Skip to main content

Solar generation forecasts

An integral part of running balancing operations on Tensor Cloud is creating accurate, reliable, and scalable asset generation forecasts. We currently provide downloadable forecasts up to 14 days for solar PV systems. We also generate consolidated forecasts on a BG-level for assets containing solar PV systems, batteries, and loads.

How to get started with forecasts

To get started with next-day generation forecasts for a given asset, you need to ensure the following:

  1. Your asset status is set to operational
  2. Your asset COD is in the past
  3. You have entered the right location ID in the settings of your asset
  4. You have uploaded historical generation data for your asset

Once you have completed step 1, Tensor Cloud automatically creates generation forecasts every 30 minutes. You can download forecasts in the main side bar menu under Forecasts. Steps 2 and 3 are required for AI-based forecasts to increase forecast accuracy.

tip

Although Tensor Cloud automatically creates forecasts for any asset set to operational with a COD in the past, even without historical generation data, we strongly recommend that you upload historical generation data for each of your assets. This will allow Tensor Cloud to create a customized AI model for each asset and ultimately generate more accurate forecasts.

Forecast data format

You can download day-ahead forecasts in CSV format. The forecast CSV files contain the following columns in order:

Column nameDescriptionUnit
datetimeStart of forecasted 30-minute time slotTimezone-aware ISO 8601 datetime format
totalForecasted total generation of all assetskWh
Name of the assetForecasted generation of each assetkWh
note

Our forecast data format does not explicitly specify start and end of each forecast. Instead, as the Japanese electricity markets operate in 30-minute increments, we assume that each forecast covers the following 30 minutes. For example, a forecast with a timestamp of 2023-05-11 05:00:00+09:00 will cover 05:00 to 05:30 on that day.

warning

We cannot guarantee the assets to be in any specific order in the forecast CSV files. Please make sure you design any downstream systems in a way that they can handle assets to be in any order.

Forecasting methodology

Generating accurate forecasts is a complex task that requires large amounts of data as well as a deep understanding of the underlying physical processes of electricity generation.

Tensor Cloud uses a combination of machine learning and physical simulations to generate forecasts. Physical simulations are what we also use for the long-term economic asset simulation on Tensor Cloud. They are based on the asset information you provide when adding new assets to Tensor Cloud. Physical simulations leverage a digital twin of your asset to estimate electricity generation and work in a similar fashion to simulations created with engineering software such as PVSyst. Machine learning prediction on the other hand uses historical generation and weather data to train an AI model that can then be used to generate forecasts.

For a given asset, if there is no or not enough historical generation data available, Tensor Cloud will default to a physical simulation-based forecast. If there is enough historical generation data available, Tensor Cloud will train a custom AI model for the asset and attempt to use it to generate forecasts.

The AI model is continuously evaluated against the physical simulation-based forecast and will only be used if it creates forecasts with a higher accuracy. From our experience, at least 50 days of historical generation data are needed to create an AI model that is more accurate than a physical simulation-based forecast.

You can find details on our forecasting methodology in the technology section.

Model re-training

Tensor Cloud regularly retrains each asset’s AI model whenever new historical generation data is available. The maximum retraining frequency is three times per week. If no new data is available for a particular asset, retraining will be skipped, and the existing model will continue to be used until new data is uploaded.