Design to Cost · Evaluate
Cost Risk Assessment
List what could blow the target: commodity swings, tolerances, volumes — and price each risk with a contingency.
- Time45 min
- FormatSmall group
- StageEvaluate
Cost Risk Assessment: what it is and why it works
Cost Risk Assessment lists the events and uncertainties that could push the cost above target, estimates the probability and cost impact of each, and converts them into contingency and mitigation actions. Typical risks in design-to-cost work include commodity price swings, exchange rates, tolerance or yield problems, volume shortfalls, single-source dependencies and late design changes. Each risk gets a price, an owner and a mitigation. The register is small enough to review at every gate, and its total exposure explains how much contingency the target needs and why.
The method works because a target without a risk allowance is a best case disguised as a plan. Pricing each risk forces precision: a vague worry becomes a number that can be compared, mitigated or accepted. For simple registers, expected value (probability times impact) sums well enough; for larger projects, a Monte Carlo simulation on the estimate gives a distribution and lets the team choose a confidence level, such as P50 or P80, as AACE practice recommends. Cost Risk Assessment connects to the Cost Window, which sets the acceptable range, to Estimate Confidence Grading, since C-grade lines carry estimate uncertainty, and to Target Attainment Review, where the register is revisited.
What you need
- Current cost estimate with its main assumptions and confidence grades
- Commodity, currency and index exposures by line
- Supplier situation: single sources, capacity, financial health
- Volume forecast and its plausible range
- Project schedule and gate dates
What you get
- A cost risk register with probability, impact and expected exposure per risk
- Contingency per risk and in total, with the chosen confidence level
- Named owners and mitigation actions for the top risks
- Triggers for releasing or increasing contingency
- A review schedule tied to project gates
When to use it
When the target has no margin for reality and the surprises come anyway.
How to do it, step by step
- List the cost risks: commodity prices, tolerances, volumes, exchange rates, single sources.
- Estimate probability and cost impact for each.
- Compute expected exposure and set contingency per risk.
- Assign a mitigation owner to the top risks.
- Review the register at each gate.
Worked example: Risk-adjusting the cost of a stainless process skid package
Illustrative scenario — figures are realistic but not from a real company.
An engineering contractor was pricing a package of eight stainless-steel dosing and neutralization skids for a chemical plant expansion, estimated at $2.4 million with delivery in 14 months. The customer wanted a fixed price.
- The team listed nine risks. The largest were a nickel surcharge increase on 316L pipe and plate, a euro-denominated instrument package, a single-source analyzer supplier, and possible scope growth from late P&ID revisions.
- For the surcharge, they estimated a 40% chance of a $90,000 increase; for the currency, a 30% chance of $45,000; for the analyzer, a 15% chance of a $60,000 expedite and redesign; for P&ID growth, a 50% chance of $70,000.
- Expected exposure across all nine risks summed to about $128,000. A simple Monte Carlo on the ranges suggested roughly $190,000 to reach P80.
- Mitigations were assigned: early mill order for plate, a currency hedge by finance, a second qualified analyzer, and a P&ID freeze date written into the contract.
Result. The bid carried $190,000 of priced contingency, with the surcharge risk partly shifted through an indexation clause. At completion, the P&ID growth and currency risks occurred, the others did not, and the project finished within budget. The lesson was that mitigations, not just contingency, reduced the exposure.
Common pitfalls and how to avoid them
- Listing risks without prices, so the register becomes a worry list.Estimate probability and cost impact for every risk, even roughly, or remove it from the register.
- Adding worst cases together, which inflates contingency beyond credibility.Use expected values or a Monte Carlo simulation, and choose an explicit confidence level.
- Treating contingency as a slush fund that is quietly consumed.Link each contingency amount to named risks and release it formally when a risk is retired.
- Ignoring correlations, such as a commodity spike that hits several lines at once.Group correlated lines into one risk or model the correlation in the simulation.
Frequently asked questions
How do you calculate cost contingency?
For a simple register, multiply each risk's probability by its cost impact and sum the expected values. For larger estimates, run a Monte Carlo simulation using ranges for uncertain lines and discrete risk events, then set contingency as the difference between the base estimate and the chosen percentile, often P50 or P80. State the method and the confidence level so reviewers know what the contingency covers.
What is the difference between contingency and management reserve?
Contingency covers identified risks and estimate uncertainty within the defined scope, and is controlled by the project. Management reserve covers unknown unknowns and is typically held above the project by management. Keeping them separate prevents identified risks from quietly consuming the reserve meant for surprises.
What does P80 mean in cost estimating?
P80 is the cost value that the simulation says has an 80% probability of not being exceeded. A P50 estimate is equally likely to be overrun or underrun. Choosing P80 rather than P50 means carrying more contingency in exchange for a lower chance of overrun; the choice depends on the organization's risk appetite and contract type.
Origin
Risk-adjusted cost estimating — project risk management practice (PMBOK lineage); Monte Carlo in cost estimating (AACE).
Related methods
- Cost Window AllocationSplit the total target into subsystems with a cost window each — no subsystem eats the budget alone.
- Estimate Confidence GradingGrade every estimate A/B/C: quoted, analogous or guessed — and never mix them in one decision.
- Target Attainment ReviewAt project milestones, measure distance to target cost and decide: redesign, renegotiate, or accept with eyes…
More in “Evaluate”
Grade the estimates and price the risks before deciding.