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, 64R-11, 65R-11, 66R-11) 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 is, when you need it, how it works, and how to tell a good one from a bad one.

14 min read

Guide

Monte Carlo Simulation Is Not Magic - What QRA Actually Does (and Doesn't Do)

What Monte Carlo simulation is, what it does well in QRA, why inputs decide the answer, how merge bias and correlation work, and how to read the S-curve for a board.

14 min read

Guide

QSRA vs QCRA: What Each Does and When You Need Each

QSRA models the schedule and QCRA models the cost. This guide sets out how each works, what each produces, how they combine, and which one answers your decision, with three short illustrative examples.

10 min read

Guide

P50, P80, P95 in Cost Estimation: Which Confidence Level Should You Actually Use?

What the percentiles mean, how far apart they sit and why, and what the Green Book 2026 and the NISTA (formerly the IPA) cost guidance require, as at October 2026.

14 min read

Guide

Capital Project Estimate Confidence Level: A Sponsor's Guide to P50, P80 and P95

For project sponsors, investment committees and capital approval boards. What a confidence level means on a cost estimate, how it relates to the IPA target ranges and the Green Book, and what to challenge in the QRA put in front of you.

13 min read

Guide

How Much Does a QRA Cost in the UK?

Indicative effort by scope tier, what drives the price, how QRA work is priced, and how to compare quotes.

8 min read

Guide

QRA Consultant vs In-House: How to Decide

When to run quantitative risk analysis with your own team, when to commission an independent analyst, and how to cost the choice.

8 min read

Guide

Three-Point Estimate Calibration: What Low, Most-Likely and High Should Mean

How to define, elicit and test the three-point inputs behind a QRA, and what the PERT formula (O + 4M + P) / 6 and its (O + 3M + P) / 5 variant actually do.

12 min read

Guide

Running a QRA Workshop That Actually Works

Most QRA workshops produce numbers nobody 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: a timed agenda, the wording to use, and how to turn the output into register entries a QRA can use.

11 min read

Guide

QSRA Readiness: What a Schedule Needs Before the Monte Carlo Runs

Why a DCMA pass does not make a schedule ready to be risk-loaded, which schedule and register checks matter, and how they fit the QRA Readiness Checklist.

11 min read

Guide

Integrated Cost-Schedule Risk Analysis: When EMV Isn't Enough

What expected value misses, how an integrated cost and schedule simulation works, how to read the joint result, and when the build is worth it.

10 min read

Guide

Writing a QRA Report That Survives Gateway Review

What separates a QRA report that gives a NISTA Gateway Review team confidence from one that prompts a recommendation to redo the work. A 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 being rebuilt. These are the structural choices that let a register feed a quantitative risk analysis directly, and the ones that force a restart.

12 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, 64R-11, 65R-11 and 66R-11 — 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. The Green Book (2026 edition) requires optimism-bias and risk adjustment but does not mandate P80; practice varies between a P50 funding case with contingency held to P80 and other confidence levels; 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.