Search in the site by keyword

reports - Deliverable

Models for the optimal management of participation in energy markets of an aggregate of distributed flexibility resources in the Italian context: mathematical formulations and preliminary results

reports - Deliverable

Models for the optimal management of participation in energy markets of an aggregate of distributed flexibility resources in the Italian context: mathematical formulations and preliminary results

This work presents three mathematical models developed to determine the technical-economic feasibility of the provision of ancillary services by aggregates of residential users. These models allow for the optimal definition of the available flexibility, the daily check of the availability of such capacity, and the forecast of energy exchanges between the aggregate and the network for the following day. The first two models are based on a deterministic mixed integer linear programming method, while the third method also considers the uncertainties in the forecasting phase by adopting a mixed integer stochastic programming model.

Due to the growth in production from non-programmable renewable sources and the decarbonisation process of the energy sector, it became necessary to increase the number of the entities authorized to provide ancillary services. In this regard, in 2017 the Autorità di Regolazione per Energia Reti e Ambiente (ARERA – Italian regulatory authority for energy networks and environment) issued resolution 300/2017/R/eel aimed at promoting the experimental provision of some ancillary services by aggregates of distributed resources (even small ones) and from renewable sources.
In this work, three models are proposed that allow for a feasibility check of the provision of ancillary services by groups of residential users. The first model enables predicting the flexibility available in the aggregate and defines the optimal quantity that can be offered in the auctions for the assignment of fixed-term contracts, as provided for by the current regulatory framework. The second model is based on a Rolling Horizon approach, which can verify the feasibility of activating flexibility, both from a technical and economic point of view. Both of these models are based on deterministic mixed integer linear optimization methods, are implemented in a Python environment and solved using the Gurobi solver. Finally, a third model is described which allows for the prediction of energy exchanges between the aggregate and the electricity grid. The latter model aims to minimize the costs of buying and selling energy by the aggregate operator and to reserve the flexibility declared in the auctions for the assignment of fixed-term contracts. Due to the high uncertainties in the forecasting of photovoltaic production and the energy demand of residential users, the proposed model is based on a mixed integer stochastic programming approach. The generation of uncertainty realization scenarios is carried out using a Monte Carlo method with subsequent reduction of the scenarios using the k-means algorithm. Finally, the preliminary results showing how an aggregate composed of approximately 1000 residential units is able to offer the electricity system 1 MW of flexibility without causing unwanted effects on users are reported and commented on.

Projects

Comments