In September 2025, Saudi Arabia's Public Investment Fund suspended construction on The Line, NEOM's signature 170-kilometre linear city. By that point, only 2.4 kilometres of foundation work had been completed, representing just 1.5 percent of the planned structure. Fifty billion dollars had already been spent. The population target for 2030 was slashed from 1.5 million to fewer than 300,000. By March 2026, tunnelling contracts with Hyundai Engineering and structural steel contracts with Eversendai were cancelled outright, and the project began pivoting toward AI data centre infrastructure. What happened at NEOM is not just a story about ambition outpacing execution. It is a story about what happens when schedule risk is not modeled quantitatively before commitments are made.
Safran Risk for Saudi giga-projects is a quantitative schedule risk analysis tool that stress-tests project timelines using Monte Carlo simulation. It imports native Primavera P6 or Microsoft Project schedules, applies probability distributions and discrete risk events to activities, and produces S-curves showing the realistic range of completion dates at defined confidence levels. For Saudi Arabia's Vision 2030 mega-projects, it replaces optimistic single-point schedules with defensible, data-driven forecasts.
Had NEOM's planners run a rigorous QSRA in Safran Risk before committing to the 2030 timeline, the S-curve would have shown what the subsequent five years proved: the probability of completing The Line by 2030 was virtually zero. The tornado chart would have identified the top risk drivers: unprecedented construction methodology, workforce mobilisation at a scale never attempted in the region, and supply chain constraints for materials that did not yet have established production lines. That insight, delivered before the first dollar was committed, would have fundamentally changed the investment decision.
This article explains how Safran Risk models schedule uncertainty on Saudi giga-projects, using NEOM as the cautionary example and drawing practical lessons for every Vision 2030 programme still in delivery.
[Featured Image: S-curve overlay showing deterministic vs P50/P80 completion dates for a giga-project timeline]
What Safran Risk Would Have Shown NEOM's Planners
A properly calibrated Safran Risk model for The Line would have produced an S-curve with the deterministic completion date sitting somewhere between P1 and P5 on the cumulative distribution. That means the original 2030 target had between a 1 and 5 percent probability of being achieved. The P80 date, the confidence level IQRM recommends for major project planning, would likely have fallen in the 2045 to 2055 range given the scale of uncertainty involved. An internal audit leaked to the Wall Street Journal reportedly confirmed a projected completion timeline stretching to 2080, which aligns with what an extreme-tail Monte Carlo analysis would produce.
The value of this information is not that it predicts the future. The value is that it forces honest conversations about feasibility before capital is deployed. When the S-curve shows the target date at P3, the decision-maker has three options: accept the risk, change the scope, or change the timeline. What they cannot do is claim the schedule is realistic. For a deeper understanding of how P-values translate into planning decisions, see P50 vs P80 vs P90: How to Choose the Right Confidence Level for Your Project.
How Safran Risk Models Schedule Uncertainty on Giga-Scale Projects
Safran Risk is purpose-built for large, complex schedules with thousands of activities, multiple critical paths, and interdependent risk events. It handles the scale that spreadsheet-based Monte Carlo tools cannot. For Saudi giga-projects, the workflow follows a structured sequence that IQRM implements across its consulting engagements. For the complete seven-phase methodology, see Schedule Risk Analysis (QSRA): Guide to Monte Carlo + Examples.
Step 1: Import the Native Schedule
/h3>Safran Risk imports Primavera P6 XCR files or Microsoft Project MPP files directly. For a project like NEOM, the P6 schedule would contain tens of thousands of activities across multiple WBS levels. The import check is critical: verify that early start and finish dates in Safran Risk match the native P6 data. On giga-projects, discrepancies often emerge because the schedule was not recalculated (F9 in P6) before export, or because resource levelling was applied differently.
Step 2: Schedule Health Check
Before any risk variables are applied, the schedule itself must be dynamically responsive. Hard constraints lock dates and prevent the simulation from shifting them realistically. Excessive lags fixate activities. Open-ended logic breaks the critical path. On Saudi giga-projects, IQRM consistently finds that 15 to 30 percent of schedule warnings must be resolved before a meaningful QSRA can be run. Skipping this step produces simulation results that look sophisticated but are fundamentally unreliable.
Step 3: Define Uncertainties and Discrete Risk Events
Duration uncertainties (business-as-usual variation) are applied as continuous distributions on each activity. For NEOM, these would capture the inherent variability in foundation construction productivity, steel erection rates, and MEP installation durations. Discrete risk events capture specific threats: regulatory approval delays, supply chain failures for bespoke materials, workforce mobilisation shortfalls. Each discrete event has a probability of occurrence and an impact distribution.
Step 4: Apply Calendar Risks for Saudi Conditions
Saudi Arabia introduces calendar disruptions that do not exist on European or North American projects. Extreme summer heat above 45 degrees Celsius triggers outdoor work bans that can eliminate productive hours for weeks at a time. Ramadan reduces working hours across the entire workforce. Sandstorms halt operations without warning. These are modeled in Safran Risk as calendar risks using Poisson distributions for frequency, not as productivity adjustments. Modeling them as productivity would double-count their effect if duration uncertainties already capture performance variation.
Step 5: Run Monte Carlo Simulation
Safran Risk runs 5,000 to 10,000 iterations using Latin Hypercube Sampling for statistical efficiency. Each iteration generates a different combination of duration outcomes, risk event occurrences, and calendar disruptions. The result is a probability distribution of completion dates, presented as an S-curve. For a project the size of NEOM, the spread between P10 and P90 could easily exceed a decade, reflecting the extreme uncertainty inherent in first-of-a-kind construction at unprecedented scale.
NEOM: What Was Planned vs What QSRA Would Have Predicted
The following table illustrates the gap between NEOM's deterministic planning assumptions and what a rigorous QSRA would have shown for The Line component.
| Metric | Deterministic Plan | QSRA P50 | QSRA P80 |
|---|---|---|---|
| The Line completion | 2030 | 2042-2050 | 2055-2065 |
| Population at 2030 | 1,500,000 | 50,000-100,000 | Under 50,000 |
| Total Phase 1 cost | $370 billion | $500-700 billion | $700+ billion |
| Trojena completion | 2028 (Asian Winter Games) | 2031-2035 | 2035+ |
| Oxagon (green hydrogen) | Mid-2026 | Mid-2026 | Late 2026 |
The final row is instructive. Oxagon's green hydrogen plant, now at 80 percent completion, was the one NEOM component built on proven technology with established supply chains. Its QSRA profile would have been dramatically tighter than The Line's because the underlying uncertainties were smaller and better understood. This is exactly how Safran Risk differentiates: components with low uncertainty produce tight S-curves; components with extreme uncertainty produce wide ones. The tool does not manufacture optimism or pessimism. It reflects the data.
Why Saudi Giga-Projects Need Safran Risk Over Spreadsheet Models
Spreadsheet-based Monte Carlo tools work for projects with a few hundred activities. Saudi giga-projects operate at a different scale entirely. The Line's schedule would contain 50,000 or more activities across civil, structural, mechanical, electrical, architectural, and commissioning disciplines. Safran Risk handles this scale natively because it reads directly from P6 and preserves the full logic network, including calendars, resource assignments, and constraint types.
P80 Completion Date = Deterministic Date + Risk-Driven Buffer (from Safran Risk S-curve)
For NEOM: P80 ≈ 2030 + 25-35 years = 2055-2065
The correlation engine in Safran Risk is equally critical. On a giga-project, risks are not independent. If the concrete supply chain is stressed, it affects foundations, superstructure, and infrastructure simultaneously. If a single contractor underperforms, every package they hold is at risk. Without correlation, the model treats each risk as isolated and produces an unrealistically narrow S-curve that underestimates the true exposure. For the detailed Safran Risk workflow including import, calibration, and simulation settings, see Schedule Risk Analysis (QSRA) with Safran Risk in 2026.
Lessons for Vision 2030 Mega-Projects Still in Delivery
NEOM's experience offers direct lessons for every Vision 2030 programme that is still active: the Red Sea tourism development, AMAALA, Diriyah Gate, Riyadh Metro extensions, and the expanding ADNOC offshore programme across the Gulf. The lesson is not that ambitious projects should be avoided. The lesson is that ambitious projects demand more rigorous risk modeling, not less.
First, run QSRA before the Final Investment Decision, not after construction starts. The S-curve produced by Safran Risk gives decision-makers the information they need to set realistic timelines and budgets. If the P80 date is unacceptable, the scope should change before capital is committed.
Second, model calendar risks specific to Saudi Arabia. Heat restrictions, Ramadan hours, sandstorms, and regional holiday schedules are not generic disruptions; they are predictable, quantifiable events that Safran Risk can model using Poisson distributions based on historical frequency data.
Third, apply positive correlation to activities that share contractors, supply chains, or workforce pools. On Saudi giga-projects, the same handful of major contractors execute multiple packages. If one package slips, the contractor's resources are diverted, and related packages slip too. Ignoring this correlation produces a falsely narrow S-curve.
Fourth, update the QSRA monthly. A risk model is not a one-time deliverable. As construction progresses, actual performance data replaces expert estimates, and the S-curve tightens or shifts accordingly. Monthly updates keep the decision-making framework current and prevent the gap between the plan and reality from widening unnoticed.
Fifth, benchmark against similar completed projects using the Risk Data Engine methodology. IQRM's RDE framework structures historical data from past projects into fitted probability distributions that feed directly into Safran Risk. For Saudi giga-projects, the benchmarking pool should include comparable programmes from the GCC and global infrastructure sectors.
Frequently Asked Questions
What is Safran Risk used for on Saudi giga-projects?
Safran Risk is a quantitative schedule risk analysis tool that imports P6 or MSP schedules and runs Monte Carlo simulations to produce probabilistic completion dates. On Saudi giga-projects, it models duration uncertainties, discrete risk events, calendar disruptions, and correlations to generate defensible P50, P80, and P90 forecasts.
How would QSRA have helped NEOM avoid schedule failure?
A rigorous QSRA would have shown that the 2030 completion date for The Line sat at approximately P1 to P5 on the S-curve, meaning a 1 to 5 percent chance of achievement. This would have forced a scope or timeline adjustment before $50 billion was committed.
What calendar risks are unique to Saudi Arabia?
Saudi Arabia faces extreme summer heat (outdoor work bans when temperatures exceed 45°C), Ramadan-reduced working hours, sandstorms, and regional public holidays. These must be modeled as calendar risks in Safran Risk using Poisson frequency distributions, not as productivity adjustments.
Why is Safran Risk better than Excel for giga-project QSRA?
Excel-based Monte Carlo tools cannot handle schedules with tens of thousands of activities, complex logic networks, resource constraints, and calendar assignments. Safran Risk imports P6 natively, preserves the full logic network, supports Latin Hypercube Sampling for statistical efficiency, and includes built-in correlation and calendar risk engines.
How many Monte Carlo iterations should you run on a Saudi mega-project?
IQRM recommends 5,000 to 10,000 iterations for statistical validity on mega-projects. Safran Risk includes a convergence option that stops automatically once the mean and P80 stabilise within a 3 percent tolerance, ensuring reliable outputs without unnecessary computation time.
What confidence level should Saudi giga-projects plan to?
IQRM recommends P80 as the standard confidence level for project planning and contingency sizing on most projects. P80 provides an 80 percent probability of achievement. For client-facing commitments on high-stakes programmes, P90 may be appropriate. P50 is used for aggressive contractor targets.
IQRM delivers specialist training and consulting in Safran Risk, Monte Carlo simulation, and quantitative schedule risk analysis for mega-projects across Saudi Arabia and the GCC. Our QRM Diploma programme equips professionals with the practical skills to build, run, and interpret QSRA models on real giga-projects.
Want to apply quantitative schedule risk analysis to your Saudi giga-project? IQRM provides Safran Risk consulting, QSRA model builds, and risk workshop facilitation across the GCC.
Written by Rami Salem, Quantitative Risk Management specialist, 15+ years in oil & gas, EPC/EPCM, and infrastructure projects.
