What Makes Mega-Projects Fail — A Field Perspective from 17 Years of Project Controls
After a decade and a half on capital programs, the same five root causes appear again and again. None of them are technical.
Mega-projects fail for boringly predictable reasons. Cost overruns of 30–80% and schedule slippages above a year are not the exception — Bent Flyvbjerg's empirical work shows they are the statistical norm. After 17 years working as a planning and project controls engineer across substations, power plants, and rail, I have stopped looking for novel causes. The same patterns recur.
1. Optimism Bias at Sanction
The single largest predictor of overrun is the realism of the FID-stage estimate. Schedules and budgets that survive the gate by being plausible to executives are almost always biased toward best case. A rigorous independent estimate review and a P80 cost-risk analysis at sanction would eliminate a third of all major overruns.
2. Schedule Logic That Hides Risk
Mega-project baselines routinely accumulate hard constraints, lag links, and out-of-sequence relationships that obscure the true critical path. A DCMA 14-point check on the sanctioned baseline tells you the schedule's honesty in 30 minutes. Most owners do not run it.
3. Resource Estimates Decoupled from the Plan
The schedule says 80,000 craft hours; the cost plan says 105,000; procurement plans for 65,000. None of these reconcile. By the time the gap is discovered in execution, the recovery options are expensive.
4. Late Procurement, Long-Lead Surprises
Long-lead equipment — transformers, GIS, turbines, large bore pipe — drives the schedule far more than construction productivity does. Yet procurement activities are often the last to be properly logiced and the first to be compressed when sanctioned dates slip.
5. Weak Contemporaneous Records
When disputes arise — and on mega-projects they always do — the party with the better contemporaneous records wins. Daily reports, time-stamped photos, properly statused schedules, and notice letters are worth more than any retrospective expert report.
What Actually Helps
Independent estimate review. Reference-class forecasting. A schedule-and-cost risk analysis pair at every stage gate. A project controls function with a reporting line independent of project management. Owner engagement at the working level, not just the steering committee. None of this is new — which is the point.
Frequently Asked Questions
What is "What Makes Mega-Projects Fail — A Field Perspective from 17 Years of Project Controls" about?
After a decade and a half on capital programs, the same five root causes appear again and again. None of them are technical. It sits within the Research stream of PMMilestone Research & Insights and is written for practising project management and project controls professionals.Who is the intended audience for this Research article?
Planning engineers, project controls engineers, cost engineers, project managers and owner-side advisors working on capital construction, infrastructure and power projects. The article assumes working familiarity with CPM scheduling, EVM and risk management.Who authored this research article?
Dr. Hassan Eliwa, PhD — Founder of PMMilestone.org, senior planning and project controls engineer with 17+ years of international field experience and a PhD from Massey University (New Zealand).Which calculators and templates pair with this article?
The EVM Calculator, SPI Calculator, CPI Calculator and Schedule Health Checker on PMMilestone.org cover the formulas referenced here. Companion templates (risk register, EVM workbook) are linked from the relevant sections.What is a common misconception this Research article corrects?
That headline SPI and CPI numbers can be read at face value. On real research programmes, schedule performance must be assessed using Earned Schedule SPI(t) past ~70% progress, and CPI is only reliable after 15–20% physical progress. The article walks through how to apply these caveats in practice.Where can I find related research and definitions?
Use the PMMilestone Encyclopedia A–Z for canonical definitions of every term referenced in the article, and the Research Articles index for adjacent long-form pieces.What are Dr. Hassan Eliwa's research areas relevant to "What Makes Mega-Projects Fail — A Field Perspective from 17 Years of Project Controls"?
Dr. Hassan Eliwa's published research areas — owner-side project controls maturity, forensic delay analysis, earned schedule reliability on long-duration projects, and quantitative schedule-risk modelling — frame the analysis here. The full author profile and publication list live on the author page.