Sizing Contingency and Reserves: What Good Looks Like vs What Bad Looks Like
Every project budget should include provision for uncertainty. The question is not whether to include contingency, but how to size it correctly, at what level to hold it, and when and how to release it. In 2022, with inflation running at elevated levels in Canada and globally, with materials and labour costs volatile, and with supply chains still recovering from pandemic disruptions, contingency sizing is more consequential than in a typical year.
What Good Looks Like
Good: Contingency is sized using a defined methodology. For capital projects, the most defensible approaches are: probabilistic cost risk analysis (Monte Carlo simulation using cost estimate uncertainties), expected value of identified risks (probability times impact for each risk), or percentage of base estimate calibrated to project stage (higher percentages in early phases when uncertainty is greatest, declining as design and procurement progress). The methodology is documented and the basis is explained to the approving authority.
Good: Contingency and management reserve are separate and held at different levels. Contingency covers known risks and estimate uncertainty -- it is the project manager's to use within defined parameters. Management reserve covers unknown unknowns -- it is held by the sponsor or programme level and released by exception. Conflating the two leads to either inadequate provision for unknowns or management reserves used for routine cost growth.
Good: Contingency drawdown is tracked against the risk register. When contingency is used, the corresponding risk is marked as realised and the remaining contingency is adjusted. This creates a running picture of how the risk profile has evolved and whether remaining contingency is adequate for remaining risks.
Good: Contingency is re-estimated at each project gate. The appropriate percentage contingency at project sanction (final investment decision) is different from the appropriate contingency at project completion. As risks are retired and uncertainty is reduced, contingency should be reviewed and potentially returned to the programme budget.
What Bad Looks Like
Bad: Contingency is sized as a fixed percentage of base estimate without reference to the actual risk profile. A 10 percent contingency applied uniformly to all projects of all types at all stages is not a methodology -- it is a convention. It may be adequate for a low-risk project with a mature design and will almost certainly be inadequate for a complex project in early design with novel technology.
Bad: Contingency is treated as a slush fund. When budget pressures arise, cost items are charged to contingency without reference to the risk register or an assessment of whether the expenditure represents a risk realisation or a scope change. This erodes contingency without creating accountability.
Bad: There is no distinction between contingency and management reserve. Everything above the base estimate is pooled and called "contingency." The project manager and the sponsor do not have a shared understanding of what the provision is for or who controls it.
Bad: In 2022, contingency is not adjusted for inflation. A base estimate developed in 2021 with a 10 percent contingency may be materially under-reserved if materials and labour costs have risen 15 to 20 percent since the estimate was prepared. Current market conditions require an explicit inflation adjustment or re-estimation of the cost base.
XNM supports public-sector and capital-project organisations in project cost management and contingency analysis. Reach out to XNM's program & project delivery advisory team to discuss contingency sizing for your project or programme.