Fabrizio Sossan

Testing of a Predictive Control Strategy for Balancing Renewable Sources in a Microgrid

This paper presents the design of a control strategy for the energy management of a grid-connected microgrid with local distributed energy resources as: 10-kW photovoltaic plant, 11-kW wind turbine, and 15-kW-190-kWh vanadium-based electric storage system. According to future regulations, the renewable energy producers will also have to provide a day-ahead hourly production plan. The overall idea is, by knowing the meteorological forecasts for the next 24 h, to dispatch the microgrid in order to be able to grant the scheduled hourly production by means of proper management of the storage system. The usage of the storage system is, however, minimized by the energy management strategy. The system design is validated by experimental testing carried out in SYSLAB, a distributed power system test facility at Riso Campus, Technical University of Denmark.

Scheduling of domestic water heater power demand for maximizing PV self-consumption using model predictive control

This paper presents a model predictive control (MPC) strategy for maximizing photo-voltaic (PV) self-consumption in a household context exploiting the flexible demand of an electric water heater. The predictive controller uses a water heater model and forecast of the hot water consumption in order to predict the future temperature of the water and it manages its state (on and off) according to the forecasted PV production, which are computed starting from forecast of the solar irradiance. Simulations for the proof of concept and for validating the proposed control strategy are proposed. Results of the control approach are compared with a traditional thermostatic controller using historical measurements of a 10 kW PV installation. Economic results based on the Italian self consumption tariffs are also reported. The model of the water heater complex is a mixed grey and white box and its parameters have been estimated using a real water heater device.

Analysis of voltage support by electric vehicles and photovoltaic in a real Danish low voltage network

With conventional generating units being replaced by renewable sources which are not required to provide same high level of ancillary services, there is an increasing need for additional resources to achieve certain standards regarding frequency and voltage. This paper investigates the potential of incorporating electric vehicles (EVs) in a low voltage distribution network with high penetration of photovoltaic installations (PVs), and focuses on analysing potential voltage support functions from EVs and PVs. In addition, the paper evaluates the benefits that reactive power control may provide with addressing the issues regarding voltage control at the expense of increased loading. Analysed real Danish low voltage network has been modelled in Matlab SimPowerSystems and is based on consumption and PV production data measured individually for number of households.

Achieving the Dispatchability of Distribution Feeders Through Prosumers Data Driven Forecasting and Model Predictive Control of Electrochemical Storage

We propose and experimentally validate a process to dispatch the operation of a distribution feeder with heterogeneous prosumers according to a trajectory with 5 min resolution, called dispatch plan, established the day before the operation. The controllable element is a utility-scale grid-connected battery energy storage system (BESS) integrated with a minimally pervasive monitoring infrastructure. The process consists of two stages: day-ahead, where the dispatch plan is determined by using forecast of the aggregated consumption and local distributed generation (prosumption), and real-time operation, where the mismatch between the actual prosumption realization and dispatch plan is compensated for thanks to adjusting the real power injections of the BESS with model predictive control (MPC). MPC accounts for BESS operational constraints thank to reduced order dynamic grey-box models identified from online measurements. The experimental validation is performed by using a grid-connected 720 kVA/500 kWh BESS to dispatch the operation of a 20-kV distribution feeder of the École Polytechnique Fédérale de Lausanne campus with both conventional consumption and distributed photo-voltaic generation.

Grey-box modeling for system identification of household refrigerators: A step toward smart appliances

This paper presents the grey-box modeling of a vapor-compression refrigeration system for residential applications based on maximum likelihood estimation of parameters in stochastic differential equations. Models obtained are useful in the view of controlling refrigerators as flexible consumption units, which operation can be shifted within temperature and operational constraints. Even if the refrigerators are not intended to be used as smart loads, validated models are useful in predicting units consumption. This information can increase the optimality of the management of other flexible units, such as heat pumps for space heating, in order to smooth the load factor during peak hours, enhance reliability and efficiency in power networks and reduce operational costs.

Domestic refrigerators temperature prediction strategy for the evaluation of the expected power consumption

This paper discusses and presents a simple temperature prediction strategy for the domestic refrigerator. The main idea is to predict the duration it takes to the cold chamber temperature to reach the thresholds according to the state of the compressor and to the last temperature measurements. The experiments are conducted at SYSLAB facility at DTU Risø Campus having a set of refrigerators working at different set point temperatures, with different ambient temperatures and under different thermal load conditions. The prediction strategy is tested using a set of different refrigerators in order to validate the performances. The challenges to calculate the time with less error pronouncement in temperature, regulating power supply and its duration are also discussed.

Evaluation of the performance of indirect control of many DSRs using hardware-in-the-loop simulations

Controlling the power consumption of many Demand Side Resources, DSRs, will be required in the future power system where a big share of the electric energy will be produced using stochastic renewable sources and the conventional power plants might not have the flexibility of providing all the regulating power. Indirect control of demand side resources is supposed to shift the electric power consumption of each single unit through broadcasting of a control signal; the flexibility in the aggregated power consumption can be used for supplying balancing power to the electric power system. Indirect control approach is convenient from communication point of view since the real-time data flow is only in one direction because the decision is computed locally according to user preferences. On the other hand, this approach results in an open loop control scheme, since it is assumed that no real-time power readings from the units can be performed. The aim of the paper is to discuss the performance of an emulated closed loop control using an estimator for predicting the aggregate power response and a regulator. By using these components it is possible to produce a control signal to broadcast to distributed demand side resources. A population of DSRs, buildings with electric space heating, is indeed simulated in a software simulation platform using an hardware in the loop approach, that allows to feedback the real heat dynamics of SYSLAB FlexHouse into the simulations for pretending more realistic result.

Dynamic behaviour of a population of controlled-by-price demand side resources

It is described that controlling or shedding by price the power consumption of a population of thermostatic loads introduces in the aggregate consumption dynamic effects that cannot be disregarded if electrical flexible demand is meant to supply power system services. It is shown that inducing a desynchronization in the consumption contributes to damp the oscillations. Results are supported by Monte Carlo simulations of a population of buildings equipped with electric space heating whose consumption is indirect controlled by a dynamic price of the electricity.

Day-ahead scheduling of a photovoltaic plant by the energy management of a storage system

The paper discusses and describes a system for energy management of a 10 kW PV plant coupled with a 15 kW-190 kWh storage system. The overall idea is, by knowing the meteorological forecast for the next 24h, to dispatch the PV system and to be able to grant the scheduled hourly energy profile by a proper management of the storage. Due to forecast inaccuracies, the energy manager controls the storage in order to ensure that the plan for hourly energy production is respected, minimizing the storage itself usage. The experimental study is carried out in SYSLAB, a distributed power system test facility at DTU Riso Campus and part of PowerLabDK. Both the PV and the storage are connected to the local network and are fully controllable through the SCADA system. The control management and the models are implemented in Matlab-Simulink, which can be interfaced with SYSLAB.

Unsupervised Disaggregation of Photovoltaic Production From Composite Power Flow Measurements of Heterogeneous Prosumers

We consider the problem of estimating the unobserved amount of photovoltaic (PV) generation and demand in a power distribution network starting from measurements of the aggregated power flow at the point of common coupling and local global horizontal irradiance (GHI). The estimation principle relies on modeling the PV generation as a function of the measured GHI, enabling the identification of PV production patterns in the aggregated power flow measurements. Four estimation algorithms are proposed: the first assumes that variability in the aggregated PV generation is given by variations of PV generation, the next two use a model of the demand to improve estimation performance, and the fourth assumes that, in a certain frequency range, the aggregated power flow is dominated by PV generation dynamics. These algorithms leverage irradiance transposition models to explore several azimuth/tilt configurations and explain PV generation patterns from multiple plants with nonuniform installation characteristics. Their estimation performance is compared and validated with measurements from a real-life setup including four houses with rooftop PV installations and battery systems for PV self-consumption.