Decentralized Energy Resources Optimized with AI (DER-OP-AI)

Challenge
Many medium-sized photovoltaic (PV) systems (10–250 kW), with or without batteries, installed on multi-storey buildings, such as housing associations and larger apartment buildings, face the same challenges: suboptimal operation, PV systems that do not produce as expected, or situations where solar power is exported while electricity is purchased at inopportune times.

Solution
Already today, it is possible to earn money by making batteries and electric vehicles available to the power grid through ancillary services. With the technology from the innovation project DER-OP-AI, it will now also be possible for housing associations and large apartment buildings with solar panels to become part of that solution.

DER-OP-AI consists of four sub-areas that are integrated into a single, overall solution.

1. Control of PV inverters and batteries
It is still common for PV inverters to be programmed to deliver the maximum possible output from the solar panels without regard to the power grid or electricity prices.

This can be a challenge for the grid, and the large supply of solar power directly leads to low and even negative prices. A negative spot price means that solar owners pay to feed their surplus electricity into the shared power grid.

With intelligent control of PV inverters, production can be curtailed to match the building’s actual consumption and avoid export. This prevents both contributing to negative electricity prices and having to pay to dispose of excess power.

2. Reliable on-site data collection and system monitoring
Common to all battery control solutions is the so-called smart meter—an electronically readable electricity meter that measures exactly the same as the building’s main meter and communicates this data in real time to the battery control system. Even when the smart meter is installed correctly, which unfortunately is not always the case, many batteries either do not function at all or operate sub-optimally.

One of the most important steps is to establish monitoring so that the battery’s condition and operation can be observed.

3. AI-based analysis and optimization
The partners in the innovation project will investigate and implement the potential of AI and AI agents as the core of the control system. The expectation is that AI can optimize energy management by forecasting solar production, consumption, and electricity prices, so that energy is used and stored in the most efficient way. The technology can also detect faults and wear at an early stage and suggest maintenance before problems arise.

Finally, AI can learn user behavior and propose adjustments that reduce both costs and the CO₂ footprint.

4. Ancillary services and aggregated volume for BSP
Energinet has a new role in the energy system as a Balance Service Partner (BSP). The results from DER-OP-AI and its partners will be able to offer aggregated capacity for up- and down-regulation, either directly to Energinet as a BSP or in cooperation with a specialized balance-responsible party or aggregator.

This part of the innovation project creates new business opportunities for project partners. At the same time, there is significant potential for customers to increase the profitability of their solar and battery installations.