Forecasting

Once you've connected sites, resources, and started activating flexibility, you can use forecasting to enhance your flexibility offerings and revenue opportunities.

What is Forecasting?

Forecasting allows you to:

• Access our predictions for your sites and resources to better plan flexibility opportunities
• Share your own forecasts with us to help improve our understanding of your resources
• Optimise revenue by providing us with better information about expected behaviour

Feature availability

Feature Flag Required

Forecasting functionality is currently available as an advanced feature that requires activation. By default:

• GET requests are limited to forecasts up to 4 hours in the future
• POST requests (uploading your own forecasts) are not available
• Access to our most advanced forecasting models is restricted

If you're interested in accessing extended forecast horizons, better models or uploading your own forecasts, please contact our support team to discuss enabling these features for your account.

Forecasts vs Baselines

It's important to understand the distinction between forecasts and baselines, as they serve different purposes in the flexibility workflow:

Forecasts

• Definition: Predictions of what would happen with no interventions from either the VPP or local optimisers.
• Purpose: Understanding the "natural" behaviour of resources for planning flexibility opportunities.
• Example: A PV resource forecast follows solar irradiance patterns - what the panels would produce if left completely uncontrolled.

Baselines

• Definition: What the VPP expects the site/resource to do if no VPP activations are sent (but local optimisation is allowed)
• Purpose: The committed behaviour that partners will follow unless specifically asked to deviate via flexibility activations
• Example: A PV resource baseline might include curtailment periods where production is (partially) curtailed by the local optimiser, even though the sun is shining.

Key Distinction

• Forecasts = "What nature would do" (no human/system intervention)
• Baselines = "What you commit to do" (with your local optimisation, but without the VPP's flexibility requests)

In practice: Partners often use forecasts as inputs to create their baselines - taking the natural resource behaviour and applying local optimisation logic to determine committed operations.

Example workflow:

1. Forecasts: "Solar panels could produce 5kW at noon, and without intervention the site would inject 3kW net into the grid"
   • (This implies forecasted 2kW baseload consumption)
2. Local optimisation: "But prices are negative! We want to avoid injection, so we'll curtail to 2kW to match our 2kW baseload forecast (self-consumption mode)"
3. Baseline: "We commit to 2kW production at noon" (what the VPP can expect)
4. Flexibility: VPP can then ask to deviate from that 2kW baseline.

Why share your Forecasts?

When you share your forecasts with us, you help us:

1. Better estimate flexibility opportunities in advance by understanding true resource potential.
2. Improve market participation by having more accurate predictions of available flexibility.
3. Optimise dispatch decisions based on actual resource capabilities rather than planned operations
4. Reduce uncertainty in our planning algorithms by knowing what resources could do if unconstrained.

Your external forecasts take priority over our internal models, giving you direct control over how we understand your resources' natural behaviour.

How it works

The forecasting system operates on a simple priority hierarchy:

1. Your uploaded forecasts (highest priority) - When you provide forecasts, we use these as the primary source of truth
2. Our advanced models (medium priority) - Available with the forecasting feature enabled
3. Our standard models (lowest priority) - Default models available to all partners

This ensures that your domain expertise about your resources is always respected while providing fallback predictions when you don't have specific forecasts to share.

Getting started

Ready to start using forecasting? The following sections will guide you through:

1. Getting Forecasts - Learn how to retrieve forecast data for your sites and resources
2. Uploading Forecasts - Discover how to share your own predictions with us

Integration Approach

We recommend starting with getting forecasts to understand the data structure and timing. Once you're familiar with the API, contact our team to enable longer forecast horizons and advanced forecasting models. If you have your own forecasting models, you can also upload your predictions to complement our system.