Infrastructure owners make an early decision that shapes almost everything that follows: how much responsibility should remain inside the institution, and how much should move to the market?

A conventional design-bid-build structure keeps substantial design responsibility with the owner before construction is tendered. Design-build moves more coordination into the contractor's organisation. Early contractor involvement brings suppliers into the project while important design and sequencing choices are still open. A public-private partnership can extend private responsibility into financing, construction, operation and maintenance for many years.

These models are usually compared through cost, programme, competition, financing, risk allocation and the type of asset being procured. There is another variable inside the choice that deserves the same attention: each model leaves the owner with a different set of responsibilities, and those responsibilities require a particular kind of organisation to carry them.

Transferring a larger volume of work can narrow the owner's workload while increasing the consequence of what remains. A design-build contractor may control hundreds of detailed design choices, while the owner retains responsibility for defining the required performance, judging material departures from it and deciding whether changes still protect the purpose of the asset. A long-term concession may place construction and operations with a private counterparty, while the public institution carries decisions on performance, affordability, contract change and public accountability across several changes of management.

The delivery model therefore creates an organisational question alongside the contractual one.

What kind of owner does this project require?

That question becomes harder to answer once a project moves beyond the documents that first made it look investable.

What appraisal leaves unresolved

The histories of two Nigerian power projects provide a useful starting point.

When the World Bank appraised the Azura Edo Independent Power Project and the Qua Iboe Independent Power Project as part of its Nigeria Power Sector Guarantees programme, their projected financial returns were remarkably close. The appraisal estimated a post-tax financial internal rate of return of about 13.5 per cent for Azura and 13.8 per cent for Qua Iboe. [1]

Both projects could therefore be represented through recognisable investment variables: construction costs, tariffs, expected output, financing structures, fuel arrangements and forecast returns. This is one of the purposes of appraisal. Governments, lenders and investors need a disciplined way to compare expected economics before committing capital and guarantees.

Their subsequent histories were very different. Azura reached financial close in December 2015 and moved through construction into operation. Qua Iboe did not reach financial close, and the unused guarantee associated with the project was later withdrawn. [2]

The comparison cannot establish a simple explanation for why one project progressed while the other did not. The public record contains too many changing conditions for that, and reducing Qua Iboe's outcome to one institutional weakness would go beyond the evidence.

What the comparison does expose is the boundary of appraisal.

Once implementation begins, assumptions become obligations held by different organisations. Gas has to be supplied. Transmission capacity has to be available. Government commitments have to be performed. Financing conditions must continue to be satisfied. Contractors, lenders, operators and public counterparties need decisions from one another at particular points in the programme.

A financial model can describe assumptions about those relationships. It cannot coordinate them when one changes.

That work continues inside the project after approval, and much of it eventually arrives back at the organisation acting for the owner. Someone has to recognise when a change in one obligation affects another, assemble the evidence needed to respond and route the issue to a person with authority to decide.

The institutional question therefore begins where the model stops.

The delivery model changes the work left with the owner

The OECD's current work on public infrastructure buyer maturity approaches the issue directly. Its 2026 programme is examining how the effectiveness of design-bid-build, design-build and early contractor involvement varies with the capability and organisational design of the public buyer using them. [3]

This matters because a delivery model does more than distribute tasks between organisations. It changes the kind of judgement the owner must retain.

Under design-bid-build, the owner carries substantial responsibility for developing and coordinating the design before construction is competitively tendered. Its ability to define scope, coordinate technical information and prepare sufficiently complete tender documents affects what the contractor is eventually being asked to price.

Design-build redistributes much of that coordination. The owner then relies more heavily on performance requirements, technical evaluation and its ability to judge whether a contractor's proposed solution satisfies the outcome originally required. Poorly defined requirements can give the contractor greater freedom while leaving the owner less able to determine whether the eventual solution serves the purpose for which the asset was commissioned.

Early contractor involvement changes the timing again. The owner receives contractor knowledge while design decisions are still developing, which can improve constructability and sequencing. It also requires enough technical and commercial confidence inside the client organisation to evaluate advice from a party that may later benefit from how those decisions are resolved.

PPP structures carry the organisational problem across a much longer period. The public institution can transfer financing, construction and operations while continuing to manage service obligations, contract performance, variations and circumstances that may change over decades.

Across these models, the retained work is different in volume and character. A smaller number of owner decisions may sit closer to the project's purpose, economics and long-term performance than the larger number of activities transferred outside the institution.

That is why delivery-model choice and buyer capability belong in the same discussion.

Formal architecture still has to become a decision

Nigeria already recognises many of these responsibilities in law and institutional guidance.

The Public Procurement Act assigns overall responsibility for procurement within a federal procuring entity to its accounting officer. Procurement planning is intended to bring together procurement, the user department, finance, planning and statistics, relevant technical personnel and legal expertise. [4]

Nigeria's PPP framework similarly provides for public-sector project teams, project officers, transaction advisers, risk assessment and contract-management responsibilities. ICRC guidance also identifies recurring weaknesses such as inadequate project ownership, the absence of an empowered public-sector champion, unrealistic development timetables and insufficient stakeholder engagement. [5]

The existence of those structures is important. It also allows the capability question to be framed more precisely.

A project can have the right professional functions involved and still struggle to reach a conclusion that crosses the boundaries between them.

Consider a proposed design change during construction. The engineer may establish that the revised solution is technically workable. The commercial team may determine that it changes price. The programme team may identify an effect on sequencing. Legal may need to determine how the contract treats the change, while finance or senior management controls the authority required to commit additional expenditure.

The project does not experience those reviews as separate organisational achievements. It experiences the date on which an authorised decision finally emerges.

This difference becomes important because infrastructure decisions regularly sit between professional disciplines. Technical questions can have commercial consequences. Commercial decisions can affect programme. Programme changes can alter financing and supplier commitments. Each function therefore sees a legitimate part of the problem while the project still needs someone to hold the combined consequence.

The procurement architecture can identify who should participate. The operating organisation determines whether that participation can become a decision.

I have encountered the same organisational threshold from the contractor's side.

At Dutum, periods of growth have shown that technical capability and institutional capability do not develop automatically at the same pace. In one period of live delivery, additional work increased demands on equipment and production capacity while commercial exposure remained after physical execution. The management system around the work had to develop as the operating load increased. [6]

That experience has shaped our investment in performance management, quality systems, digital visibility, knowledge sharing, document management and governance. These systems perform different tasks, but each reduces the amount of organisational performance that depends on one person carrying information, judgement or follow-up in his head. [6]

The relevance for an infrastructure owner is straightforward. Complexity places demands on the organisation carrying the work, regardless of which side of the contract that organisation occupies.

A sophisticated procurement structure therefore needs an owner organisation capable of operating at the sophistication the contract assumes.

Follow a transferred risk

Risk allocation is one place where the quality of that judgement becomes visible.

Suppose a project contains uncertainty around ground conditions. The contract can assign responsibility for those conditions to the contractor, and the allocation can be legally clear. The contractor still needs to decide what that uncertainty is worth before submitting a price.

Where the available information is reliable, bidders can estimate the exposure with greater confidence. Where information is incomplete, bidders have several responses available: include contingency, qualify the offer, seek exceptions, price conservatively or decide that the opportunity does not justify the uncertainty.

The contractual allocation therefore interacts with the information available to the market and the ability of the party receiving the risk to control its outcome.

The same mechanism appears with access, inflation, utility relocation, demand assumptions and design completeness. The owner may prefer a particular risk to sit with the contractor, yet the eventual economic effect depends on how clearly that risk can be understood and managed when the tender is priced.

ICRC's PPP guidance treats risk identification, valuation and allocation as connected activities. OECD's STEPS methodology similarly begins with the characteristics of the project, the structure of the relevant supply market and the buyer's own technical and institutional capabilities before reaching decisions about packaging and contracting. [5][7]

These approaches place a demanding responsibility on the buyer.

Someone inside the owner's decision structure must understand enough about the risk to assess the information available, the party able to influence the outcome and the likely market response to the proposed allocation. Without that judgement, the project can transfer responsibility formally while remaining uncertain about the price it is paying to do so.

The contractor's bid will eventually reveal part of that answer.

Several bidders may include similar contingency. Some may qualify the same condition. Competition may narrow because particular firms decide they cannot price the exposure responsibly. Each response gives the owner information about what the market thinks the project is asking it to carry.

A capable buyer can use those signals to reconsider the allocation, provide better information or decide that the premium is justified by the benefit of transfer. That decision requires an organisation able to interpret what the commercial response is telling it about the project.

Risk transfer therefore depends on buyer judgement before it becomes contract language.

Follow an interface

The same need appears when responsibility is already divided across organisations.

Infrastructure projects are full of boundaries. One organisation controls an asset, another controls access, another grants an approval, another supplies a utility and a contractor sequences its work around commitments from several of them.

The Azura project illustrates the density of those relationships. The generating plant depended on arrangements involving gas supply, transmission, a power purchaser, government support, lenders, sponsors, contractors and other public institutions. [1] The contracts could identify responsibilities across those parties, while project performance still depended on their obligations connecting in practice.

An infrastructure owner therefore needs visibility over the interfaces that sit between formal responsibilities.

Take a simpler construction example. A contractor submits a technical detail that must be approved before a particular workfront can proceed. The design team reviews the submission, but the proposed solution affects an external utility, so another organisation must also respond.

The project report may correctly record the status as “awaiting approval.”

The useful management question is what sits behind that status.

The supplier supporting the work may have priced materials for a limited period. A specialist subcontractor may have reserved a mobilisation window. Labour and equipment may have been sequenced around the expected approval date. The technical question is therefore connected to commitments that begin moving while the approval remains open.

At some point the issue needs to leave routine review and enter a different decision route.

The project team needs to know which evidence is still missing, who owns the external interface, who can accept any resulting cost or programme consequence and where the decision escalates if the original authority cannot conclude it.

That is enough for P3. The detailed economics of what happens after the decision remains open belong in a separate analysis of decision latency. The important point here is organisational: an interface spread across several competent parties still needs an owner capable of seeing the combined consequence.

This is where buyer capability becomes visible to the contractor.

A contractor does not experience the client's organisational structure through the names of its committees. It experiences whether access arrives, whether information can be approved, whether an instruction comes from someone authorised to issue it and whether an unresolved matter can reach the right authority while the programme can still respond efficiently.

Eventually, the internal design of the owner organisation appears outside that organisation as commercial behaviour.

What has to remain after the advisers leave

Infrastructure buyers also rely on specialist advisers because the range of expertise required by a major transaction is wider than most institutions should permanently maintain in-house.

Financial advisers can model a project's economics. Lawyers can structure contracts. Engineers can test technical assumptions. Environmental and social specialists can examine obligations that require deep expertise. The public project team benefits from being able to assemble those capabilities around a transaction when required.

The organisational question returns when that transaction ends.

ICRC's PPP Manual makes the public project team responsible for managing transaction advisers and provides for the transfer of project documentation generated during their work. [5] Documentation is an essential starting point, although the value of the record depends on what has actually been preserved.

A final report can record the recommendation without capturing the reasoning that gave the recommendation its boundaries.

Years later, another team may need to know why one risk was retained while another was transferred, which technical assumption caused the greatest concern, why a bidder qualification was accepted or which alternative commercial structure was examined and rejected. These questions matter when circumstances change and the institution has to determine whether an earlier decision still applies.

The useful record therefore connects important decisions with their evidence and context.

It should be possible to recover the issue under consideration, the alternatives examined, the assumptions relied upon, the authority that concluded the matter and the consequence the decision was intended to control. That information allows a future team to reopen an assumption without accidentally reopening every settled issue around it.

Document management becomes organisational memory when it preserves enough reasoning for future judgement.

The same principle applies beyond PPP transactions. Staff move. Consultants finish assignments. political leadership changes. Contractors hand over assets. A project with a twenty-year economic life will almost certainly outlast many of the people who made its earliest decisions.

The owner institution remains.

Its ability to retain judgement across those changes is part of the capability the original delivery model required.

Design the buyer before the tender

The practical implication is that delivery strategy should include an organisational test.

Before approving a procurement model, the owner should be able to describe the important responsibilities that remain inside its organisation and show how those responsibilities will be carried during preparation, procurement, delivery and operation.

The exercise does not require another large procedural manual. It requires specific answers.

Proposed project condition Capability the owner should be able to identify
Design or performance responsibility will move to a contractor Who defines the required outcome and who can judge a material departure from it
A material risk will be transferred What evidence supports the allocation, who can influence the risk and how the market is likely to price it
Delivery depends on another institution or external system Who owns the interface, what decision is required and where the issue escalates
Changes can affect scope, cost or programme Which authority level can accept each consequence
Specialist advisers hold significant project knowledge What expertise, evidence and decision record remain inside the institution after their assignment
The contract continues into long-term operation Who will manage performance, change, institutional memory and future disputes

The most useful findings will often appear where the table cannot be completed.

If a project plans to transfer a technical risk while nobody within the owner's decision structure can explain the exposure being transferred, the gap should affect the procurement strategy. If several departments participate in a material interface but no person can conclude the decision, the project has an organisational dependency that has not yet been resolved.

The response may differ from project to project. An institution can build permanent capability where the need will recur, procure specialist support where the expertise is episodic, alter delegated authorities, change how work is packaged or select a delivery model that better matches the organisation available to govern it.

Those choices belong inside project preparation because the market will eventually price the consequences of getting them wrong.

Nigeria already has procurement legislation, regulatory institutions, PPP frameworks, transaction advisers and a substantial history of delivering complex infrastructure. The next question is increasingly about what happens inside the organisations that use those tools when responsibilities cross professional, contractual and institutional boundaries.

A delivery model can allocate design, construction, financing and operations across the market. It will still leave the owner with decisions that require technical understanding, commercial judgement, authority, coordination and memory.

That remaining work has to live somewhere.

Before approving how an infrastructure project will be delivered, the owner should therefore be able to answer one additional question with the same seriousness given to cost, financing and risk:

What must our organisation be capable of doing for this delivery model to work?

Where the answer exposes a capability that does not yet exist, building the buyer is part of preparing the project.

References

  1. World Bank, Nigeria Power Sector Guarantees Project: Project Appraisal Document, 2014.

  2. World Bank, Nigeria Power Sector Guarantees Project: Implementation Completion and Results Report, 2020.

  3. OECD, Public Infrastructure Buyer Maturity and Effective Use of Delivery Models, 31 March 2026.

  4. Federal Republic of Nigeria, Public Procurement Act 2007.

  5. Infrastructure Concession Regulatory Commission, PPP Manual for Nigeria and related project-development guidance.

  6. Temitope Runsewe, From Legacy to Leadership, LSDPC Strategy Retreat, September 2026.

  7. OECD, Support Tool for Effective Procurement Strategies (STEPS).