Opinion

19650 review: the major project perspective part 2

Lawrence Chapman

BIM lead, Sizewell C

15 April 2026

Having detailed yesterday (14 April) the aspects of the 19650 revisions he rates, today Lawrence Chapman reveals where he thinks the changes don’t go far enough for major projects.

19650 review image in pink (contains elements from Dreamstime). Image: Heather Rugeley/Dreamstime — Image: Heather Rugeley/Dreamstime

Let’s start with the sequential model and programme reality. The nine-step process is presented as strictly sequential: all activities in Step 1 must be complete before Step 2 begins; Step 2 before Step 3; and so on. The document states explicitly that “the order in which the nine steps are presented shall be the order in which they are undertaken”. On a straightforward single-asset project, this is workable. On a major programme, it describes a fiction.

At Sizewell C, as on HS2 and virtually every other major programme of comparable scale and duration, the organisational information management strategy (Step 1) is never finished: it evolves continuously as the programme matures, the regulatory environment changes, and new assets and work packages come into scope. Meanwhile, multiple projects (Steps 3 through 9) are running simultaneously, all drawing on and contributing to the same asset information model (AIM). The standard offers no guidance on how these concurrent flows are governed, how conflicting information production activities are prioritised, or how the integrity of the AIM is maintained when dozens of appointed parties are simultaneously aggregating information into it.

The Infrastructure and Project Authority’s Transforming Infrastructure Performance programme has recognised that the governance of information in complex, multi-party programme environments requires specific design: it cannot be derived simply by applying a single-project process model at scale. The revised standard should have addressed this directly. It does not.

Alliance contracting and the three-party model

The standard’s fundamental organising structure – appointing party, lead appointed party, appointed party – is a conventional procurement hierarchy. The Major Projects Association (MPA) has consistently highlighted that major programme delivery is increasingly structured around alliance contracts, integrated project teams and joint ventures, in which the distinction between appointing party and appointed party is deliberately attenuated through shared risk and reward.

The NEC4 Alliance Contract, used on an expanding range of major infrastructure programmes, creates a multi-party environment that the standard’s three-party model simply cannot describe. This was a gap in the 2018 edition. It remains a gap in 2026. The MPA’s work on collaborative contracting models and the Infrastructure Client Group’s Project 13 enterprise model both offer frameworks that could have informed a more nuanced treatment of multi-party governance. Neither is referenced.

Regulatory hold points and safety-critical programmes

For programmes in regulated sectors (nuclear new build, aviation, rail), information management is not merely a project management function: it is a regulatory compliance function. The Office for Nuclear Regulation’s (ONR) licence conditions, Network Rail’s engineering assurance standards and equivalent frameworks in other sectors impose mandatory hold points in the information production process at which regulatory review and approval must occur before work can proceed.

The revised standard’s acceptance workflow acknowledges in a footnote that “there may be external stakeholders, such as shareholders or investors, that form part of the appointing party’s acceptance process”. This is a significant understatement of the regulatory reality on nuclear programmes, where the ONR is not an optional external stakeholder: it is a statutory consultee whose acceptance of safety-case-relevant information models is a legal prerequisite for continuing work.

This is not a minor implementation detail: it is a structural incompatibility that regulated programme teams will need to resolve locally, generating the inconsistency and risk that the standard is meant to prevent.

The mobilisation gap

Step 7 – mobilisation – is the shortest substantive step in the document: three activities across two pages. For major programme practitioners, this is the step’s most significant flaw because mobilisation is where information management failures most commonly originate. The Get It Right Initiative’s analysis of root causes of information quality failure consistently identifies inadequate mobilisation, particularly insufficiently tested CDE configurations, undertrained information producers and untested cross-organisation information exchange workflows, as a primary driver of downstream rework.

On a major programme, CDE mobilisation involves configuring access rights for potentially hundreds of users across dozens of organisations, establishing metadata schemas that will govern information retrieval for decades, testing status code workflows across multiple integrated platforms, and validating cross-CDE information exchange protocols that will carry safety-critical data. This is not a three-activity process. The absorption of Part 3’s more detailed mobilisation guidance into the new Part 2 has, in this area at least, produced a leaner document at the cost of essential operational detail.

Project close-out and the information handover problem

Step 9 – project completion – consists of two activities: archive information models and publish lessons learned. It makes no provision for validating that the AIM is complete against the original information requirements; no provision for governing the transfer of information between the construction-phase AIM and the operational AIM; no provision for the formal handover of intellectual property rights in information containers; and no structured framework for the lessons-learned process.

The IPA has identified poor information handover at project close-out as a consistent source of operational risk on major assets. When the people who built an asset cannot demonstrate to the people who will operate it that the AIM reflects the as-built reality, and when the information that should have been produced during construction is absent, incomplete or unstructured, the operational consequences can be severe. A standard that aspires to whole-lifecycle information management should treat project close-out with the same rigour it applies to project initiation.

Risks and gaps

No bridge to project controls standards; integrated project information model unaddressed.
Strictly sequential nine-step model cannot accommodate concurrent major programme projects.
No guidance on alliance or JV governance; three-party model is insufficient.
Regulatory hold points absent from the acceptance workflow.
Federated CDE governance treated as a technology problem, not a process problem.
Mobilisation step underdeveloped relative to its risk significance.
AIM completeness validation absent from project close-out.
‘Shall consider/record if not taken forward’ mechanism creates audit risk on regulated programmes.
Absorption of Part 3 guidance leaves an implementation vacuum.