Part V — Procurement, Contracting & Programme Delivery · Chapters 13–14
Building It Right
and Governing
What You Build
Two chapters on the contract models that determine who bears risk in construction, the governance frameworks that determine whether programmes deliver what they promised, and what happens when both go wrong simultaneously.
Parts I through IV built the foundations, the financial toolkit, and the risk and regulatory framework. Part V addresses the moment where strategy meets execution — the procurement decisions that determine how infrastructure is contracted and built, and the governance frameworks that determine whether programmes deliver what they promised at the cost and time they projected.
Chapter 13 develops the procurement strategy framework: the eight contract models that govern how infrastructure is procured, the five-question decision process that determines which model fits each situation, the NEC contract family in detail, and the project alliancing model that represents the most collaborative — and most demanding — approach available. The Snowy 2.0 case study shows what happens when the wrong contract model is chosen: an EPC fixed-price structure applied to geologically uncertain underground works, producing a contractual dispute that more than doubled the cost of Australia’s most ambitious pumped hydro project.
Chapter 14 addresses performance management and programme governance: the Earned Value Management toolkit, the IPA Gateway Review process, the benefits realisation framework, and the five governance failure modes that recur across major infrastructure programmes. The HS2 case study — scope inflation, optimism bias, leadership discontinuity, and the cost trajectory from £17 billion at approval to over £54 billion by 2023 — is the most extensively documented governance failure in recent UK infrastructure history, and one of the most instructive.
Together, the two chapters make a paired argument: the contract model determines the incentives that govern how contractors behave; the governance architecture determines the incentives that govern how clients behave. Both must be right for a major infrastructure programme to succeed.
Chapters 13 and 14 · Snowy 2.0 (procurement mismatch) · HS2 (governance failure) · Eight contract models · NEC Options A–F · Project alliancing · EVM toolkit · IPA Gateway Reviews · Benefits realisation
“A fixed-price contract that transfers risk to a contractor who cannot manage it is not a risk transfer. It is a risk deferral — to the moment the contractor becomes insolvent or submits claims that exceed the contract value.”
Contract model → contractor incentives. Governance architecture → client incentives. Both must be right. Most major programme failures trace to one or both being wrong.
Procurement Strategy & Contract Models
From adversarial to collaborative: selecting the contract that delivers long-run value
Chapter 13 opens with the most important insight in procurement strategy: the choice of contract model is not a commercial decision — it is a governance decision. The model selected determines what the contractor is financially motivated to do over the life of the contract, how risks are managed when they materialise, and what the relationship between client and contractor will look like when things go wrong. These consequences persist for years or decades and are largely irreversible once the contract is signed.
The five-question procurement decision framework drives model selection: How well-defined is the scope? The more uncertain the scope, the less suitable fixed-price models are. How technically novel is the work? Novel work requires contractor input into design; models that exclude the contractor from design produce worse outcomes. How urgent is programme delivery? Design-and-build models are faster; traditional sequential models are slower but allow more client control. How competitive is the market? Few qualified contractors reduce the benefits of competitive tendering. Who should bear long-term maintenance responsibility? Bundled models align design and maintenance incentives; traditional models separate them.
The chapter then develops the NEC contract family in depth — the six options from fixed price (Option A) through management contract (Option F) — and the project alliancing model, which takes collaborative contracting to its logical conclusion by creating a single commercial framework in which all parties share project outcomes rather than allocating risk through contract clauses. The conditions under which alliancing outperforms traditional contracting — genuine scope complexity, long duration, high relationship investment — are also the conditions under which it is most demanding for the client.
A fixed-price contract transfers risk to the contractor on paper. But risk cannot be eliminated by contract — it can only be allocated. When a contractor cannot manage the risk they have accepted, it returns to the client through claims, delay, or insolvency.Chapter 13 — Procurement Strategy & Contract Models
The chapter closes with contractor selection and supply chain resilience — arguing that the consistent evaluation of quality over price produces better long-run outcomes than the industry norm of 60–80% price weighting. An abnormally low tender is not a procurement success; it is a risk deferred to the construction phase, when the contractor discovers the gap between their price and their actual costs.
Contract Model Landscape: Four Key Types
Contractor: construction
Established methodology. Client controls design; contractor builds to specification. Sequential — design complete before tendering. Best for well-defined scope with strong client design team.
Open-book cost management
Target agreed at contract award; actual cost compared to target; pain/gain shared per agreed ratio. Most widely used model for complex UK infrastructure. Requires genuine collaborative intent and client NEC capability.
Maximum risk transfer on paper. Fixed price, schedule, and performance warranty. Requires very well-defined technical scope — when scope is uncertain, EPC becomes a claims factory.
Unified commercial framework, open-book accounting, no-blame culture, integrated team from project inception. Highest governance overhead. Requires genuine client technical capability — cannot be delegated.
NEC Contract Family: Six Options
Lump sum per completed activity. Contractor bears resource and cost risk within each activity. Provides programme certainty for the client.
→ Well-defined scope · client wants programme certaintyRates × measured quantities; remeasured on completion. Contractor bears rate risk; client shares volume risk. Common for civil engineering with uncertain quantities.
→ Volume uncertainty · stable scopeTarget cost + shared pain/gain on the difference. Open-book cost management. Most widely used for major UK infrastructure. Requires genuine collaborative culture.
→ Complex/uncertain scope · dominant UK modelTarget cost + shared pain/gain; bill of quantities used for target setting. As Option C but with volume uncertainty layered on top.
→ Both scope and volume uncertaintyAll defined costs reimbursed plus fee. Contractor bears no cost risk. Client bears maximum exposure. No incentive to minimise cost without additional performance measures.
→ Emergency works · very uncertain scopeManagement contractor paid fee; all substantive works subcontracted. Client retains risk of subcontract performance. Appropriate for very large programmes needing management coordination.
→ Large programmes · management coordination- EPC fixed-price contract selected for the main tunnel and cavern works — designed to transfer cost and programme risk to the contractor
- Premise: scope defined with sufficient certainty to support a credible fixed price at tender
- Problem: deep alpine tunnelling geology cannot be fully characterised by any pre-contract investigation
- Geological conditions encountered — water inflows, fault zones, swelling ground — differed materially from design assumptions
- Original 2025 completion was politically driven; reference class forecasting would have predicted 2028–2030 for comparable projects
- Ground conditions dispute: contractor claimed compensable “ground risk event”; Snowy Hydro maintained risk was priced. Contractual dispute dominated the delivery programme.
- Cost more than doubled: AUD 5.1bn at approval → AUD 12bn+ revised estimate
- Programme slipped from 2025 to 2029, with continuing uncertainty
- Better model: target cost or alliance with explicit ground condition risk sharing. Geological uncertainty is genuinely incompatible with fixed-price EPC — the risk transfer is theoretical.
- Energy storage rationale remains sound: 350,000 MWh may still justify AUD 12bn for NEM transition — the financial and strategic value analyses must be conducted separately
EPC is inappropriate for genuinely uncertain scope
Fixed-price contracting requires scope certainty to support a credible price. For geotechnically uncertain underground works this condition is not met — risk transfer becomes theoretical.
Reference class forecasting must discipline programme ambitions
The original 2025 completion was politically driven, not empirically grounded. Reference class forecasting of comparable alpine tunnelling projects would have produced a 2028–2030 programme at approval.
Tailored ground risk allocation outperforms binary models
The binary choice — all to contractor (EPC) or all to client (cost-plus) — is rarely optimal. Construction methodology risk to contractor; residual geological uncertainty shared: the appropriate allocation for underground works.
Strategic and financial value must be analysed separately
At AUD 12bn the financial return is materially weaker than at AUD 5bn. The strategic value of 350,000 MWh of storage for the NEM may still justify the investment — but the two analyses are distinct and must be presented separately.
Performance Management & Governance
Measuring what matters, governing what is difficult, and realising what was promised
Chapter 14 makes the governance argument that underpins the entire book: the best investment appraisal, delivered through poor governance, will produce a worse outcome than a more ordinary appraisal delivered through excellent governance. Governance is the performance multiplier. And the IPA’s evidence on major projects consistently shows that governance quality — measured through programme board composition, SRO continuity, reporting discipline, and benefits realisation oversight — is among the most powerful predictors of programme outcome.
The Earned Value Management toolkit provides the quantitative foundation for programme performance reporting. The three primary data points — Planned Value (what should have been done), Earned Value (what has been done, measured in budget terms), and Actual Cost (what has been spent) — generate a set of derived metrics that give the programme board an integrated picture of schedule performance, cost performance, and projected final cost that no single metric can provide. The most important: Estimate at Completion (EAC = BAC/CPI) — the projection of total programme cost if current cost efficiency continues. An EAC that diverges from the Budget at Completion early in a programme is the most reliable single indicator that the programme is in difficulty.
The benefits realisation section argues that the weakest point in most infrastructure governance frameworks is the connection between approved business case benefits and measured outturn benefits. Benefits that are not measured cannot be verified; benefits that are not verified cannot inform the next appraisal; appraisals that are not informed by outturn performance cannot be calibrated. This feedback loop — from approved business case to benefits realisation to appraisal methodology improvement — is the mechanism by which the infrastructure profession learns from its forecasting errors. Without it, optimism bias reproduces itself indefinitely.
The best investment appraisal, delivered through poor governance, will produce a worse outcome than a more ordinary appraisal delivered through excellent governance. Governance is the performance multiplier.Chapter 14 — Performance Management & Governance
The five governance failure modes — optimism cascade, accountability diffusion, reporting capture, decision avoidance, and benefit disconnection — close the analytical section of the chapter. Each is documented in the HS2 record; together they constitute a near-complete taxonomy of how major infrastructure programme governance fails. The chapter does not present HS2 as uniquely badly governed — it presents it as unusually well-documented, and argues that most large programmes share these failure modes to varying degrees.
Earned Value Management: The Seven Core Metrics
Budgeted cost of work scheduled to be complete by today. Read from the approved baseline S-curve.
Baseline S-curveBudgeted cost of work actually performed. Progress measured in budget-dollars, not real spend.
% complete × BACActual cost incurred for work performed to date. What has been spent from the financial ledger.
Financial ledgerSchedule performance index. >1.0 = ahead; <1.0 = behind. SPI 0.85 means 85% of planned progress achieved.
EV ÷ PVCost performance index. >1.0 = under budget per unit of work; <1.0 = over budget. Most important single metric.
EV ÷ ACProjected total cost if current CPI continues. THE governance metric. Early divergence from BAC is the primary warning signal.
BAC ÷ CPICPI required on remaining work to finish within BAC. >1.2 = effectively impossible without a programme reset.
(BAC − EV) ÷ (BAC − AC)IPA Gateway Review: Six Stages
Five Governance Failure Modes
Each successive cost estimate anchors on the previous one; upward revisions are communicated in smaller increments to minimise political shock.
→ Antidote: mandatory reference class forecasting at approval; independent cost reviewMultiple sponsors, multiple governance layers, unclear decision rights — nobody is clearly accountable for the programme outcome as a whole.
→ Antidote: named SRO with clear authority; SRO tenure incentivised; board composition enforcedReporting systems are controlled by programme teams who have an incentive to present progress favourably; challenging data is smoothed or delayed.
→ Antidote: independent assurance; EVM with external validation; whistleblower protectionsDifficult decisions — scope reductions, budget resets, contractor changes — are deferred because their short-term political cost exceeds the long-term financial benefit.
→ Antidote: IPA gateway authority; pre-defined decision triggers; external review at hard thresholdsBusiness case benefits are never measured at outturn; the feedback loop from approved forecast to actual performance is never closed.
→ Antidote: mandatory Gate 5 reviews; benefits realisation plan before approval; independent measurement- £17.4 billion at 2011 Phase 1 approval in 2011 prices — already optimistic relative to reference class for urban rail tunnelling
- 750+ scope changes from outline to detailed design: each individually justified, collectively transforming the programme scope without commensurate cost discipline
- Multiple CEO, Finance Director, and Board Chair changes created repeated institutional knowledge loss — SRO continuity was not protected as a programme asset
- Revised estimate reached £45–54 billion (2019 prices) by 2019 — the Oakervee Review identified systemic governance weaknesses as the primary driver
- All five governance failure modes were present and documented: optimism cascade (sequential small revisions), accountability diffusion (unclear sponsor hierarchy), reporting transparency concerns (internal estimates more pessimistic than public disclosures), decision avoidance (failure to reset scope), benefit disconnection (no outturn BCR tracking)
- Phase 2 cancellation (October 2023): announced at a party conference without full BCR analysis for Phase 1-only configuration — the governance failure extended to the cancellation decision itself
- Governance reforms 2020–2023: new CEO, restructured board, enhanced HM Treasury oversight, revised programme — IPA rating improved from red to amber-red
- Lesson: not a technical failure. No engineering challenge on HS2 was beyond British capability. Every major cost driver was a governance failure.
Reference class forecasting must discipline original estimates
The gap from £17bn to £54bn is consistent with Flyvbjerg’s reference class for urban rail. The approval estimate was optimistic by design. Mandatory reference class adjustment would have produced a more defensible and more accurate starting point.
Scope change control is the most important single governance mechanism
The accumulation of 750+ scope changes — each individually justified, collectively transforming the programme — drove more cost growth than any other single factor. A robust scope change control process with cumulative cost impact assessment is the primary antidote.
SRO continuity is a programme asset that must be protected
Each leadership transition cost institutional knowledge, stakeholder relationships, and commercial authority with the contractor. The governance framework should create incentives for longer SRO tenure — not merely respond to departures.
Reporting transparency builds the political resilience programmes need
The perception that HS2’s estimates were not fully transparent damaged political sustainability as much as the cost growth itself. Honest range disclosure, delivered consistently from the outset, builds more durable political support than optimistic point estimates that are later revised upward.
What Part V Establishes — and the Paired Argument It Makes
Part V’s two chapters make a paired argument: the contract model determines the incentives that govern how contractors behave; the governance architecture determines the incentives that govern how clients behave. Both must be right for a major infrastructure programme to succeed.
Snowy 2.0 illustrates what happens when the contract model is wrong: the right incentive (cost certainty via fixed price) was applied in the wrong context (geologically uncertain underground works), producing theoretical risk transfer that converted in practice into a contractual dispute. HS2 illustrates what happens when the governance architecture is wrong: all five failure modes — optimism cascade, accountability diffusion, reporting capture, decision avoidance, benefit disconnection — operated simultaneously and in reinforcing combination.
The lesson that connects them: the technical and financial quality of an infrastructure programme is determined much earlier in the procurement and governance cycle than most organisations appreciate. By the time a project is under construction, most of the important governance decisions have already been made — for better or worse. The infrastructure strategist’s highest-leverage interventions are the ones made before contracts are signed and before governance frameworks are established.
The procurement decision framework and contract model landscape — from traditional DBB through project alliancing — that matches model to context, scope certainty, and risk profile.
The EVM performance toolkit, IPA Gateway process, benefits realisation framework, and five governance failure mode taxonomy that determines whether programmes deliver what they promised.
That procurement and governance decisions, made before shovels break ground, determine programme outcomes more powerfully than any subsequent management intervention.
The BCR approved in Chapter 8’s investment appraisal is only realised if the programme is delivered through the procurement model of Chapter 13 and the governance of Chapter 14. Analysis and delivery are inseparable.
Digital twins and data-driven asset management · ESG, sustainability and climate adaptation · The future of infrastructure investment — Virtual Singapore, Netherlands Delta Programme, G20 Global Infrastructure Hub
“The infrastructure strategist’s highest-leverage interventions are the ones made before contracts are signed and before governance frameworks are established.”
Snowy 2.0: AUD 5.1bn → AUD 12bn+ · HS2 Phase 1: £17.4bn → £45–54bn · 750+ scope changes on HS2 · NEC Option C dominant model for major UK infrastructure · TCPI >1.2 = near-impossible without reset
Continue to Part VI
Digital twins and data-driven asset management, ESG and climate adaptation, and the future of infrastructure investment.