I-NERGY Services for optimal grid flexibility through demand response

The exploitation of energy flexibility in an electricity grid is achieved through Demand Response (DR). DR can be considered as the main demand-side management mechanism in smart grids. The term «Demand Response» refers specifically the changes in electricity usage by end-users / prosumers (industrial, commercial or domestic) [1]. In this context, end-users undertake to change their conventional consumption patterns using temporary on-site power generation or by reducing / shifting electricity consumption away from time periods of low RES generation and / or high demand always responding to signals coming from the administrator or electricity provider as a result of DR programs.

There is a wide classification of DR programs depending on the response factor [2]. However, such programs can be briefly divided as follows:

  • Price-based DR programs. In this setting, the price of electricity changes at different time periods, in order to motivate end-users to change their energy consumption patterns. The schemes that fall into this category are usage duration, peak value, and real-time value (RTP). The simplest application of such a program is the well-known night tariff.
  • Contract-based DR plans. This type of system encourages end-users to reduce their electricity consumption on demand or by prior arrangement. Such control strategy requires the design of incentives or contracts proposed to consumers, taking into account their behaviors and preferences. To achieve this goal, DR solutions use extensive AI-based solutions.

To lead the successful implementation of demand response programs and fully leverage prosumers’ capabilities to provide flexibility, which is valuable for the smart grid, I-NERGY will develop the following AI-empowered services:

  • Flexibility based prosumer clustering. This service allows for deciding similar flexibility providers within a large number of available prosumers scattered a selected area, which include flexible loads, battery storage, RES local generation and EVs.
  • Load and flexibility forecasting. This service allows for predicting prosumers load and generation behavioural patterns along with the flexibility stemming from said patterns.

These services, that also take into account external factors such as meteorological measurements, will allow the aggregation of community-level loads and flexibility into a unified decentralised Virtual Power Plant (VPP). Therefore, flexibility trading is facilitated and requests conveyed by the market are better matched leading to satisfaction of both prosumers, aggregators and grid operators.

[1] I. Antonopoulos et al., “Artificial intelligence and machine learning approaches to energy demand-side response: A systematic review,” Renew. Sustain. Energy Rev., vol. 130, p. 109899, Sep. 2020, doi: 10.1016/J.RSER.2020.109899.

[2] J. S. Vardakas, N. Zorba, and C. V. Verikoukis, “A Survey on Demand Response Programs in Smart Grids: Pricing Methods and Optimization Algorithms,” IEEE Commun. Surv. Tutorials, vol. 17, no. 1, pp. 152–178, Jan. 2015, doi: 10.1109/COMST.2014.2341586.