Schedule Risk Analysis: How Schedule Uncertainty Drives Cost Contingency on Capital Projects
Cost contingency on capital projects is usually presented as the answer to a single question: how much extra money should we hold against unidentified risks and uncertainty? The honest answer cannot be given without first answering a different question — how much schedule do we expect to consume, and with what probability? On capital projects, time and cost are not independent variables. Every additional month of execution carries fixed indirect costs, escalation drift, prolonged contractor preliminaries, and exposure to weather windows or labour-market shifts. A cost contingency that ignores schedule uncertainty is structurally an underestimate.
Schedule risk analysis — usually called SRA — is the technique that quantifies this exposure. It produces a probabilistic distribution of project duration, identifies the activities most likely to drive that distribution, and converts time-based outcomes into cost outcomes that feed the contingency calculation. This article covers why schedule and cost risk cannot be modelled in isolation, how SRA is performed mechanically, how its output translates into cost contingency, and the common pitfalls that decouple the two.
Why Schedule and Cost Risk Cannot Be Modelled Separately
Capital projects accumulate cost in two distinct ways. Direct costs — equipment, materials, installation labour — scale with quantity and unit rate. Indirect costs — site supervision, equipment hire, project management, owner team, temporary works, insurance — scale with duration. A two-month delay in completion does not change the steel quantity, but it does extend every weekly indirect line for an additional eight weeks. On large capital projects, indirects routinely run between 12% and 25% of total project cost, which means schedule slippage feeds directly into a sizeable cost line that quantity-only contingency models miss entirely.
Time-related cost exposure compounds with escalation. On a five-year capital project, a six-month delay does not just extend indirects — it pushes residual scope into a later inflation regime. Steel and labour priced against a 2026 base date will be installed against 2027 conditions. Escalation indices that sit benignly in the assumed schedule become a meaningful risk variable when the schedule lengthens. Modelling cost risk without schedule risk treats escalation as deterministic when it is, in practice, conditional on duration.
There is a third, less mechanical reason. Schedule pressure changes contractor and owner behaviour in ways that affect cost. A project running late frequently absorbs unplanned acceleration costs — overtime premiums, out-of-sequence work, expedited fabrication, double-shifting — that do not appear in any baseline estimate but are a foreseeable consequence of duration risk. Capturing these effects requires a probabilistic view of duration, which is what schedule risk analysis exists to produce.
Modelling cost risk without schedule risk treats escalation as deterministic when it is, in practice, conditional on duration.
The Mechanics of a Schedule Risk Analysis
A schedule risk analysis runs Monte Carlo simulation against the project schedule with three-point duration ranges applied to activities — or activity groups, where the schedule is too detailed to range every line. Most projects also apply probabilistic branching for risk events with discrete probability — regulatory delay, geotechnical findings, vendor performance — which can either occur or not, with separate impact when they do. The mechanics break into three steps.
Range estimates on activity durations
Each activity in the analysis requires three duration values — minimum, most likely, and maximum — usually distributed using a triangular or PERT-beta function. The most common error is treating the deterministic schedule duration as the most likely value when it has typically been built from optimistic planning assumptions. A defensible most-likely duration in an SRA is usually 5–15% longer than the deterministic baseline, with maximum values that reflect the long tail of execution conditions actually observed on comparable projects.
Probabilistic logic and risk events
Beyond range distributions, an SRA incorporates discrete risk events with probability and conditional impact. A late regulatory approval may have a 30% chance of occurring with a two-month impact on a critical-path activity. The simulation combines these with the activity ranges to produce a probabilistic distribution of project completion dates that reflects both ordinary variability and named risks.
Criticality and sensitivity output
The simulation reports far more than a P50 or P80 completion date. Criticality indices identify which activities most frequently sit on the critical path across simulated scenarios, and sensitivity tornadoes rank the activities or risk events whose duration uncertainty most influences project completion variance. These outputs drive remediation: schedule contingency is most efficiently invested in compressing or de-risking the activities the tornado names, not in adding flat float to every line.
Translating Schedule Risk into Cost Exposure
The output of an SRA is a distribution of project completion dates. The challenge is converting that distribution into a cost distribution that feeds the contingency calculation. Three categories of conversion are typically applied; the result is added to the scope-driven cost contingency derived from cost-only Monte Carlo analysis.
Direct time-cost extension. Every additional day of execution multiplies a daily indirect cost rate. The daily rate combines site supervision, project management, owner team, equipment hire, accommodation, utilities, and any other line that scales with duration rather than quantity. On large projects this rate runs anywhere from $50,000 to $500,000 per day depending on size and location. A project whose P80 schedule outcome is 60 days longer than baseline is therefore 60 times the daily rate above baseline before any other cost effect is considered.
Escalation reprofiling. Scope completed later than baseline absorbs more inflation. This is calculated by re-applying the escalation curve to remaining scope under the simulated completion profile, then comparing integrated escalation cost against the deterministic baseline. On a five-year project at 4% annual escalation, a one-year delay typically adds 1–2% of total project cost in escalation alone, which is rarely captured in scope-only contingency.
Acceleration provisions. Where the SRA forecasts a late completion that the project will not accept, contingency must include the cost of compressing the residual schedule. The probability-weighted acceleration cost — overtime premiums, expedited fabrication, additional supervision — is added to cost contingency if recovery is the assumed response to slippage. Without this provision, the contingency presumes the project quietly absorbs the late completion, which is rarely how owners actually behave.
Common Pitfalls in Coupling SRA to Cost Contingency
Even on projects that run a credible schedule risk analysis, the link to cost contingency is often broken. Four patterns recur.
- Treating SRA and cost risk analysis as parallel exercises. If the SRA is run by the planner and the cost risk model is run by the cost engineer with no shared inputs, the cost model’s duration assumption frequently does not match the SRA’s P-level outcome. The cost contingency is then either silently double-counting time risk or silently ignoring it.
- Using deterministic durations as most-likely inputs. Setting the most-likely activity duration equal to the planning baseline collapses the SRA to the same answer as the planning schedule. The simulation produces a tight distribution centred on the baseline date, missing the asymmetric upside risk that real capital projects carry.
- Excluding soft activities from the schedule. Many baseline schedules do not include owner approvals, permitting windows, or vendor fabrication lead times — duration-sensitive activities that are not part of construction logic. SRA performed on a construction-only schedule misses some of the most consequential slippage drivers.
- Ignoring recovery behaviour. Once a project is identifiably late, contractors and owners typically invest in acceleration. If the cost contingency assumes the schedule simply runs out, it omits what is often the most likely actual cost outcome — recovery cost — and tells the project nothing about how much acceleration headroom is funded.
Key Takeaways
- Schedule risk and cost risk are not independent on capital projects — duration drives indirect costs, escalation reprofiling, and acceleration provisions that cost-only contingency models cannot capture.
- A schedule risk analysis combines three-point activity durations, discrete risk events, and Monte Carlo simulation to produce a probabilistic distribution of project completion dates.
- Translate SRA output into cost terms via three conversions: daily indirect-cost extension, escalation reprofiling on residual scope, and probability-weighted acceleration provision.
- Sum the time-driven contingency with the scope-driven contingency carefully — double-counting and silent omission are the most common errors when the two analyses are run in isolation.
- Use SRA criticality indices and sensitivity tornadoes to prioritise the activities where additional float, schedule contingency, and de-risking effort yield the highest return.