Safran Risk for Sellafield Nuclear Decommissioning: UK Schedule Risk Analysis

Nuclear decommissioning programmes operate in a risk environment unlike any other sector. At Sellafield, Europe's largest and most complex nuclear site, the schedule for retrieving legacy waste, demolishing redundant facilities, and remediating contaminated land spans decades, involves technologies that have never been deployed at scale, and must comply with regulatory frameworks where safety takes absolute precedence over programme efficiency. A single unforeseen radiological condition can halt operations for months.

Safran Risk is a specialist schedule risk analysis software that integrates directly with Primavera P6 and Microsoft Project to perform Monte Carlo simulation on project schedules. It enables risk analysts to model duration uncertainty, discrete risk events, and correlation between activities, producing probabilistic forecasts expressed as confidence levels (P50, P80, P90) that support decisions on schedule contingency, resource planning, and regulatory milestone commitments.

For Sellafield Ltd and the Nuclear Decommissioning Authority (NDA), Safran Risk provides the analytical capability to move beyond deterministic programme targets and understand the true range of probable completion dates for each major decommissioning workstream, giving leadership the evidence base to set realistic milestones, allocate contingency where it is most needed, and defend programme timelines to regulators and government stakeholders.

Here is how Safran Risk applies to the Sellafield decommissioning programme, and why it has become the tool of choice for nuclear schedule risk analysis in the UK.

Why Sellafield Requires Specialist Schedule Risk Analysis

Sellafield's decommissioning portfolio includes some of the most technically challenging projects ever undertaken in the nuclear industry. The legacy ponds and silos, built in the 1950s and 1960s, contain decades of accumulated nuclear waste in facilities that were never designed for retrieval operations. The Pile Fuel Cladding Silo (PFCS), Magnox Swarf Storage Silo (MSSS), and First Generation Magnox Storage Pond (FGMSP) each present unique engineering challenges where construction methodologies must be developed, tested, and approved by the Office for Nuclear Regulation (ONR) before deployment.

The schedule risk profile of these projects is fundamentally different from conventional construction. Duration uncertainty is driven by radiological characterisation findings that can change the retrieval approach mid-programme, equipment reliability in harsh radiological environments where maintenance access is severely limited, regulatory approval timelines that depend on safety case development and review cycles, and the availability of specialist workforce with nuclear security clearances. Historical data from nuclear decommissioning programmes worldwide shows that schedule overruns of 100-300% against original deterministic estimates are common, making probabilistic forecasting not merely useful but essential for credible programme management.

How Safran Risk Works for Nuclear Decommissioning

Safran Risk connects directly to the Primavera P6 schedule that governs Sellafield's decommissioning programme, importing the full CPM logic network including activity durations, relationships, constraints, and resource assignments. This native integration means the risk model always reflects the current approved baseline schedule, eliminating the version control issues that plague spreadsheet-based risk analysis approaches.

Duration Uncertainty Modelling

Every critical and near-critical activity in the decommissioning schedule receives a three-point duration range reflecting inherent variability. For waste retrieval operations in the legacy silos, the optimistic duration might assume equipment operates at design throughput with minimal downtime, the most likely duration reflects realistic operating conditions with planned maintenance windows, and the pessimistic duration accounts for equipment failures, unexpected waste forms, and radiological surprises that require methodology changes. Safran Risk applies BetaPERT or triangular distributions to these ranges, with the distribution shape selected based on data maturity and the nature of the uncertainty.

Discrete Risk Event Modelling

Beyond duration variability, Safran Risk models discrete risk events that may or may not occur. These include discovery of previously uncharacterised waste inventory requiring a revised safety case, failure of bespoke retrieval equipment requiring redesign and remanufacture, regulatory holds pending resolution of safety case queries, and supply chain delays for specialist nuclear-grade components with limited global manufacturing capacity. Each risk receives a probability of occurrence and an impact distribution, and Safran Risk maps these to the specific activities they would affect if they materialise.

Correlation and Systemic Risk

Nuclear decommissioning programmes exhibit strong correlation patterns. If retrieval equipment underperforms in one silo, similar equipment in adjacent facilities is likely to face comparable issues. If regulatory approval timelines extend for one safety case, the precedent affects all subsequent submissions. Safran Risk allows analysts to define correlation coefficients between related activities and risk events, ensuring the simulation produces realistic output distributions that reflect these systemic dependencies rather than artificially narrow ranges.

QSRA Outputs for the Sellafield Programme

Running Safran Risk on the Sellafield decommissioning schedule produces a comprehensive set of outputs that support programme governance at every level.

The wide spread between P50 and P90 reflects the genuine uncertainty inherent in first-of-a-kind nuclear decommissioning operations. This is not a modelling artefact; it reflects the reality that retrieval technologies have never been deployed at this scale, waste inventories are partially characterised, and regulatory pathways remain subject to evolving safety standards.

Key Risk Drivers from the Tornado Analysis

Why Safran Risk is the Preferred Tool for Nuclear QSRA

Native P6 integration eliminates model drift. Safran Risk reads directly from Primavera P6, ensuring the risk model always reflects the current approved baseline. In a programme where schedule updates occur monthly and change control is rigorous, this eliminates the risk of analysing an outdated schedule version.

Correlation modelling produces realistic outputs. Nuclear decommissioning programmes exhibit strong systemic risks that affect multiple workstreams simultaneously. Safran Risk's correlation engine ensures these dependencies are captured, preventing the dangerously optimistic results that uncorrelated models produce.

Audit-trail transparency satisfies regulatory scrutiny. Every assumption, distribution, risk mapping, and correlation coefficient in Safran Risk is documented and traceable. When the ONR or NDA queries a confidence level or contingency recommendation, the complete analytical basis can be demonstrated.

Scenario comparison supports decision-making. Safran Risk allows analysts to run multiple scenarios (with and without specific mitigations, accelerated vs. standard sequences) and compare the resulting S-curves side by side. This enables leadership to quantify the schedule benefit of investing in risk mitigation before committing resources.

Scalability handles programme-level complexity. Sellafield's decommissioning schedule contains thousands of activities across multiple interconnected projects. Safran Risk handles this scale efficiently, running 10,000 Monte Carlo iterations on schedules with 5,000+ activities within minutes.

Schedule Risk Governance at Sellafield

QSRA using Safran Risk is not a one-time exercise at Sellafield. It forms part of the ongoing programme governance framework, with quarterly risk model updates that incorporate actual progress, newly identified risks, retired risks, and updated characterisation data. This living risk model provides the NDA with a continuously updated view of programme confidence that informs funding decisions, regulatory reporting, and strategic planning.

The outputs feed directly into Sellafield Ltd's Performance Plan submissions to the NDA, where probabilistic completion forecasts and risk-based contingency allocations are required. Without Safran Risk's analytical rigour, these submissions would rely on deterministic estimates that historical evidence consistently shows to be unreliable for nuclear decommissioning programmes.

Safran Risk transforms Sellafield's decommissioning schedule from a planning tool into a decision support instrument, giving the NDA, government, and regulators the quantified confidence they need to oversee a programme that will define UK nuclear decommissioning capability for generations.

Frequently Asked Questions

What is Safran Risk?

Safran Risk is a schedule and cost risk analysis software that performs Monte Carlo simulation on project schedules imported from Primavera P6 or Microsoft Project. It models duration uncertainty, discrete risk events, and correlation to produce probabilistic forecasts at defined confidence levels (P50, P80, P90).

Why is Safran Risk preferred for nuclear projects?

Nuclear projects require auditable, transparent risk analysis that can withstand regulatory scrutiny. Safran Risk's native P6 integration, comprehensive audit trail, and correlation modelling capabilities make it the standard tool for UK nuclear programme risk analysis, used across Sellafield, Hinkley Point C, and other major nuclear programmes.

How does Safran Risk handle correlation in nuclear decommissioning?

Safran Risk applies correlation coefficients between related activities and risk events to ensure the simulation captures systemic dependencies. For nuclear decommissioning, this means recognising that equipment reliability issues, regulatory delays, and workforce constraints tend to affect multiple projects simultaneously rather than independently.

What confidence level does the NDA require for programme milestones?

The NDA typically requires P80 confidence levels for major programme milestones in Sellafield's Performance Plan. This means there should be an 80% probability that the milestone will be achieved by the stated date. Some high-profile commitments may require P90 confidence to provide additional assurance to government stakeholders.

Can Safran Risk model multiple decommissioning projects simultaneously?

Yes, Safran Risk can model programme-level schedules containing multiple interconnected projects. This is essential at Sellafield where the legacy ponds, silos, and new build facilities share resources, access routes, and regulatory approval pathways. Programme-level modelling captures these inter-project dependencies that individual project analyses would miss.

How often should the Safran Risk model be updated?

Quarterly updates are standard practice for active nuclear decommissioning programmes, with full model refreshes at major programme gates. Monthly updates may be appropriate during critical phases such as active retrieval operations or regulatory approval periods. Each update incorporates actual progress, new risk information, and updated characterisation data.