Elsevier

Energy

Volume 198, 1 May 2020, 117188
Energy

Energy scheduling of a smart microgrid with shared photovoltaic panels and storage: The case of the Ballen marina in Samsø

https://doi.org/10.1016/j.energy.2020.117188Get rights and content

Highlights

  • A new framework for the smart energy management of a marina is proposed.

  • Controllable loads, shared resources, and demand-supply balance are considered.

  • A model predictive control approach for the optimal energy scheduling is detailed.

  • Energy cost and self reliance are analyzed extensively using a public real dataset.

  • The proposed approach effectively improves profit and sustainability.

Abstract

This paper focuses on the Model Predictive Control (MPC) based energy scheduling of a smart microgrid equipped with non-controllable (i.e., with fixed power profile) and controllable (i.e., with flexible and programmable operation) electrical appliances, as well as photovoltaic (PV) panels, and a battery energy storage system (BESS). The proposed control strategy aims at a simultaneous optimal planning of the controllable loads, the shared resources (i.e., the storage system charge/discharge and renewable energy usage), and the energy exchange with the grid. The control scheme relies on an iterative finite horizon on-line optimization, implementing a mixed integer linear programming energy scheduling algorithm to maximize the self-supply with solar energy and/or minimize the daily cost of energy bought from the grid under time-varying energy pricing. At each time step, the resulting optimization problem is solved providing the optimal operations of controllable loads, the optimal amount of energy to be bought/sold from/to the grid, and the optimal charging/discharging profile for the BESS.

The proposed energy scheduling approach is applied to the demand side management control of the marina of Ballen, Samsø (Denmark), where a smart microgrid is currently being implemented as a demonstrator in the Horizon2020 European research project SMILE. Simulations considering the marina electric consumption (340 boat sockets, a service building equipped with a sauna and a wastewater pumping station, and the harbour master’s office equipped with a heat pump), PV production (60kWp), and the BESS (237 kWh capacity) based on a public real dataset are carried out on a one year time series with a 1 h resolution. Simulations indicate that the proposed approach allows 90% exploitation of the production of the PV plant. Furthermore, results are compared to a naïve control approach. The MPC based energy scheduling improves the self-supply by 1.6% compared to the naïve control. Optimization of the business economy using the MPC approach, instead, yields to 8.2% savings in the yearly energy cost with respect to the naïve approach.

Introduction

Due to increasing concerns on environment sustainability, policy makers, researchers, and practitioners have recently devoted attention to strategic actions aimed at energy efficiency and energy savings [[1], [2], [3]]. The growing energy demand and penetration of renewable energies pose several challenges to grid operators and thus motivate the acclaimed roll-out of the smart grid concept [4]. This concept refers to the enhancement of the traditional power grid infrastructure, especially in terms of pervasive use of information and communications technologies [[5], [6], [7]]. Indeed, to reduce the generation of energy from fossil fuels and expand the use of renewable energy sources, new advanced solutions for the future power grids need to be explored [8,9]. Smart grids are also attractive from a management and control perspective, since decision and control actions aimed at optimally controlling power flows between components should be considered in order to achieve the economic and environmental profitability of their operations [10]. In particular, the motivation for moving from the traditional energy systems paradigm of “generation-follows-load” to the new paradigm of “load-follows-generation” has led to developing Demand Side Management (DSM) methodologies [11]. These methodologies play an important role for the sustainable development of both district heating [12] and electrical distribution networks [13]. The term DSM is generally used to indicate a set of actions aimed at efficiently controlling and managing the energy consumption of a site, mainly to decrease the costs incurred for the supply of electricity, including network and system charges, and reduce the need for investments in networks and/or power plants for meeting peak demands [14]. These optimization actions are aimed at modifying the characteristics of the energy consumption, in terms of the overall entity of consumption, time profile of consumption, and grid connection parameters, in order to bring the energy demand and supply closer to a perceived optimum [14].

In basic implementations of DSM, an in-depth analysis of energy consumption of a site (e.g., building, network of buildings, facility, district, etc.) allows determining its own peculiarities of each site and understanding if the consumption habits can be optimized without the use of additional tools [15]. Alternatively, the on-site installation of renewable sources (e.g., photo-voltaic panels, wind turbines), cogeneration plants as well as energy storage systems (e.g., electrochemical, thermal storages) -also known as distributed generation and storage-enables collection of energy from many sources and leads to lower environmental impacts while improving security of supply [16]. However, such a variety of small, grid-connected devices implies the need of an Energy Management System (EMS), which allows to monitor the dynamics of all the involved assets (loads, batteries, generators), optimize in real-time the performance of storage and generation systems, and leverage on deployed assets to provide the power grid with supplementary services [17]. Indeed, all these components must autonomously and efficiently cooperate for the optimal demand response of end-users and efficient interaction with power grid, smart appliances, renewable energy sources (RESs), battery energy storage systems (BESSs) and plug-in electric vehicles (PEVs) [14]. As a consequence, the necessity of developing an effective optimal energy scheduling framework tackling all these objectives is apparent.

Coping with this challenge, this paper focuses on the Model Predictive Control (MPC) based energy scheduling of a smart microgrid equipped with non-controllable (i.e., with fixed power profile) and controllable (i.e., with flexible and programmable operation) electrical appliances, as well as photovoltaic (PV) panels, and a BESS that allows storing surplus of energy for later use. The proposed control strategy aims at a simultaneous optimal planning of the controllable loads, the shared resources (i.e., the storage system charge/discharge and renewable energy usage), and the energy exchange with the grid. The control scheme relies on an iterative finite horizon on-line optimization, implementing a mixed integer linear programming energy scheduling algorithm to maximize the self-supply with solar energy and/or minimize the daily cost of energy bought from the grid under time-varying energy pricing. At each time step, the resulting optimization problem is solved providing the optimal operations of controllable loads, the optimal amount of energy to be bought/sold from/to the grid, and the optimal charging/discharging profile for the BESS. The proposed energy scheduling approach is applied to the demand side management control of the marina of Ballen, Samsø (Denmark), where a smart microgrid is currently being implemented as a demonstrator in the Horizon2020 European research project SMILE [54]. Simulations considering the marina electric consumption (340 boat sockets, a service building equipped with a sauna and a wastewater pumping station, and the Harbour Master’s office equipped with a heat pump), PV production (60kWp), and the BESS (237 kWh capacity) based on a real public dataset [59] are carried out on a one year time series with a 1 h resolution. Simulations indicate that the proposed approach allows 90% exploitation of the production of the PV plant. Furthermore, results are compared to a naïve control approach. The MPC based energy scheduling improves the self-supply by 1.6% compared to the naïve control. Optimization of the business economy using the MPC approach, instead, yields to 8.2% savings in the yearly energy cost with respect to the naïve approach.

The remainder of this paper is organized as follows. Section II provides an overview of the literature on microgrid energy scheduling and positions the paper contribution. Section III describes the detailed model of the system under study. The naïve and MPC-based formulations of the scheduling problem are presented in Section IV and Section V, respectively. The case study results and analysis are provided in Section VI to demonstrate the benefits of the proposed strategy. Finally, the paper ends with conclusions and future work in Section VII.

Section snippets

Literature review

There is a large literature on energy management of microgrids. In the context of DSM, the aim of these systems is essentially to enable users to respond to market mechanisms and determine the optimal schedule of the energy activities. The literature reports several methods for energy scheduling. In general, two types of strategies can be found as described in detail in the sequel: offline scheduling [[18], [19], [20]] and real-time scheduling [[22], [23], [24], [25], [26], [27], [28], [29],

System model

In this section we present a mathematical model related to the microgrid of a marina, which is aimed at formulating the energy scheduling problem for the electrical appliances and the shared resources, i.e., the RES and the BESS as well as the demand-supply balance and constraints.

We consider the marina as a small port area including facilities such as docks equipped with electrical sockets for boats, a service building used by sailors, and an office building for the harbour master. The

Naïve energy scheduling strategy

For the sake of defining a baseline strategy, in this section we describe the simplest possible approach for determining time slot-by-time slot the amount of energy to be consumed by each CL, the amount of energy to be charged/discharged into/from the BESS, and the amount of bought/sold from/to the grid.

The proposed naïve control strategy is formally expressed by Algorithm 1 (Fig. 2). This algorithm is composed by two stages: the naïve scheduling of CLs and the scheduling of the BESS and energy

MPC energy scheduling strategy

. Naïve energy scheduling of the sauna heater

Inputs: {Text(τ)}, {Ts(τ)}, {Tsmin(τ)},{Tsmax(τ)}
Procedure:
1 set τ1
2 set δsh0
3 iterate
4 set τsstartbeginning of the nearest sauna period of use
5 set τsstop end of the nearest sauna period of use
6 if τsstartΔτspreheatingτ<τsstart
7 set δsh1
8 elseif τsstartτ<τsstop
9 if Ts(τ)<(Tsmin(τ)+Tsmax(τ))/2
10 set δsh1
11 end
12 else
13 set δsh0
14 end
15 ys(τ)=δshE¯s
16 set ττ+1
17 end
Outputs: {ys(τ)}

. Naïve energy scheduling algorithm of the BESS and

Case study

In this section we show the performance of the proposed MPC algorithm through numerical experiments. Specifically, we consider a period of analysis of one year (i.e., T=[0,8,784]) with a prediction horizon of H=24 h (i.e., H(t)=[t+1,t+24]) and a sampling time of 1 h. The obtained results are reported in the sequel and are analyzed and compared with the previously defined naïve control strategy. In particular, five different cases are analyzed (Table 1). In the first case the naïve control

Conclusions

Smart energy systems must be developed in close correlation and interplay with advanced control strategies to meet the challenge of integrated fluctuating renewable energy sources (RES) and achieve a balance between generation and consumption. Microgrids with energy storage capacity and load flexibility play a crucial role in this process. Coping with this, we propose a Model Predictive Control (MPC) based energy scheduling for the Demand Side Management (DSM) of a microgrid equipped with

Acknowledgement

This paper presents work that is partially supported by the Horizon 2020 research programme through the project SMILE (Smart Island Energy systems) under grant agreement no. 731249. The article is based on an oral presentation delivered at the 5th conference on Smart Energy Systems (SES) in Copenhagen, Denmark, September 10–11, 2019.

References (60)

  • S. Lee et al.

    Optimal power management for nanogrids based on technical information of electric appliances

    Energy Build

    (2019)
  • A. Fratean et al.

    Control strategies for decreasing energy costs and increasing self-consumption in nearly zero-energy buildings

    Sustainable cities and society

    (2018)
  • D. Kolokotsa et al.

    Predictive control techniques for energy and indoor environmental quality management in buildings

    Build Environ

    (2009)
  • S. Long et al.

    Generalised control-oriented modelling framework for multi-energy systems

    Appl Energy

    (2019)
  • I. Prodan et al.

    A model predictive control framework for reliable microgrid energy management

    Int J Electr Power Energy Syst

    (2014)
  • Q. Zhang et al.

    Enterprise-wide optimization for industrial demand side management: fundamentals, advances, and perspectives

    Chem Eng Res Des

    (2016)
  • B. Matthews et al.

    Demand side management of a run-of-mine ore milling circuit

    Contr Eng Pract

    (2013)
  • A.J. Van Staden et al.

    A model predictive control strategy for load shifting in a water pumping scheme with maximum demand charges

    Appl Energy

    (2011)
  • C.E. Garcia et al.

    Model predictive control: theory and practice—a survey

    Automatica

    (1989)
  • B.P. Esther et al.

    A survey on residential demand side management architecture, approaches, optimization models and methods

    Renew Sustain Energy Rev

    (2016)
  • P.A. Østergaard et al.

    Business and socioeconomic assessment of introducing heat pumps with heat storage in small-scale district heating systems

    Renew Energy

    (2019)
  • J. Jantzen et al.

    Sociotechnical transition to smart energy: the case of Samso 1997-2030

    Energy

    (2018)
  • H.M. Marczinkowski et al.

    Evaluation of electricity storage versus thermal storage as part of two different energy planning approaches for the islands Samsø and Orkney

    Energy

    (2019)
  • L. Ranieri et al.

    Energy recovery from the organic fraction of municipal solid waste: a real options-based facility assessment

    Sustainability

    (2018)
  • X. Fang et al.

    Smart grid — the new and improved power grid: a survey

    IEEE Communications Surveys Tutorials

    (2012)
  • R. Ma, H.-H. Chen, Y.-R. Huang, and W. Meng, “Smart grid communication: its challenges and opportunities,” IEEE Trans....
  • C. Talarico et al.

    A high precision phase control unit for DDS-based PLLs for 2.4-GHz ISM band applications

    IEEE 58th International Midwest Symposium on Circuits and Systems (MWSCAS)

    (2015)
  • G. Casalino et al.

    Incremental adaptive semi-supervised fuzzy clustering for data stream classification

    IEEE Conference on Evolving and Adaptive Intelligent Systems (EAIS)

    (2018)
  • You H. Lund et al.

    The status of 4th generation district heating: research and results

    Energy

    (December 2018)
  • P. Palensky et al.

    Demand side management: demand response, intelligent energy systems, and smart loads

    IEEE Transactions on Industrial Informatics

    (Jun. 2011)
  • Cited by (74)

    • Loads scheduling for demand response in energy communities

      2023, Computers and Operations Research
    View all citing articles on Scopus
    View full text