| Vose Software

Industry: Engineering
Product: Tamara
Application: Mitigating Budget Overruns in Infrastructure Projects


The Base Estimate Was $60M. The Simulation Says the Project Needs a $30M Contingency to Reach P80

An engineering firm published a deterministic estimate for an urban highway-expansion project: an eleven-activity most-likely roll-up summing to a $60M base cost. It was the number that went to the client and into the firm's margin forecast. It was also the answer to a question no project sponsor should ask — what does this cost if every activity lands on its most-likely value and no risk materialises? On a job riddled with utility relocation and unknown ground conditions, that is the one outcome that essentially never happens.

Rebuilt in Tamara, Vose Software's Monte Carlo project risk tool, with Beta-PERT durations and costs on every activity and six discrete risk events layered on top, the deterministic cost becomes a distribution. The cost chart is the headline of this case:

Project cost distribution before and after mitigation against the budget

The distribution carries a mean of $83M, a P90 of $94M, and a P80 of $90M. The gap between the published $60M base and the P80 cost is $30M — that is the contingency the project actually requires, and it was invisible to the deterministic sum. Against a working budget of $88M, the unmitigated project carries a 26% probability of overrun; a targeted mitigation package — advance utility investigation, fixed-price material contracts and a geotechnical/ground-conditions allowance — pulls the green distribution left to a $77M mean and an $86M P90 and cuts that to 6%.

Why the deterministic budget fails

A summed most-likely estimate misses two things at once. First, activity costs and durations are right-skewed — each can come in a little under but can overrun a lot, so the mean cost ($83M) sits well above the $60M roll-up. Second, discrete risks compound: utility relocation overruns, unforeseen ground conditions and contractor claims each carry their own cost tail, and the budget that ignores them is not conservative — it is simply wrong on the high side of the odds.

The discrete risks that drive the overrun

The six discrete events were modelled as Bernoulli risks — each may or may not occur, but if it does it adds delay and cost. Ranking them by expected schedule impact (probability × delay) gives a clean Pareto, and these same events carry the heavy cost tails:

Pareto of discrete risk events by expected schedule impact

Four of the six events carry ~80% of the expected discrete-event delay — a utility-relocation overrun (4.7 weeks expected), unforeseen site/ground conditions (3.8), a contractor claim/rework (2.8) and weather lost days (2.7). Material price escalation carries no schedule impact — it is a pure cost risk, and the Pareto isolates it correctly. These are the shortlist the mitigation package targets, because they dominate both the cost and the schedule tail.

Cost and schedule overrun together

The schedule tail is not a separate problem from the budget; through extended general-conditions and traffic-management overhead it is part of the budget. Tamara plots finish day against total cost, one cell per simulated outcome:

Joint density of finish day versus total cost, split into quadrants by the budget and the plan

The cloud runs diagonally because every day past the deterministic plan adds overhead. 26% of runs are both late and over budget — the budget overrun and the schedule overrun are not separate risks to manage, which is why Tamara models the two together.

Where the duration risk lives

Tamara reports each activity's cruciality — the rank-correlation between its duration and the project finish, i.e. how much its variability moves the end date.

Schedule tornado ranking activities by correlation with the completion date

Utility relocation dominates (cruciality 0.54), followed by earthworks and drainage (0.50) and the bridge and structures (0.41). Utility relocation is both the top schedule driver and a leading cost driver — exactly the kind of dual-impact activity that a deterministic plan buries and a probabilistic one surfaces.

The schedule behind the cost

The same simulation that produces the cost distribution also produces the schedule:

Stochastic S-curve of project completion date versus the deterministic plan

The deterministic 705-day plan has only a 2% chance of being met; the P50 is day 868 and the P90 day 984, with a P80 finish 239 days — about 7.8 months — beyond the plan. That schedule tail feeds straight back into the budget through extended overhead, and mitigation trims the P80 finish by 55 days (944 → 889 days) along with the cost.

What Tamara changed

  • The published budget moved from the $60M deterministic base to a risk-informed contingency, exposing the $30M gap to P80 before it became an overrun claim.
  • The overrun probability was made explicit (26%), turning "are we covered?" into a number the sponsor could act on.
  • A mitigation package was approved on its joint cost-and-schedule effect, cutting overrun probability from 26% to 6%.
  • Risk-reduction effort was concentrated on utility relocation and ground conditions — the dual-impact drivers — instead of spread evenly across the bar chart.

Tamara Functionality Used

  • Monte Carlo simulation of cost and schedule over the full activity network, with Beta-PERT durations and costs.
  • Cumulative cost distribution and P-value contingency analysis quantifying the gap between the deterministic base and the P80 budget.
  • Discrete risk-event modelling (Bernoulli occurrence × Triangular impact) on durations and costs, including a pure cost-only escalation risk.
  • Joint cost–schedule density quantifying the probability of being late and over budget together.
  • Criticality and cruciality analysis identifying the dual cost-and-schedule drivers.
  • Scenario comparison quantifying a mitigation package's before/after impact on the overrun probability and the P80 finish.

A budget is not a single number; it is a distribution with a base nobody hits and a contingency the project genuinely needs. Tamara is what turns "will we overrun?" into a probability the firm and its client can both plan against.