Modelling and Forecasting of Photovoltaic Generation for Microgrid Applications: from Theory to Validation

The penetration of stochastic renewable generation in modern power systems requires to reconsider conventional practices to ensure the reliable functioning of the electrical network. Decentralized control schemes for distributed energy resources (DERs) have gained attention to support the grid operation. In order to cope with the uncertainties of the DERs, predictive schemes that leverage on forecast of renewable generation recently came into prominence. The period of the control action typically depends on the availability of the reserve in the grid. For the case of microgrids, their limited physical extension and the lack of spatial smoothing imply fast power fluctuations and the necessity of coupling energy management strategies with real-time control. Among the DERs, small-scale photovoltaic (PV) systems are expected to represent most of the future available capacity, and consequently, solar resource assessment and power forecasting are of fundamental importance. This thesis focuses on developing forecasting methods and generation models to support the integration of photovoltaic systems in microgrids, considering short-term temporal horizons (below one hour) and fine spatial resolution (single site installations). In particular, we aim at computing probabilistic prediction intervals (PIs), i.e. we include information accounting for the intrinsic uncertainty of the prediction. In this respect, nonparametric tools to deliver PIs from sub-second to intra-hour forecasting horizons are proposed and benchmarked. They forecast the AC power and/or the global horizontal irradiance (GHI) by extracting selected endogenous influential variables from historical time series. The methods are shown to outperform available state-of-the-art techniques, and are able to capture the fastest fluctuations of small-scale PV plants. Then, we investigate how the inclusion of features from ground all-sky images can be used to improve time-series-based forecasting tools, thanks to identifying clouds movement. In this respect, we define a toolchain that allows predicting the cloud cover of the sun disk, through image processing and cloud motion identification. Furthermore, a methodology to estimate the irradiance from all-sky images is proposed, investigating the possibility of using an all-sky camera as an irradiance sensor. Next, we consider the problem of having power measurements that are corrupted by exogenous control actions (e.g. curtailment) and, therefore, not representative of the true potential of the PV plant. We propose a model-based strategy to reconstruct the maximum power production of a PV power plant thanks to integrating measurements of the PV cell temperature, system DC voltage and current. The strategy can improve time series-based direct power forecasting techniques when the production of the PV system is curtailed and thus the measured power does not correspond to the maximum available. The proposed methods to model and forecast the PV generation are then integrated in a single chain that allows to deliver power PIs that are able to account for the overall uncertainty of a PV system at a predefined confidence level. In the last part of the thesis, the proposed methods are experimentally validated in a real microgrid by considering possible applications in modern power systems.

Model Predictive Control Strategies for Polygeneration systems and microgrids

Ramanunni Parakkal Menon

Increasing electricity and thermal demand in all sectors, an increasing focus on the reduction in carbon emissions and use of nuclear power, advent of distributed generation and greater use of renewable technologies on an aging electrical and thermal grid system has necessitated the need for modern control and management systems. These new control and management systems need to be able to integrate new technologies, stochasticities and maximise the utilisation of the existing infrastructure while satisfying demands, without requiring complete overhaul of the pre-existing centralised grid system and the transmission and distribution systems. A model predictive control system has been proposed and demonstrated here which is able to create strategies for thermal and electrical systems such that the grid efficiency and security is maintained while minimising resource usage and emissions, while, simultaneously reducing the operating costs in the grid. The model predictive control(MPC) utilises a fully energetic approach for low-voltage microgrids and houses in the residential and commercial sector which comprises of CHP units, heat pumps, storage systems(electric and thermal) and stochastic renewable resources, while accounting for the varying dynamics of the electrical and thermal systems. Finally, validation of the MPC is performed on a testbed with physical units and building emulators which have access to meteorological and resource market data. The capability of the MPC to provide strategies for systems with photovoltaics (PV), heat pumps and CHP units is demonstrated. The MPC implementation developed is input into an optimal system design algorithm based on a multi-objective optimisation genetic algorithm developed for microgrids and urban systems/grids with end-users and polygeneration systems and storage devices. The optimal design of the system is so that the optimal sizes of the polygeneration systems can be identified. This will help in maximising the utilisation of heat pumps, storage devices and other systems in a LV microgrid equipped with an MPC-based thermo-electric energy management system. The work also aims to compare the cost effectiveness versus ability of thermal storage devices compared to electrical storage devices for the same grid in question.

EPFL2017

Solar irradiance estimations for modeling the variability of photovoltaic generation and assessing violations of grid constraints: A comparison between satellite and pyranometers measurements with load flow simulations

Rahul Kumar Gupta, Mario Paolone, Enrica Scolari, Fabrizio Sossan

Global horizontal irradiance (GHI) is typically used to model the power output of distributed photovoltaic (PV) generation. On the one hand, satellite estimations are nonpervasive and already available from commercial providers, but they have a limited spatiotemporal resolution. On the other hand, local estimations, e.g., from pyranometers, sky-cameras, and monitored PV plants, capture local irradiance patterns and dynamics, but they require in situ monitoring infrastructure and upgrading the asset of electrical operators. Considering that in most power systems, PV generation is typically the aggregated contribution of many distributed plants, are local GHI estimations necessary to characterize the variability of the power flow at the grid connection point (GCP) and detect violations of the limits of voltages and line currents accurately? To reply, we consider GHI measurements from a dense network of pyranometers (used to model the ground truth GHI potential), satellite estimations for the same area, and information about a medium and low voltage distribution system. We perform load flows at different levels of installed PV capacity and compare the nodal voltages, line currents, and the power at the GCP when the irradiance is from pyranometers and when from satellite estimations, deriving conclusions on the necessity, or not, of highly spatiotemporally resolved irradiance estimations.

2019

Real-Time Control Framework for Active Distribution Networks Theoretical Definition and Experimental Validation

Lorenzo Enrique Reyes Chamorro

The great challenge of massively integrating the volatile distributed power-generation into the power system is strongly related to the evolution of their operation and control. The literature of the last decade has suggested two models for such an evolution: (i) the supergrid model, based on enhanced continental/intercontinental network interconnections (mainly DC) for bulk transmission, (ii) the microgrid mode, where small medium/low voltage networks interfacing heterogeneous resources, such as local generation, energy storage and active customers, are intelligently managed so that they are operated as independent cells capable of providing different services from each other and operate in islanded mode. Irrespective of the model that will eventually emerge, the control of heterogeneous distributed resources represents a fundamental challenge for both supergrid and microgrid models. This requires the definition of scalable and composable control methods that guarantee the optimal and feasible operation of distribution grids in order to satisfy local objectives (e.g., distribution grid power balance), as well as the provision of ancillary services to the external bulk transmission (e.g., primary and secondary frequency supports). Several control methodologies have been proposed to achieve these goals, and the majority of them have been inspired by the classic time-layered approach traditionally adopted in power systems that are associated with different time-scales and extension of the controlling area, i.e. primary, secondary and tertiary controls, ranging from sub-seconds to hours, respectively. In the context of microgrids, these three levels of control can be associated with a decision process that can be centralized (i.e., a dedicated central controller decides on the operation of the system resources) and/or decentralized (each element decides based on its own rules). In the current literature, the former is used for long-term, whereas the latter for short-term decisions. In particular, primary controls are typically deployed through fully decentralized schemes mainly relying on the use of droop control. With this in mind, in this thesis we propose, and experimentally validate, a novel control framework called COMMELEC â A Composable Framework for Real-Time Control of Active Distribution Networks, Using Explicit Power Set-Points. It controls a power grid in real-time based on a multi-agent structure, using a simple and low-bandwidth communication protocol. Such a framework enables a controller to easily steer an entire network as an equivalent energy resource, thus making an entire system able to provide grid support by exploiting the flexibility of its components in real-time. The main features of the framework are (i) that it is able to indirectly control the reserve of the storage systems, thus maximizing the autonomy of the islanding operation, (ii) that it keeps the system in feasible operation conditions and better explores, compared to traditional techniques, the various degrees of freedom that characterize the system, and (iii) that it maintains the system power-equilibrium without using the frequency as a global variable, even being able to do so in inertia-less systems. Our framework has been extensively validated, first by simulations but, more importantly, in a real-scale microgrid laboratory specially designed and setup for this goal. This is the first real-scale experiment that proves the applicability of a droop-less explicit power-flow control mechanism in microgrids.