A contingency reserve is the budget set aside for identified risks, things you know could go wrong but haven't happened yet. Known unknowns. On a £40M highways project, that might be 5% for weather delays, ground condition surprises, or utility diversions that the survey flagged as "possible." The contingency sits inside the Performance Measurement Baseline (PMB), which means it's part of the Budget at Completion (BAC) your earned value metrics are measured against.
This page is part of the earned value definitions glossary. If you're looking for how contingency fits into the broader budget structure, see the Contract Budget Base (CBB) page.
The Definition
Contingency Reserve = budget for identified risks (known unknowns), allocated within the PMBIt's not a slush fund. It's not padding. Every pound in the contingency reserve should trace back to a specific risk on the risk register, with a probability and a cost impact. If you can't point to a risk and say "that's what this £200K is for," it's not contingency. It's guesswork.
The formula is simple in theory:
Contingency Reserve = Sum of (Probability x Impact) for all identified risks
In practice, most commercial teams use a blend of quantitative risk analysis and professional judgement. Pure expected monetary value calculations give you a number. Experience tells you whether that number feels right.
Where Contingency Sits in the Budget Hierarchy
This is the distinction that trips up every graduate QS I've ever mentored. There are three levels of budget protection, and they're not the same thing:
┌──────────────────────────────────────────────────────┐ │ CONTRACT BUDGET BASE (CBB) │ │ │ │ ┌────────────────────────────────────────────────┐ │ │ │ PERFORMANCE MEASUREMENT BASELINE (PMB) │ │ │ │ │ │ │ │ ┌──────────────────────────────────────────┐ │ │ │ │ │ BASELINE BUDGET │ │ │ │ │ │ (Work packages, cost accounts, │ │ │ │ │ │ distributed across WBS) │ │ │ │ │ │ │ │ │ │ │ │ £38,000,000 │ │ │ │ │ └───────────────────────â–█â────────────────────┘ │ │ │ │ │ │ │ │ ┌──────────────────────────────────────────┐ │ │ │ │ │ CONTINGENCY RESERVE │ │ │ │ │ │ (Known unknowns – identified risks) │ │ │ │ │ │ │ │ │ │ │ │ £2,000,000 (5%) │ │ │ │ │ └──────────────────────────────────────────┘ │ │ │ │ │ │ │ │ PMB Total = £40,000,000 │ │ │ └────────────────────────────────────────────────┘ │ │ │ │ ┌────────────────────────────────────────────────┐ │ │ │ MANAGEMENT RESERVE (MR) │ │ │ │ (Unknown unknowns – unidentified risks) │ │ │ │ │ │ │ │ £2,500,000 │ │ │ └────────────────────────────────────────────────┘ │ │ │ │ CBB Total = £42,500,000 │ └──────────────────────────────────────────────────────┘
Contingency reserve = inside the PMB. Managed by the project team. Draws down as identified risks materialise or expire. Your CPI and SPI are calculated against a baseline that includes it.
Management reserve = outside the PMB but inside the CBB. Controlled by senior management. Used for risks nobody saw coming. Accessing it requires formal approval and changes BAC.
Profit margin = not shown here at all. Entirely separate from both reserves. Don't confuse them.
Contingency vs Management Reserve: The Differences That Matter
I see these two confused constantly. On one HS2 package I reviewed, the commercial team had lumped everything into a single "risk pot" with no distinction between identified and unidentified risks. Made the entire reserve meaningless because nobody knew what it was for.
| Aspect | Contingency Reserve | Management Reserve |
|---|---|---|
| What it covers | Identified risks (known unknowns) | Unidentified risks (unknown unknowns) |
| Inside PMB? | Yes | No |
| Inside CBB? | Yes | Yes |
| Who controls it | Project Manager / Commercial Manager | Senior management / Project Director |
| Affects BAC? | No (already in BAC) | Yes (accessing MR increases BAC) |
| Affects CPI/SPI? | Already baked in | Changes baseline when drawn |
| Example | Weather delays, utility clashes, design changes | Global pandemic, unprecedented material shortage |
| How it's sized | Expected monetary value of specific risks | Typically 3-10% of PMB, based on project complexity |
| Draw-down process | PM releases to control accounts as risks materialise | Formal change request to sponsor/director |
Sizing the Contingency: Not Just a Percentage
The lazy approach is slapping 5% on the budget and calling it contingency. That's not risk management. That's hope.
Proper contingency sizing starts with the risk register:
- Identify every quantifiable risk
- Estimate probability (%) and cost impact (£)
- Calculate expected monetary value (EMV = P x I) for each risk
- Sum the EMVs
- Sense-check against historical data from similar projects
For a more rigorous approach, use Monte Carlo simulation. Run 10,000 iterations with ranges for each risk's probability and impact. The P80 output (80th percentile) is a common contingency figure. It means you're 80% confident the contingency will be sufficient.
On UK construction, typical contingency percentages land between 3% and 8% of BAC, depending on project complexity:
| Project Type | Typical Contingency Range | Why |
|---|---|---|
| Repeat build (housing, standard warehouse) | 2-3% | Well-understood risks, limited uncertainty |
| Infrastructure (highways, rail) | 4-6% | Ground conditions, utilities, weather |
| Complex civil engineering (tunnels, marine) | 6-10% | High uncertainty, limited precedent |
| Refurbishment / brownfield | 5-8% | Hidden conditions, asbestos, structural unknowns |
Worked Example: Contingency on a Highways Project
Worked ExampleScenario: A Tier 1 contractor wins a £40M NEC4 Option C highways improvement scheme in the West Midlands. Programme runs January 2026 to March 2028 (27 months). During pre-construction, the commercial team builds the contingency reserve from the risk register.
Key identified risks:
| Risk ID | Description | Probability | Impact | EMV |
|---|---|---|---|---|
| R-003 | Uncharted utilities in central reservation | 60% | £800,000 | £480,000 |
| R-007 | Adverse weather (winter 2026/27 earthworks) | 70% | £450,000 | £315,000 |
| R-011 | Design development on retaining wall details | 40% | £600,000 | £240,000 |
| R-015 | Statutory undertaker delays (Western Power) | 50% | £700,000 | £350,000 |
| R-019 | Material price escalation (steel, concrete) | 30% | £500,000 | £150,000 |
| R-022 | Additional traffic management (revised TMP) | 45% | £400,000 | £180,000 |
| R-026 | Ground contamination (Phase 2 assessment pending) | 35% | £800,000 | £280,000 |
Total EMV = £1,995,000
Rounded to £2,000,000 (5% of the £40M target). The commercial team reviews this against benchmark data from three similar Highways England schemes and confirms 5% is reasonable.
Budget structure:
| Component | Amount | % of CBB |
|---|---|---|
| Distributed budget (work packages) | £38,000,000 | 89.4% |
| Contingency reserve | £2,000,000 | 4.7% |
| PMB (BAC) | £40,000,000 | 94.1% |
| Management reserve | £2,500,000 | 5.9% |
| CBB | £42,500,000 | 100% |
Month 8 update: Risk R-003 materialises. Uncharted BT duct discovered during excavation. Actual cost: £720,000. The commercial manager releases £720,000 from contingency to the affected control account. This is a transfer within the PMB, BAC doesn't change, CPI isn't distorted.
Meanwhile, risk R-019 (material escalation) is re-assessed as low probability. The £150,000 EMV is returned to the contingency pool.
Contingency status at month 8:
| Item | Amount |
|---|---|
| Opening contingency | £2,000,000 |
| Released to control accounts | -£720,000 |
| Returned from expired risk | +£150,000 |
| Remaining contingency | £1,430,000 |
On NEC4, the utility diversion also qualifies as a compensation event under clause 60.1(12) if the duct wasn't shown in the Site Information. If the CE is accepted and implemented at £720,000, the target increases by £720,000 and the contingency draw-down is effectively offset. That's the beauty of NEC4 target cost, the CE mechanism and the contingency reserve work in parallel.
Common Mistakes
1. No traceability to the risk register. If you can't map every pound of contingency to a specific identified risk, you don't have a contingency reserve. You have a guess. Auditors and client commercial teams will challenge unsubstantiated contingency during open-book reviews.
2. Confusing contingency with management reserve. Contingency is for risks you've identified. MR is for risks you haven't. Lumping them together means you don't know when you've exhausted your identified risk budget and need to escalate to senior management.
3. Never drawing down contingency. Some teams hoard contingency like it's their retirement fund. If a risk materialises, release the budget. That's what it's there for. Unreleased contingency sitting in the PMB while control accounts show overspend is bad EVM discipline.
4. Not updating the risk register. Contingency should be a living number. Risks expire. New risks emerge. If you set the contingency at month 0 and never revisit it, by month 12 you're carrying budget for risks that no longer exist and none for risks that have appeared since.
5. Using contingency to hide overruns. This is the one that gets commercial managers in trouble. If a control account is overspending because of poor productivity (not a risk event), don't paper over it with contingency. That masks the real CPI problem. Use contingency only for identified risk events.
How Gather helps. Gather's AI reads your site diaries daily and maps progress against your cost-loaded programme, giving you accurate earned value data without manual spreadsheet updates. Book a demo to see it working on a live NEC4 project.
Frequently Asked Questions
Is contingency reserve included in BAC?
Yes. Contingency reserve sits inside the Performance Measurement Baseline, which means it's part of BAC. When you calculate CPI (EV / AC), the BAC against which EV is measured includes the contingency allowance. Management reserve is the one that sits outside BAC.
How much contingency should I allow on a UK construction project?
It depends on the project's risk profile. A straightforward repeat-build project might need 2-3%. A complex infrastructure scheme with ground risk, utility interfaces, and winter working might need 6-8%. The right number comes from your quantified risk register, not from a percentage pulled from thin air. If pushed for a starting point, 5% is a defensible default for most mid-complexity infrastructure work, but always substantiate it.
What happens when contingency runs out?
If all contingency is consumed and another identified risk materialises, you have two options: absorb the cost within existing control accounts (squeezing other budgets) or request management reserve. Accessing MR increases BAC and requires senior management approval. It's a formal process, and it should be. It signals that the project's risk profile has exceeded expectations.
Can contingency be released at the end of the project?
Yes, and it should be. Unused contingency at project completion is effectively a budget saving. On NEC4 Option C, unused contingency within the target contributes to the Contractor's share under the pain/gain mechanism. If the total Defined Cost comes in below the target (which includes the contingency), the gain is shared between Contractor and Client per the share percentages in Contract Data.
.webp)
