Learn how Tyba's latest updates will boost your revenue capture and streamline your development and financing processes.
We are excited to share several updates across the Tyba product. These include:
As always, we welcome your feedback on what features are most helpful for your team. We will be focusing on a variety of usability and automation improvements over the next few months, so please don’t hesitate to reach out with suggestions.
This month, our ML/AI team made major strides in our prediction accuracy, reducing the Mean Absolute Error (MAE) of our short-term price predictions in ERCOT and CAISO to $3-8/MWh (depending on the node) in the Day Ahead Market.
YTD Price Prediction Performance (Day-Ahead Energy @ Houston Hub)
Note: Achieved an MAE of $5.4/MWh YTD. Even where we did not predict the magnitude of a large spike, we most often predicted the existence of a spike, which is essential for energy storage applications.
We plan to share more detailed performance metrics in future blog posts, and are currently conducting ISO-wide backtesting. So far, we have seen comparable improvements for Real Time Market and Ancillary Service prices.
Accessing price forecasts
Tyba has updated the price forecasts being used in our Virtual Bidder module for existing customer’s projects. You will be able to see the uplift from this in your day-to-day revenue outcomes in the “Daily View” and over time in the “Summary View.”
Please reach out to Tyba if you would like to add new project locations to this module, or fill out this form (note the particular node, if known).
Benefits
Improved price forecasts directly translate to increased revenue for storage projects, given limited cycles to provide energy and ancillary services. However, Tyba helps customers translate these price predictions into actionable strategies — in operations, through virtual trading, and in the development and financing of projects.