We receive the weather forecast from a third party. Based on this, we conduct a physical simulation to determine the expected electricity generation from wind and solar.
We use the open source tools pyCity and richardsonpy to make a stochastic forecast of electricity consumption .
The installation of devices is currently underway in Bedburg and the go-live is scheduled for the end of November.
In this project, our technology is employed to cover only electricity needs – Xnet covers heating.
The weather forecast is retrieved with a seven day horizon. The resolution depends on the horizon: for the following 24 hours the forecast has a 15-minute resolution, the following three days has a 1-hour resolution and the following week a 3-hour resolution.
This data is used to calculate an optimization schedule for the following three days and constraints are sent to the local gateway for the next 24 hours. These constraints are updated every 15 minutes.
We think a wide coverage of devices is essential for the widespread applicability of a solution. Moreover, we think it is important to have an open system that allows external controls from third parties, like RWTH in this case, so they can deploy their own logic on top and tailor the solution to the individual use case.
The communication protocol between the individual device and our IoT-gateway, the gridBox, depends on the device.
Our integration team is continuously integrating new devices to improve the coverage of our XENON platform. You can learn more about their work here.
Through those integrations, our platform is able to provide an abstraction layer for various energy devices. This allows platform users to interact with energy assets, regardless of the underlying protocol.
No, we do not offer this. However, we do provide the necessary data to perform such an analysis. The RWTH has done this here. As shown in the presentation (slide 27), emissions and the degree of self-sufficiency are a function of the installed battery capacity. This allows the determination of a battery size that minimizes costs.
The gridBox is designed to continue operation when offline. Once back online, it feeds all data that was recorded during the network interruption into the cloud.