Topic hub
Quantitative risk analysis (QRA) — practitioner resources
How quantitative risk analysis is actually run on UK programmes — workshop, calibration, Monte Carlo, P-level interpretation, and writing up a QRA report a board can act on.
About this topic
Quantitative risk analysis (QRA) is a Monte Carlo simulation of a project’s cost and schedule risks. It produces a confidence range — typically P50, P80 and P95 outcomes — rather than a single number. A QRA lets a board or sponsor decide how much contingency to approve at a defensible confidence level, and identifies the top risks driving that exposure.
SOMA delivers QRA to AACE International recommended practices (57R-09, 113R-20, 118R-21, 123R-22) using Safran Risk, Primavera Risk Analysis, @Risk or Acumen Risk. The discipline splits into Quantitative Schedule Risk Analysis (QSRA) — a Monte Carlo against the integrated schedule — and Quantitative Cost Risk Analysis (QCRA) — the same technique applied to the cost estimate. Integrated QCSRA combines the two so cost and schedule outcomes can be sampled together.
On UK programmes, QRA is most often run ahead of final investment decision (FID) or business-case approval, at major contract-award gates, and as an independent assurance check for sponsors. Done properly, it pulls together a credible risk register, calibrated three-point estimates, and a model whose correlation, distribution, and merge-bias treatment can be defended at gateway review. Done badly, it produces a number that nobody quite stands behind — which is no use to anyone trying to make a decision.
The guides below cover the practitioner mechanics: how to design and facilitate the workshop, how to calibrate three-point estimates so the tails are not artificially compressed, how to interpret confidence levels, when to integrate cost and schedule risk, and how to write up the report so the recipient can act on it. They are written for risk leads, controls managers and PMO directors on live UK programmes.
Guides on this topic
14 guides in this cluster
Guide
The Honest Guide to QRA
What Quantitative Risk Analysis actually is, when you need it, how it works, and how to tell a good one from a bad one.
10 min read
Guide
Monte Carlo Simulation Is Not Magic — What QRA Actually Does (and Doesn't Do)
What Monte Carlo simulation actually is in three sentences, what it does well in QRA, garbage-in-garbage-out, merge bias, correlation, and how to read the S-curve output for a board or finance committee.
9 min read
Guide
QSRA vs QCRA: Meaning, Methodology, and When Each Is the Right Answer
Two of the most important tools in quantitative risk analysis, frequently confused. Here is what each acronym means, how the methodologies differ, what each produces, and how to decide which your programme needs — with worked UK rail, water and nuclear examples.
8 min read
Guide
P50, P80, P95 in Cost Estimation: Which Confidence Level Should You Actually Use?
P50 is the IPA-required central estimate for UK capital cost. P80 is the UK departmental sensitivity convention. P95 is for safety-critical programmes and portfolio-level safeguards. How to pick the right confidence level for project sanction — and what HM Treasury Green Book and IPA Cost Estimating Guidance actually say, versus the working conventions departments use in practice.
9 min read
Guide
Capital Project Estimate Confidence Level — A Sponsor's Guide to P50, P80, P95 and IPA Cost Bands
For project sponsors, treasury teams, investment committees and capital approval boards. What "confidence level" actually means on a cost estimate, what the IPA Cost Estimating Requirements specify at each stage gate, when P80 is the right upper-bound sensitivity, and what to challenge in the QRA your project team has put in front of you.
12 min read
Guide
How Much Does a QRA Cost in the UK?
What actually drives the price of a Quantitative Risk Analysis, how QRA work is priced, and how to compare quotes without getting caught out.
8 min read
Guide
QRA Consultant vs In-House: How to Decide
When it makes sense to run a Quantitative Risk Analysis with your own team, when you need an independent consultant, and the hidden costs of getting that call wrong.
9 min read
Guide
Three-Point Estimate Calibration — What Low, Most-Likely, High Should Actually Mean
The difference between a QRA that works and a QRA that produces theatre usually comes down to how the three-point estimates were elicited. A practical guide to doing it properly — and to the PERT formula (O + 4M + P) / 6 and the (O + 3M + P) / 5 variant.
10 min read
Guide
Running a QRA Workshop That Actually Works
Most QRA workshops produce numbers nobody really stands behind. This is how to run one that produces inputs you can defend at a gateway review.
13 min read
Guide
How to Run a Pre-Mortem: The Risk Workshop That Finds What Brainstorming Misses
Gary Klein's pre-mortem technique, adapted for UK project controls — the facilitation script, the questions that work, and how to turn the output into a risk register that survives gateway review.
8 min read
Guide
QSRA Readiness — What a Schedule Needs Before the Monte Carlo Runs
Why DCMA 14-point, CIOB PP21 and Acumen Fuse SQI each answer a different question, what risk-load readiness actually means, and the specific failure modes that kill a QSRA before the simulation has a chance.
11 min read
Guide
Integrated Cost-Schedule Risk Analysis — When EMV Isn't Enough
Why running cost QRA and schedule QRA separately misses the real exposure, how integrated analysis actually works, and when to recommend it to a client.
11 min read
Guide
Writing a QRA Report That Survives Gateway Review
What makes the difference between a QRA report that gives an IPA reviewer confidence and one that triggers a recommendation to redo the work. Practical guide for UK public-sector programmes.
8 min read
Guide
Structuring a Risk Register So It's QRA-Ready From Day One
Most risk registers cannot be quantified without rebuilding them. This guide covers the structural choices that make a register QRA-compatible from the start — and the ones that force you to start over.
9 min read
Frequently asked
Quantitative risk analysis (QRA) — questions we get asked
- What is quantitative risk analysis (QRA)?
- QRA is a Monte Carlo simulation of a project’s cost and schedule risks. It produces a probabilistic confidence range — typically P50, P80 and P95 — rather than a single deterministic number. The output supports a defensible decision about how much contingency to hold and where the top exposures sit.
- What’s the difference between QSRA and QCRA?
- QSRA — Quantitative Schedule Risk Analysis — is a Monte Carlo against the integrated schedule, producing P-level dates. QCRA — Quantitative Cost Risk Analysis — is the same technique applied to the cost estimate, producing P-level totals. Integrated QCSRA combines the two so cost and schedule outcomes are sampled together rather than in separate models.
- Which QRA tools does SOMA use?
- SOMA is tool-agnostic. We deliver in Safran Risk, Primavera Risk Analysis, @Risk and Acumen Risk depending on the client’s estate and the model needs. The methodology — AACE 57R-09, 113R-20, 118R-21 and 123R-22 — is the constant; the tool is selected to fit.
- When in a programme should a QRA be run?
- Most commonly ahead of final investment decision or business-case approval, at major contract-award gates, and as a periodic independent assurance check during delivery. A one-off QRA run at sanction and never revisited is of limited use by the time the programme is under way; the most valuable QRAs are run live and updated as the risk profile changes.
- What is Monte Carlo simulation in project management?
- Monte Carlo simulation runs a programme’s cost or schedule model thousands of times, each iteration sampling random values from the probability distributions assigned to each risk and uncertain duration or cost. The result is a probability distribution of total outcomes — the basis for P50, P80 and P95 confidence levels. It does not predict the future; it characterises the range of plausible outcomes given the inputs.
- What do P50, P80 and P95 actually mean?
- P50 is the outcome that has a 50% probability of being met or bettered — the median of the simulation. P80 is the outcome that has an 80% probability of being met or bettered. P95 is the same at 95% confidence. UK Treasury Green Book guidance typically supports a P50 funding case with contingency held to P80; gateway reviews on infrastructure programmes often examine the P50–P80 range as the working envelope.
- Is Primavera Risk Analysis still supported?
- Oracle has placed Primavera Risk Analysis (formerly Pertmaster) on extended support — security fixes only, no new feature development. It is still in widespread use across UK infrastructure and defence programmes, and remains a defensible choice for QSRA work where the schedule is in P6. For new estates, Safran Risk, Acumen Risk and @Risk are the more actively developed alternatives. The methodology — AACE recommended practice — is what matters at gateway review; the tool is a delivery choice.
- Does a QRA need a fully integrated cost and schedule model?
- Not always. Many programmes start with a QSRA against the schedule and a separate QCRA against the cost estimate, then move to integrated QCSRA at major gates where the trade-off between cost and time contingency is a live decision. Integration adds modelling effort and demands a clean WBS/CBS alignment; the benefit is that cost-time trade-offs (acceleration, scope changes, mitigation buys) can be evaluated against a single sampled outcome rather than two disconnected models.
Got a programme where this matters?
Most engagements start with a short call. We work out whether we’re the right fit, then come back with a short scoping note — scope, duration, team, indicative cost.