At handover, every review arrives after the leverage has gone

Handover assurance has a scheduling problem that no amount of rigour fixes.

The review is commissioned because handover is happening. It is resourced, scoped and executed on that basis. And handover, as an event, is the precise moment at which the contractor's obligation ends, the delivery team begins demobilising, and the commercial leverage that could have compelled anything disappears.

So the work is done well, the report is accurate, and the findings are unactionable. Everything it identifies is now a cost to the owner rather than an obligation of somebody else.

The same review, three months earlier

Nothing about the review itself needs to change. The questions are identical. What changes is that the answers can still alter what gets produced, because the party who would have to produce it is still on site, still unpaid in part, and still holds an interest in a certificate that has not yet been issued.

This is the cheapest available improvement in the whole lifecycle and it is a scheduling decision rather than a methodology. It requires somebody to commission handover assurance against a date that has not happened yet, which requires them to be thinking about handover during delivery, which is when nobody is thinking about handover.

The review is not late because anybody delayed it. It is late because it was triggered by the thing it needed to precede.

What gets waived, and by a legitimate route

There is a specific mechanism worth understanding, because it operates without anybody breaching a control.

Completion dates are visible. They are reported upward, they appear in announcements, they carry political weight, and a slip in one is felt immediately by people senior enough to act. Information conditions have none of those properties. Nobody outside the project knows whether the as-built records were complete on the day the certificate was signed.

When the two conflict, and near the end of a large programme they always conflict, the information condition gives way. Not through misconduct. Through a judgement, made under pressure, by someone entitled to make it, that the date matters more than the document and that the document can be chased afterwards.

The instruments have improved here in a way worth acknowledging. FIDIC's 2017 edition moved the obligation to supply as-built records and operation and maintenance manuals into Sub-Clause 10.1, making it an express requirement of taking over, and coupled it with a requirement for the Engineer to issue a Notice of No-objection. That is a real strengthening in a region where FIDIC based conditions are the norm, and it converts asset information from a schedule item into a completion condition.

The examinable question is therefore narrow and answerable. Read the taking-over certificate against the conditions the contract attaches to it. List what was outstanding on the date of issue. For each item, find who accepted it and what was recorded about the consequence.

That last part is where it usually stops, because the acceptance is almost always recorded and the consequence almost never is.

What the consequence turns out to cost

The test that costs nothing

There is one assurance activity at this stage that is genuinely cheap, genuinely decisive, and almost never performed.

Take a competent person who was not involved in the project. Give them only the handover pack. Ask them to answer three operating questions of the kind the asset will actually generate in its first year: where the isolation point is for a given system, what the maintenance interval is for a particular item and where that came from, what to do when a specific alarm activates.

If they can answer from the documents, the pack works. If they cannot, the pack is complete against its register and useless against reality, which is the normal condition and is invisible to a document review.

The reason this test is rarely run is that it can fail. A document register check produces a percentage. This produces a yes or a no, in front of witnesses, at a point when somebody would have to do something about it.

Two reconciliations, in both directions

The other thing created at this stage is the record of what the organisation now owns, and it is created twice, by two functions, from different sources.

Take ten assets from the financial register and locate them physically. Then take ten physical assets and find them in the register. The second direction is the one that fails first and the one nobody runs, because the register is treated as the authority and reality is treated as the thing that should match it.

Then compare the componentisation used for depreciation against the breakdown used for maintenance. Where they differ, establish which one anybody actually manages the asset by. Neither is wrong. They simply do not reconcile, and the discovery is usually deferred until something has to be revalued, insured or replaced.

Why the deadline for this arrived from the accounting side

The organisation, not the asset

Readiness is generally assessed as a property of the asset, and the asset is usually the part that is ready. It has been commissioned, tested and certified by people whose job was to do exactly that.

What is missing is the organisation. The roles, the trained people in those roles, the spares, the systems, the term maintenance contract, and the authority to spend money on a Tuesday without a committee. These have long lead times, no contractor is responsible for any of them, and they are therefore nobody's critical path until the week they become everybody's.

The single most useful question at this stage is whether the operator can raise, approve and pay for a corrective work order on day one. It tests the systems, the delegations, the contracts and the staffing simultaneously, and it has a yes or no answer.

The framework for this stage sets out all five domains and the published clause behind each, including the two the 2024 revision of ISO 55001 changed most, data and information at 7.6 and knowledge at 7.7.

Sources. FIDIC, Conditions of Contract, 2017 edition, Sub-Clause 10.1 on taking over the works and Sub-Clause 10.3 on interference with tests on completion, compared with the 1999 edition. ISO 19650-3, information management during the operational phase of assets. ISO 55001:2024, Clause 7.6 on data and information and Clause 7.7 on knowledge, both changed in the second edition, ISO standard 83054. IPSAS 33, first-time adoption of accrual basis IPSAS. ISO 21502:2020, Clause 6, covering project closure and transition. Standards published behind a paywall are cited without a link.