CATEPS MICROGRID LIVING LAB: putting demand flexibility to the test

The concept of the energy transition is built on the foundation of decarbonisation—in other words, the gradual replacement of fossil fuel-based energy generation, which emits large volumes of GHG and contributes to global warming, with renewable energy sources (wind power, solar, etc.) that are free from such emissions.

From ‘consumer’ to ‘prosumer’

One of the sectors most deeply involved in this process is electricity generation and distribution. Indeed, it is undergoing a transformation—from large, centralised power plants (nuclear, gas, coal) to a multitude of smaller-scale generation and storage points (distributed energy resources, or DERs, like self-consumption, EVs, or batteries) connected to the electricity system via medium- and low-voltage distribution grids.

While those large plants were theoretically capable of generating electricity permanently, these smaller, renewable-based sources depend on intermittent resources (sun, wind, water, etc.) that must be efficiently managed to align electricity generation with demand. Additionally, energy must be stored for times when these primary resources are unavailable.

In this new model, the presence of DERs transforms the traditional consumer into a more active and relevant player in the energy system: a prosumer (producer + consumer).

What is demand flexibility?

This leads to the concept of demand flexibility, also known as Active Demand Management. This paradigm refers to the prosumer’s ability to voluntarily alter their usual consumption pattern—typically reducing it—in response to signals from the electricity system, such as price signals or financial incentives.

Effectively managing this activity requires accurate forecasting of electricity generation and demand, as well as optimised energy distribution. In this context, AI tools become indispensable, as they can process large data volumes and generate models capable of managing complex systems.

The growing presence of DERs in distribution grids demands that Distribution System Operators (DSOs) have robust capabilities to analyse and control these resources. Rather than following the traditional ‘iron alternative’—oversizing infrastructure in a ‘fit and forget’ approach—DSOs are now adopting a strategy of active system management (ASM), which includes demand flexibility control. This approach requires in-depth knowledge of the grid’s configuration (topology) and operation (generation/consumption models). In this regard, data analytics and AI techniques offer a powerful solution to tackle these challenges.

Demand aggregators, energy Communities, and local flexibility Markets

So far, demand flexibility in Spain has been limited to energy-intensive consumers (>5 MW) through legacy interruptibility programmes. The main challenge now—economic, technical, and regulatory—is to allow a broader range of players to participate in such programmes, including small and medium-sized consumers (industrial, commercial, and residential), either directly or via demand aggregators.

Demand aggregators are a promising new player. These agents manage the combined demand of a group of prosumers. The EU is paying special attention to Local Energy Communities (LECs) within this aggregated demand model. These LECs are defined by a community’s energy resources (a neighbourhood, municipality, or industrial park) that also involve a social component.

The EU considers the development of such systems a priority. Therefore, it advocates for the creation and implementation of so-called Local Flexibility Markets, with the participation of market operators, DSOs, and direct consumers and/or aggregators.

CATEPS uGrid Living Lab: a microgrid for advancing Local Flexibility Markets

To foster the development of these Local Flexibility Markets, the University of Seville has launched a research project at its School of Engineering. This initiative enables experimentation with AI in these markets and the deployment of new management technologies. The CATEPS uGrid Living Lab is a modular infrastructure equipped with various DERs across the campus. 

A real-world testbed for AI-driven demand flexibility

The main research objective of the facility is to develop new AI algorithms—especially machine learning and deep learning techniques—for managing demand flexibility, which can be tested in a real-world environment (Living Lab). The research also explores telecontrol protocols, IoT-based data capture devices, and blockchain-style decentralised architectures.

Several projects are currently under development at the facility in collaboration with Endesa, including DER4ALL: a scalable, multi-level, interoperable, and secure DERMS based on AI, edge computing, and decentralised architectures, focused on the research and development of platforms for managing distributed energy resources (DERs). These platforms are designed to be scalable, multi-level, interoperable, and secure. They leverage AI, edge computing, and decentralised architectures using blockchain technology to enable flexibility services based on the efficient management of distributed energy resources.

The project pursues four key objectives:

1. Forecasting demand, renewable generation, and operational optimisation.
2. Self-configuring the platform based on the relevant grid level (initially focused on end users, aggregators, energy hubs, or energy communities).
3. Applying decentralised architectures for asset traceability and management across various smart grid levels.
4. Deploying the platform in a real-world installation at the CATEPS Microgrid Living Lab, including equipment to emulate battery-integrated electric vehicles (V2X).

From a technical and scientific standpoint, the project will help develop a new solution that enables the rapid, scalable, secure, and efficient integration of low-voltage renewable energy resources while providing flexibility and traceability services.

From an economic and social perspective, it offers a tool for enabling circular energy economies. This tool contributes to promoting sustainable installations, facilitating user participation in electricity markets, boosting competitiveness, encouraging self-consumption, and lowering energy prices and costs—all while enhancing grid reliability.

On the other hand, the AI4FLEX project—focused on improving distribution grid management and demand-side flexibility through data analytics and AI—sets out three main objectives:

1. A demand flexibility management framework to enable optimal operation of the distribution system, based on advanced data analytics and AI applied to available information for congestion management and DER integration through forecasting and optimisation.
2. Integration of heterogeneous data sources with AMI (advanced metering infrastructure) to detect non-technical losses in distribution grids through topology discovery.
3. Use case: optimal flexibility management for the integration of electric vehicles at the distribution level, including modelling battery degradation.

By applying these three objectives, distribution system operators could gain access to a powerful tool that would allow them to operate the grid more efficiently. It will also contribute to enhancing grid resilience and its capacity to integrate generation.

Testing local energy communities

The building is also part of the eCitySevilla project, which seeks to establish a net-zero emissions zone in the Cartuja area. This initiative enables the facility to test multiple use cases from the project, including the development of platforms for balancing and integrating energy within local energy communities.