Publications

1) Vovos, P. and Georgakas, K. (2021) Smart Boilers: Grid Support Services from Non-Critical Loads. Journal of Power and Energy Engineering, 8 ,pp. 23-45. doi: 10.4236/jpee.2020.812003.

2) Maryam Mohiti, Mohammadreza Mazidi, David Steen, Le Anh Tuan (2022), "A Risk-Averse Energy Management System for Optimal Heat and Power Scheduling in Local Energy Communities " in IEEE International Conference on Environment and Electrical Engineering and 2022 IEEE Industrial and Commercial Power Systems Europe (EEEIC / I&CPS Europe), Prague, 28 June-01 July 2022.

Events

1) "PV generation self-consumption", a presentation by Panagis Vovos for Patras IEEE Student Branch web-conference on "Smart Cities and modern Photovoltaics", May 2021. 
Click here to watch (in Greek) !

Reports

1) Simulation of Smart Boilers (D2.1), July 2021.
Deliverable D2.1 is the written report on the output provided by Task 2.1, which contains the operating analysis of Smart Boilers (SBs) through simulations. The ability of smart boilers to precisely control active power, thus heating rate, and simultaneously control reactive power and harmonic content will also be verified. The following activities will be carried out:
• Analysis of SBs through simulation with valuable inputs from similar converter analysis after close cooperation between UoP and UoPe.
• Modelling of the thermal response of the boiler with input data and parameters provided by Maltezos SA (subcontractor).
• Integration of SBs models with thermal response to achieve a complete simulation model of SBs.
The outcome is a complete, advanced, simulation tool for SBs, providing both their electrical and thermal response. 

2) Ancillary service definition (D3.1), October 2021.
In this report, a list of the ancillary services (AS) which can be provided to distribution system operators (DSOs) is defined, focussing on ancillary services provided by PV-dominated near-zero energy building (NZEB)-based local energy communities (LECs).
The defined ASs aim to allow larger PV penetration in distribution networks without affecting its secure operation. Initially, a literature study on the ASs that can be provided in system operators is performed, emphasising on the services that allow larger exploitation of solar energy for provision of services to distribution systems. Subsequently, the concise list of ancillary services is determined, and each service is analysed in detail, along with their impact on solar energy and distribution networks. Then, the feasibility and the applicability of the defined ancillary services, as well as their value to distribution system and their ability to facilitate increased PV penetration in distribution level is evaluated.
The outcome of this report will be used by Task 3.3 for the validation of the defined ancillary services in simulation environment as well as by Task 6.2 for the real-life demonstration of the ancillary service on Chalmers campus network.   

3) PV and load forecasting algorithms (D4.1), February 2022.
In this report, the developed forecasting algorithms for PV generation and electricity load of the local energy communities (LECs) are described. The forecasts will be as input to the energy management system of the LECs for short term (intra-day and near real time) and long term (day-ahead) management of their resources. Accordingly, advanced machine learning and mathematical based forecast algorithms with two different time steps of 1-hour and 15-minutes and three horizons of 15-minutes, 1-hour and 24-hours ahead are provided for the prediction of the PV sites and building loads of the LEC located at Chalmers University of Technology campus. To develop the machine learning algorithms, measurement data of PV historical production is collected form Chalmers campus PV sites, while weather variables are retrieved from numerical weather prediction models. For each forecasting algorithm based on the data, an individual forecasting method is presented; the forecast for short term and very short term (1-hour ahead and 15 minutes ahead) are based on dynamic recurrent neural networks (DRNNs) while the 24-hours ahead Artificial Neural Networks and hybrid methods are utilized for the prediction. Mathematical based algorithms such as state-space model identification regression method (SSIR) are also performed as comparison and a potential algorithm for the forecast system.
The results of the presented methods are validated and compared with the other state-of-art methods, which show the superiority of the presented method. Furthermore, the realization and future exploitation of the forecast system is briefly described. The presented forecast methods utilize weather prediction variables instead of measurements. Therefore, the accuracy of the weather predictions highly influences the accuracy of the PV forecast. However, their results are more viable in real time exploitation where real weather measurements are not available. The results of the algorithms on the real-life data of two PV sites and building loads in three mentioned time horizons at Chalmers are presented, to illustrate the accuracy and effectiveness of the provided methods. The short term and very short-term prediction show high accuracy while the accuracy decreases as the forecast horizon increases to 24 hours ahead. The presented forecasting tool of this report will be used in as input to the EMS of Task 3.3 as well as Task 6.2 for the real-life demonstrations. In SUNSETs the output of the forecasting system will be used in short term (intra-day and near real time) and long term (day-ahead) management of resources. Consequently, the forecasts are provided in two different time steps of 1-hour and 15-minutes with three horizons of 15-minutes, 1-hour and 24-hours ahead.

4) ​Optimization model of suggested PV system for NZEBs with respect to electricity costs and PV self-consumption (D7.1), April 2022 .
In this report we will study the optimal management of energy-consumption in a photovoltaic system with energy storage, through an optimization algorithm, which incorporates optimal scheduling of controlled loads, with a view to maximum net profit and minimum energy costs. In particular, we look at the latest measures of Greek and Swedish PV development using net metering systems. In such, the energy generated by the photovoltaic system can be either consumed by local consumers or fed into the grid. The final cost shown in the electricity bill depends on the energy produced by the photovoltaic system, the energy consumed by the grid and the energy fed into the grid. Therefore, the actual electricity consumption is very important for estimating the benefits of using this renewable energy source.

5) Test cases and demonstration plan (D6.5), May 2022.
In this report the outcome of the work in Task 6.5 is presented. The test cases that can be demonstrated different test sites of SUNSETS will be presented, along with the most significant technical requirements related to their implementation. The test cases are prioritized based on their significance and applicability, while the technical requirements include the required resolution of the measurements. . Furthermore, the devices that will be controlled in each test case have been identified. Therefore, it is expected that the definition of the test cases and the respective technical requirements specification will facilitate the preparation of each test site for the demonstration activities.

6) Assessment of Hardware and Control of Smart Boilers (D2.2), July 2022.
In this report a description of the hardware built so far for the purposes of the project and the device control aspects are presented. This presentation comes in a brief extend and does not enter in details related to all the considerations prior to the final component and design choices, with the exception of Section 3, where the reasons for selecting the specific topology are analyzed.The report opens with a general presentation of the energy storage and grid quality services concepts as realized with a Smart Boiler. It is explained that a non-critical load as a boiler is the perfect candidate to support these goals, under the currently configured conditions where the increased penetration depth of the renewables leads to unpredictable fluctuations of active/reactive power flow and potentially unacceptable harmonic distortion of the voltage waveform.Then follows the explanation of selecting a single switch topology with DC input voltage as rectified from the mains. This discussion highlights the advantages of the converter, which are simplicity, robustness, reliability, endurance, small size, low cost and high efficiency.
The main part of the report is dedicated to a one-by-one presentation and description of the several components that form the converter at both the power circuit and the control circuit. Features and specifications of these components together with power loss calculations and measurements, time response oscillograms and other technical details fully justify the design options of the research team and offer a comprehensive illustration of the device operation and performance.Next, the main aspects of the closed loop control strategy are presented, with graphs and block diagrams to clarify the background design and the resultant performance.
Some thoughts about future work are listed at the closing of the manuscript.