Enabling peer-to-peer energy trading by leveraging prosumer analytics

Small energy actors –especially small businesses and residential end-users– with distributed energy generation and storage units connected behind the meter (prosumers) are currently at the centre of the energy transition. Much research has focused on their engagement, either in terms of directly controlling their consumption or by exposing them to time-varying price signals reflecting the power system or wholesale electricity market conditions. However, centralized ways of stimulating the prosumers’ activation have not been proven successful due to the compromise in the end-users’ comfort and the limited economic incentives provided.

With the aim to increase the engagement of small prosumers, P2P-TALES proposes the use of data-driven knowledge extraction techniques in order to identify clusters of prosumers with similar or dissimilar energy-related behavioural characteristics, preferences and flexibility potential and create incentive schemes that reflect their financial and non-monetary interests.

Consequently, based on the prosumer analytics and clustering, P2P-TALES aspires to address several challenges related to distributed transaction systems through which energy is traded in a peer-to-peer (P2P) fashion, via two main research projects.

First, P2P-TALES aims at the development of trading protocols among prosumers, and possibly among autonomous trading agents, to improve sustainability and maximize value of innovative technologies installed behind the meter. This can be achieved by learning and promoting synergies between prosumers in a distributed fashion, on the basis of complementary objectives or alternative asset ownership. The developed methods will be also made robust against the external uncertain variables, e.g. the availability of renewable energy sources, by extending the latest developments in stochastic optimization from centralized optimization problems to multi-agent equilibrium problems. Distributed trading mechanisms shall induce prosumers, selfish entities in principle, to maximize their own value and make optimal use of the shared resources, e.g. community storages. The principles of such a distributed trading economy are then closely related to those of the sharing economy, whose outcomes will be used as benchmarks.

Second, P2P-TALES will attempt to map the cyber P2P transactions onto the physical distribution system and the actual flows of power, with the goal to assess the operational and reliability impacts. This can be achieved via the definition of a distribution system reliability assessment framework, intended as a practical methodology to quantify reliability in the light of flows and network conditions induced by coordination mechanisms such as P2P trading. The conceived reliability assessment framework will be beneficial for distribution network operators (DNO), especially for enhancing the effectiveness of their decision making job. In turn, this will allow them to procure system services from energy prosumers, e.g. for congestion management or for improving the operational characteristics of their grid, both preventively and in real-time.

In summary, P2P-TALES will strive to (1) cluster the behaviour of energy prosumers by applying machine learning methods to users’ big data sets, (2) conceive distributed trading protocols among prosumers by developing distributed learning methods for multi-agent equilibrium problems, and (3) assess the impact of trading mechanisms on the distribution network and in turn define system services that can be offered by prosumers so that integration of the cyber and physical systems can be achieved.

Project team members

Dr. ing. S. (Sergio) Grammatico (co-applicant)
PhD Aitazaz Ali Raja

Prof. dr. ir. J.G. (Han) Slootweg (PI) TU Eindhoven
Dr. N. (Nikolaos) Paterakis (Co-applicant) TU Eindhoven


Behavioural analytics, distributed energy resources, distributed equilibrium learning, distribution system optimization, peer-to-peer trading, sharing economy.

Sponsored by

NWO ESI-bida


  • Enexis B.V.
  • ICT Automatisering Nederland B.V.
  • SynerScope B.V.