank an operator's compliance, assert completion of unobserved work or claim that a particular liability has disappeared unless the relevant evidence directly supports that conclusion.

Field Note 05 / Closure decisions are often won or lost at the interfaces

END-FN-05 · Rev B · 14 September 2026 · Technical note

1. Describe the end state in observable terms

“Decommissioned” is too broad to serve as a completion criterion. Describe what has been removed, what remains, the condition of anything retained and the evidence required to demonstrate that condition. Keep physical completion, completion of a contract and a regulatory decision as separate milestones.

For Commonwealth offshore petroleum, NOPSEMA describes full property removal as the base case. Alternative approaches need the relevant demonstration and applicable requirements addressed. An option study should therefore state its permissioning assumptions explicitly, rather than treat a potentially lower-cost end state as available by default. NOPSEMA decommissioning overview ↗ (opens in a new tab).

2. Follow each item across the boundary

Trace a representative item from its present location to its final destination. That journey can cross operating systems, isolation boundaries, temporary works, lifting and transport arrangements, port access and a receiving facility. At each handover, identify the incoming condition, outgoing condition and evidence passed with the item.

A removal package may assume that equipment is clean enough for transport while a waste package assumes that decontamination will happen offshore. Both packages can look complete when read separately. A boundary review exposes the missing task. The same reasoning applies to survey coverage, pipeline ends, subsea debris and the records that demonstrate well work has been completed.

3. Align contract completion with evidence completion

Create an interface register before procurement. For each boundary, record the supplying party, receiving party, acceptance criteria, dependency, decision owner and required record. Include the assumption that would force the handover to be redesigned. Where an outcome depends on a third-party facility, establish what information that facility needs to decide whether it can receive the material.

The engineering task is to make these conditions explicit. The applicable transport, waste, environmental, safety and offshore requirements depend on the actual route and jurisdiction. A single offshore permissioning document should not be assumed to settle all onshore obligations. The government's Roadmap discusses decommissioning as a value chain that includes materials management and onshore capability. Australia's Offshore Resources Decommissioning Roadmap ↗ (opens in a new tab).

4. Plan the closure evidence before execution

Decide in advance which observations, inspections, surveys and records will demonstrate the required end state. Otherwise, the project may discover that an area is inaccessible or a component has been disposed of before necessary evidence was captured. A well-defined evidence requirement also gives contractors a more useful completion test than a general promise to provide a final report.

NOPSEMA's Section 270 policy addresses property, wells, natural resources and seabed or subsoil condition in its advice on title surrender. It notes that compliance with accepted permissioning documents is a primary means of demonstrating the relevant matters, but is not an assurance of NOPSEMA's satisfaction at surrender. Section 270 policy ↓ (opens in a new tab).

A practical close-out review reconciles the asset inventory, work completed, approved variations, remaining obligations and evidence still outstanding. Every residual item should have an owner and a defined next decision. That creates a usable account of progress without compressing several different kinds of completion into one green status.

Field Note 06 / Estimate drift: explain what changed before explaining the total

END-FN-06 · Rev B · 14 September 2026 · Technical note

1. Preserve the previous basis

Begin a revision by retaining the prior estimate, scope, assumptions, pricing basis and schedule. Do not overwrite the evidence needed to explain the movement. Establish which comparison is being made: like-for-like work, the remaining work after execution, or the full program including actual expenditure.

Classify changes before adding them together. Useful categories include physical scope, new asset evidence, execution method, schedule, market rates, actual performance and financial assumptions. Apply a consistent order to the reconciliation because interacting changes can produce different intermediate values. A visible reconciliation is more informative than assigning every difference to escalation.

2. Identify the mechanism behind the movement

A later execution date can change more than discounting. It can alter inspection needs, maintenance duration, equipment condition, vessel access and the validity of a contractor quotation. A newly discovered weight discrepancy may require a different removal method rather than a simple quantity adjustment. Explain the mechanism so that the next decision targets the cause.

Do not treat better information as automatic cost reduction. Inspection can confirm the original basis, narrow a range or reveal additional work. All three outcomes can improve the estimate. The useful question is whether the revised assumptions are better supported, not whether the total moved in the preferred direction.

NOPSEMA's proactive-planning paper warns that inadequate maintenance can make later removal more difficult and costly. That provides a regulatory context for reviewing the engineering effects of deferral; it does not establish a universal escalation rate. Planning for proactive decommissioning ↓ (opens in a new tab).

3. Reconcile engineering cost and financial reporting

AASB 137 requires provisions to be reviewed at each reporting date. AASB Interpretation 1 addresses changes in existing decommissioning liabilities, including revised cash flows, discount rates and the passage of time. The accounting treatment depends on the applicable facts and standards. An engineering cost movement should therefore be reconciled separately from the finance team's measurement and presentation adjustments. AASB 137, paragraph 59 ↗ (opens in a new tab), AASB Interpretation 1, paragraphs 2 to 4 and 8 ↗ (opens in a new tab).

The January 2026 amendments add a disclosure example involving high future decommissioning costs with an immaterial discounted carrying amount. It illustrates why information about the obligation can still be material, including where earlier closure is a significant risk. A small present value is not a physical measure of the work remaining. AASB 2026-1, application p. 5 and Example 2A p. 7 ↓ (opens in a new tab).

The amendments apply to annual periods beginning on or after 1 July 2025 that end on or after 28 February 2026. Earlier application is permitted. Applicability depends on the entity and reporting period.

4. Turn variance into an action

For each material driver, identify whether it can be resolved through evidence, a commercial enquiry, a method decision or active risk management. Assign a date for reconsideration and specify the evidence that would change the assumption. Keep improvement actions distinct from unsupported savings commitments.

The revision is useful when management can see what was learned, what was spent, what changed externally and what remains uncertain. A stable total can still deserve attention if offsetting movements conceal a less mature method or a growing schedule dependency. The movement report should make those changes visible even when the headline number looks reassuring.