Curtailment Risk Modelling
Executive Summary
Key Takeaways
- ✓ Curtailment risk modelling should be built from historical curtailment data at the specific grid node or zone the project connects to, not a generic industry-average curtailment percentage.
- ✓ Because grid congestion frequently worsens as more generation connects to the same constrained network over time, curtailment risk should be modelled as a forward-looking trend, not a static rate held constant across the asset's full operating life.
- ✓ The applicable compensation mechanism (compensated or uncompensated curtailment) should be modelled precisely against the project's actual PPA, regulatory framework, or interconnection agreement, since this materially changes the revenue consequence of the same physical curtailment level.
- ✓ Curtailment risk should be tested through dedicated sensitivity analysis reflecting a worse-than-historical congestion scenario, given the trend risk of increasing curtailment in many constrained network zones.
- ✓ Curtailment risk modelling should be built as an extension of, not a replacement for, the basic curtailment mechanic already covered in the model's technical output chain.
Objective¶
This guide covers the modelling methodology for forward-looking curtailment risk within Energy Financial Modelling, extending the basic curtailment mechanic already covered in the technical output chain with a full sourcing, forecasting, and sensitivity methodology.
Sourcing from Node- or Zone-Specific Historical Data¶
Curtailment risk should be modelled using historical curtailment data specific to the grid node or zone the project actually connects to, not a generic industry-average curtailment percentage. Curtailment is driven by local transmission capacity relative to the concentration of generation connected to that same network segment, meaning risk varies materially between a well-connected node and a congested one — a generic average obscures this location-specific reality.
Modelling Curtailment as a Trend, Not a Static Rate¶
Grid congestion frequently worsens over the course of an asset's operating life as more generation connects to the same constrained network, particularly in zones experiencing rapid renewable buildout relative to available transmission capacity. A static curtailment rate held constant across the full modelled operating life can understate curtailment risk in later years relative to the level observed at the time the model was originally built. Curtailment risk should instead be modelled as a forward-looking trend, informed by known or planned network upgrades and the pipeline of other generation projects expected to connect to the same constrained zone.
Compensation Mechanism Precision¶
Whether curtailed output is compensated depends entirely on the project's actual PPA terms, applicable regulatory framework, or interconnection agreement — some arrangements compensate curtailed energy as if it had been generated and delivered, while others do not. This materially changes the revenue consequence of the same physical curtailment level, so the model should confirm and reflect the actual applicable compensation mechanism rather than assuming a default treatment.
Sensitivity Analysis for Curtailment Risk¶
Because of the trend risk of increasing curtailment in many constrained network zones, a dedicated downside sensitivity scenario testing a worse-than-historical congestion level should be included, rather than relying solely on historical curtailment data as representative of the asset's full remaining operating life. This is particularly important for projects in known or emerging congestion zones, where historical data may understate the curtailment risk actually facing the asset going forward.
Common Construction Pitfalls¶
Generic industry-average curtailment rate. Applying a standard industry curtailment percentage regardless of the project's specific node or zone ignores materially different local congestion conditions.
Static rate assumed across the full operating life. Holding curtailment constant at a historical level without modelling the trend risk of worsening congestion understates long-term curtailment exposure.
Compensation mechanism assumed by default. Applying a default compensated or uncompensated treatment without confirming the actual PPA, regulatory, or interconnection agreement terms misstates the revenue consequence of curtailment.
Recommended Practices¶
- Source curtailment assumptions from historical data specific to the project's actual grid node or zone.
- Model curtailment as a forward-looking trend informed by known network upgrades and the generation connection pipeline in the same zone.
- Confirm and apply the project's actual compensation mechanism for curtailed output.
- Test a dedicated worse-than-historical congestion sensitivity scenario.
Continue Reading¶
Related Pillars¶
Related Technical Guides¶
Related Glossary¶
Related Industries¶
How OXXON tests thisRun a free structural check with FMAE
Frequently Asked Questions
Why isn't a generic industry-average curtailment assumption sufficient?
Because curtailment risk is highly specific to the grid node or zone a project connects to, depending on local transmission capacity and the concentration of generation connected to that same network segment — a generic industry-average figure can materially understate or overstate the risk at a specific, more or less congested location.
Why should curtailment be modelled as a trend rather than a static rate?
Because grid congestion frequently worsens over time as more generation connects to the same constrained network, particularly in zones with high renewable buildout relative to available transmission capacity — a static curtailment rate held constant across the asset's full operating life can understate curtailment risk in later years relative to the year the model was originally built.
How should the compensation mechanism be represented?
Precisely against the project's actual PPA, regulatory framework, or interconnection agreement terms — some arrangements compensate curtailed output as if it had been generated and delivered, while others do not, and this materially changes the revenue consequence of the same physical curtailment level, so the model should not assume a default treatment without confirming the actual applicable terms.
What sensitivity analysis should be applied to curtailment risk?
A dedicated downside scenario testing a worse-than-historical congestion level, reflecting the trend risk of increasing curtailment in many constrained network zones, rather than relying solely on historical curtailment data as if it were representative of the asset's full remaining operating life.
How does this guide relate to the general curtailment mechanic in the technical output chain?
This guide extends, rather than replaces, the basic curtailment reduction already built into the technical output chain — it covers the modelling methodology for sourcing, forecasting, and sensitivity-testing that curtailment assumption, rather than introducing a separate mechanic.
References
Related Articles
Energy Financial Modelling
Energy financial modelling is the discipline of building financial models for power generation assets, independent power producers, and renewable energy projects — structured around a technical output schedule and an electricity revenue stack that a standard corporate or general project finance model has no direct equivalent for. This page is the hub for the Knowledge Centre's energy and power modelling content: how a power project model is architected, how electricity markets and dispatch mechanics translate into revenue, and how power purchase agreements, capacity payments, and merchant exposure combine into a project's revenue structure. Technology-specific renewable energy models (solar, wind, storage, hydro, and others), technical and commercial modelling mechanics, and institutional practice for this asset class are indexed here as the domain expands.
Curtailment
Curtailment is a reduction in a generation asset's output due to grid capacity constraints or contractual limits, independent of the equipment's own availability or resource conditions. It should be modelled as its own distinct output reduction, separate from availability, so that grid or contractual exposure can be tested and reported independently of equipment uptime.
Transmission and Grid Models
Connecting a power project to the electricity grid involves interconnection capital cost, ongoing transmission losses between the point of generation and the point of sale, queue position risk in congested interconnection processes, and potential responsibility for network upgrade costs beyond the project's own connection. This guide covers how each of these transmission and grid mechanics should be modelled, distinct from the generation and revenue mechanics covered elsewhere in this pillar.
Power Project Financial Model Structure
A power generation financial model is architected around a technical output schedule — generation volume for a variable-output asset or available capacity for a dispatchable one — that drives every downstream calculation: the electricity revenue stack, the operating cost build, and, where the asset is project-financed, debt sculpting and covenant testing. This guide sets out that architecture as a sequence of explicit, separately built modules, distinct from a standard corporate model's revenue-growth-first structure.
Financial Model Audit for Renewables
Renewable energy financial models combine standard project finance debt sculpting with technical assumptions specific to the energy source, resource yield (solar irradiance or wind speed), equipment degradation over the asset life, and curtailment risk, that directly determine the cash flow feeding the debt structure. Power purchase agreement pricing and tenor, and the merchant tail risk once a PPA expires, add a further layer of revenue structure specific to this sector. This page sets out the modelling risks specific to renewables, the audit findings that recur across solar, wind, and storage financings, and what lenders typically expect before financial close.