Search

International audience

Advanced data-driven methods can facilitate the participation of renewable energy sources in competitive electricity markets by leveraging available contextual information, such as weather and market conditions. However, the underpinning assumption is that data will always be available in an operational setting, which is not always the case in industrial applications. In this work, we present a feature-driven method that both directly forecasts the trading decisions of a renewable producer participating in a day-ahead market, and is resilient to missing data in an operational setting. Specifically, we leverage robust optimization to formulate a feature-driven method that minimizes the worst-case trading cost when a subset of features used during model training is missing at test time. The proposed approach is validated in numerical experiments against impute-then-regress benchmarks, with the results showcasing that it leads to improved trading performance when data are missing.

In a context of massive renovation of residential housing, stakeholders need decision-support tools based on knowledge of the current building stock and an accurate simulation of energy demand. For this purpose, we developed a val- idation/calibration method on a territorial/national scale in order to represent the real consumption of housing. This methodological approach provides (1) more reliable identification of energy-saving measures (changes in technology or behaviour) and (2) improved knowledge of the energy simulation tool and its post-calibration performance for optimisation issues. The main contribution of the calibration method described in this paper is the geographical scale concerned: all French residential housing has been modelled, simulated and calibrated with national data (geometries and attributes) on buildings. Furthermore, some occupants' socio-professional characteristics have been taken into account to reflect their actual energy behaviours. This is different from traditional approaches that focus only on a few buildings or archetypes. This paper also describes the application of this methodology on an Open Source simulation software in order to be easily verifiable and usable. This linear model will also be used to optimise renovation solutions at territorial scale in future work. All data used in this paper are Open Data and thus available to the scientific community. This method enabled more than 18 million buildings to be calibrated while reducing the Normalized Root Mean Square Error, between simulated and real energy annual consumption, from 52% to 24% for gas and from 24% to 15% for electricity. In addition, the user of the method is free to prioritise either the maximum error reduction or the number of calibration coefficients if a simpler model is desired. This paper also discusses the results obtained from this method for future improvement.

International audience; Challenging times for European electricity security have recently brought light to the importance of a robust power sector with well-functioning electricity markets. Increased uncertainty disrupts the standard practice of decision-making in energy systems. The increased penetration of Renewable Energy Sources (RES) such as wind and photovoltaic plants adds to this uncertainty due to the weather dependency of their electricity production. Artificial Intelligence (AI) based tools have proven their efficiency in different applications in the energy sector ranging from forecasting to optimization and decision making. They permit to simplify modeling chains and to improve performance due to higher learning capabilities compared to state-of-the-art methods. However, decision-makers of the energy sector, especially in high-risk situations, need to understand how decision-aid tools construct their outputs from the data. AI-based tools are often seen as black-box models and this penalizes their acceptability by end-users (traders, power system operators a.o.). The lack of interpretability of AI tools is a major challenge for the wider adoption of AI in the energy sector and a fundamental requirement to better support humans in the decision-aid process. Agents of energy systems expect very high levels of reliability for the various services they provide. As energy systems are impacted by multiple uncertainty sources (e.g. available power of RES plants, climate change, market conditions), developed AI tools should not only be performant on average situations but be able to guarantee robust solutions in the case of extreme events. Therefore, our research focuses on optimizing understandable symbolic representations of data-driven decision-aid models for human operators in the energy sector.

International audience; The European Horizon 2020 project Smart4RES (http://www.smart4res.eu), which started in 2019 and runs until April 2023, aims at improving modelling and forecasting of weather variables necessary to optimize the integration of weather-dependent renewable energy (RES) production (i.e. wind, solar) into power systems and electricity markets. It gathers experts from several disciplines ranging from meteorology, data science, power systems a.o. It aims to contribute to the pathway towards energy systems with very high RES penetrations by 2030 and beyond.This presentation has a double objective:(1) To present a comprehensive overview in terms of KPI improvements of the final results obtained by the project. These results cover thematic objectives including:Improvement of weather and RES forecasting;Streamlined extraction of optimal value from the data through data sharing, data market places, and novel business models for the data;New data-driven optimization and decision-aid tools for market and grid management applications;Validation of new models in living labs and assessment of forecasting value vs costly remedies to hedge uncertainties (i.e. storage). The results obtained are numerous. Without being exhaustive, they include: improved forecasting of weather variables with focus on extreme situations and also through innovative measuring settings (i.e. a network of sky cameras); A seamless approach to couple outputs from different ensemble numerical weather prediction (NWP) models with different temporal resolutions; Advances from ultra-high resolution NWPs based on Large Eddy Simulation; Approaches for RES production forecasting aiming at efficiently combining highly dimensionally input (various types of satellite images, NWPs, spatially distributed measurements etc.); Seamless probabilistic RES forecasting covering multiple time frames and data inputs; Resilient energy forecasting. In the front of applications methods are proposed to optimally use forecasts for the management of storage systems coupled with renewables, for the optimal trading of renewables in multiple markets and for grid management optimization and dynamic security assessment. Prescriptive analytics and explainable AI methods are proposed to optimize decision making. A cost benefit analysis is performed to assess the contribution of different types of data in forecasting problems.(2) To present hierarchized proposals for future research directions. An international workshop is organized by the project (14/04/2023), where experts are invited to assess where RES predictability stands today and propose research directions for the future. In this presentation we will present the conclusions of this workshop. This will be a useful insight for academics, industrials as well as policy makers in the field.Smart4RES Team:Simon Camal, Georges Kariniotakis, Dennis van der Meer, Konstantinos Parginos, Luka Santosuosso, Akylas Stratigakos [MINES Paris, PSL University, Centre PERSEE, France]; Gregor Giebel, Tuhfe Göçmen, Liyang Han [DTU, Denmark]; Pierre Pinson [Imperial College, UK]; Ricardo Bessa; Carla Goncalves, Rui Sousa [INESC TEC, Portugal]; Ivana Aleksovska, Bastien Alonzo, Marie Cassas, Quentin Libois, Marie-Adèle Magnaldo, Laure Raynaud [Meteo France, France]; Gerwin van Dalum, Gerrit Deen, Daan Houf, Remco Verzijlbergh [Whiffle, The Netherlands]; Matthias Lange, Björn Witha [Energy and Meteo Systems, Germany]; Niklas Blum, Jorge Lezaca, Bijan Nouri, Stefan Wilbert [DLR, Germany]; Maria Ines Marques, Catarina Mendes Martins, Manuel Silva [EDP, Portugal]; Wouter De Boer, Marcel Eijgelaar, Ganesh Sauba [DNV, The Netherlands]; John Karakitsios, Theodoros Konstantinou, Dimitrios Lagos, George Sideratos [NTUA/ICCS, Greece]; Theodora Anastopoulou, Efrosini Korka, Christos Vitellas [DEDDIE, Greece]; Stephanie Petit, Clementine Coujard [Dowel Innovation, France]

Standard practice of decision-making in energy systems relies largely on complex modeling chains to address technical constraints and integrate numerous sources of uncertainty. The increased penetration of Renewable Energy Sources (RES) such as solar and wind plants adds complexity due to the weather dependency of their electricity production. Artificial Intelligence (AI) based tools have proven their efficiency in different applications in the energy sector ranging from forecasting to optimization and decision making. They permit to simplify modeling chains and to improve performance due to higher learning capabilities compared to state-of-the-art methods. However, decision-makers of the energy sector need to understand how decision-aid tools construct their outputs from the data. AI-based tools are often seen as black-box models and this penalizes their acceptability by end-users (traders, power system operators a.o.). The lack of interpretability of AI tools is a major challenge for the wider adoption of AI in the energy sector and a fundamental requirement to better support humans in the decision-aid process. Agents of energy systems expect very high levels of reliability for the various services they provide. As energy systems are impacted by multiple uncertainty sources (e.g. available power of RES plants, weather and meteorological conditions, market conditions), developed AI tools should not only be performant on average situations but be able to guarantee robust solutions in the case of an extreme event. Therefore, our research focuses on understandable representations of data-driven decision-aid models for human operators in the energy sector. In order to enhance the interpretability of the AI models, a technique borrowed from the computer science domain is explored and further developed. Genetic programming and more precisely Symbolic Regression is used to derive a symbolic representation for the data-driven model that can take the form of a single equation. This equation results according to a specific reward function. The optimal solutions are selected naturally mimicking the biological theory of survival of the fittest. The main outcome is the production of symbolic representations of the AI models that require minimum changes when applied to different case studies. In this presentation a real-world use case is considered, to demonstrate the added value of the proposed tools for decision-making when trading the production of wind and solar power plants to the day-ahead market. An annual period of data is considered to train and test the proposed model. The typical modeling chain involves as many as 12 models for forecasting RES production, weather and meteorological conditions, together with stochastic optimization to derive trading decisions. A single AI-based model here replaces this complex chain. Such simplification is a significant enhancement to the modeling chain interpretability and facilitates trust to the human decision-maker. This work is carried out in part in the frame of the European project Smart4RES (Grant No 864337) supported by the H2020 Framework Program and in part in the frame the Marie-Curie COFUND project Ai4theSciences (Grant No 945304)

This paper presents the results of a series of tests on the provision of ancillary services by a Virtual Power Plant composed of variable Renewable Energy Resources. The ancillary services provided are Frequency Containment Reserve and Frequency Restoration Reserve, roughly primary and secondary reserve, and the resources used by the virtual power plant were wind and photovoltaics plant for a total of 273 MW of installed power on 17 generators. Renewable dispatchable units such as batteries, hydro or biomass were not considered in order to quantify the performance with the most critical generation technologies. Symmetric reserve activation of up to 10 MW. The quality of the reserve delivery was measured according to purpose-developed metrics and telecommunication failures were considered the main source of errors.

International audience

International audience; Straw bale building construction is attracting a revived public interest because of its potential for reduced carbon footprint, hygrothermal comfort, and energy savings at an affordable price. The present paper aims to summarize the current knowledge on straw bale construction, using available data from academic, industry, and public agencies sources. The main findings on straw fibers, bales, walls, and buildings are presented. The literature shows a wide variability of results, which reflects the diversity of straw material and of straw construction techniques. It is found that the effective thermal conductivity, density, specific heat, and elastic modulus of straw bales used in construction are in the range 0.033–0.19 W/(m·K), 80–150 kg/m3, 1075–2000 J/(kg·K), and 150–350 kPa respectively. Most straw-based multilayered walls comply with fire resistance regulations, and their U-value and sound reduction index range from 0.11 to 0.28 W/m2 K and 42 to 53 dB respectively, depending on the wall layout. When compared to standard buildings, straw bale buildings do provide yearly reductions in carbon emissions and energy consumption. The reductions often match those obtained after applying energy-saving technologies in standard buildings. The paper ends by discussing the future research needed to foster the dissemination of straw bale construction.

International audience

International audience; One of the levers of the Energy Transition is to massively integrate Renewable Energies (EnR) into power systems. This poses several challenges to the different actors involved; TSOs, DSOs, aggregators, market operators, etc. Today, all these players make extensive use of short-term forecasts of renewable energy production over time horizons ranging from a few minutes to a few days, to make decisions concerning the management of the electricity system and their assets. Research in the field of renewable energy forecasting is very active. The objective is always to improve the accuracy of the models, to better estimate the associated uncertainties, but also to optimize their use in applications. In this presentation we will illustrate how the state of the art has evolved throughout the year. New research directions will also be presented with a focus on some disruptive approaches developed within the framework of the European H2020 Smart4RES project that we coordinate.; Un des leviers de la Transition Energétique est d’intégrer massivement des Energies Renouvelables (EnR) dans les systèmes électriques. Cela pose plusieurs défis aux différents acteurs impliqués ; GRT, GRD, agrégateurs, opérateurs de marché,… Aujourd'hui, tous ces acteurs utilisent largement les prévisions à court terme de la production EnR à des horizons allant de quelques minutes à quelques jours, pour prendre des décisions concernant la gestion du système électrique et des actifs. La recherche dans le domaine de la prévision EnR est très active. L’objectif est toujours d'améliorer la précision des modèles, de mieux estimer les incertitudes associées, mais aussi d’optimiser leur utilisation dans les applications. Dans cette présentation nous allons illustrer comment l’état de l'art a évolué. De nouvelles directions de recherche seront également présentées en mettant l'accent sur certaines approches disruptives développées dans le cadre du projet européen H2020 Smart4RES que nous coordonnons.

Considering the critical roles of hydrogen in energy transition and the renewable character of biogas, an integrated process linking ionic liquid (IL) based biogas upgrading and thermal plasma (TP) assisted hydrogen production is conceptually proposed and studied from the process intensification point of view. To select a practically suitable IL absorbent for biogas upgrading, an IL screening is first conducted from an experimental database exhaustively collected from the literature. Following the thermodynamic screening and the assessment of important physical properties, the retained IL is evaluated in the conceptual biogas upgrading process. After that, the upgraded biogas with high biomethane purity is fed into a simulated TP reactor for the production of hydrogen by decarbonisation, where solid carbon could be simultaneously obtained as a second product. The improvement of the combined process is further examined by strategies of heat and power integration. The configuration of the whole integrated process is finally presented, showing a promising scenario for energy efficient and sustainable production of hydrogen.

International audience; Virtual power plants aggregating multiple renewable energy sources such as Photovoltaics and Wind are promising candidates for the provision of balancing ancillary services. A requisite for the provision of these services is that forecasts of aggregated production need to be highly reliable in order to minimize the risk of not providing the service. Yet, a reliability greater than 99% is unattainable for standard forecasting models. This work proposes alternative models for the day-ahead prediction of the lowest quantiles (0.1% to 0.9 %) of renewable Virtual power plant production. The proposed approaches derive conditional quantile forecasts of aggregated Wind/PV/Hydro production, obtained from tailored parametric models and machine learning models, including a Convolutional Neural Network architecture for predicting extremes. Reliability deviation is reduced up to 50 % and probabilistic skill score up to 18% compared to Quantile Regression Forest. Forecasting models are subsequently applied to the provision of downward reserve capacity by a renewable Virtual power plant. Increased forecasting reliability leads to a higher reliability of the reserve capacity, but reduces the average reserve volume offered by the renewable aggregation.

International audience; Short-term forecasting is key to the safe, reliable, and economic operation of modern power systems. As the majority of modern forecasting tools are purely data-driven, their performance relies heavily on the quality and availability of data. In this work, we examine forecasting when a subset of features used during model training becomes unavailable (deleted or missing) in an operational setting, a subject that has been largely overlooked by previous works. Several reasons could lead to feature deletion, including malicious data-integrity attacks, network latency, and equipment malfunctions, among others. We leverage tools from robust optimization and machine learning and formulate a linear regression model that is optimally resilient to the deletion of features at test time. The robust counterpart of the proposed model is a linear program whose size grows polynomially with the number of training observations and the number of features; we further provide a decomposition algorithm based on the alternating direction method of multipliers to deal with large problem instances that are typically found in energy forecasting applications. We further extend to the case of Probabilistic Forecasting by robustifying the standard linear quantile regression model. To validate empirically the proposed approach, we examine several prevalent forecasting practices in power systems, namely electricity prices, load, wind production, and solar production forecasting. We compare against regularized and randomization-based models and benchmark their performance for the case of feature deletion. The results show that the proposed solution successfully mitigates the adverse effects of missing features, leading to the lowest overall performance degradation. Further, it successfully hedges against the most adverse scenario of deleting the most important feature from the test set. The results persist both for point and probabilistic forecasts and across the different series. Overall, this work highlights the benefits of leveraging robust optimization and provides a new perspective on how to deal with feature uncertainty in energy forecasting applications.

International audience; Standard practice of decision-making in energy systems relies largely on complex modeling chains to address technical constraints and integrate numerous sources of uncertainty. The increased penetration of Renewable Energy Sources (RES) such as wind and photovoltaic plants adds complexity due to the weather dependency of their electricity production. Artificial Intelligence (AI) based tools have proven their efficiency in different applications in the energy sector ranging from forecasting to optimization and decision making. They permit to simplify modeling chains and to improve performance due to higher learning capabilities compared to state-of-the-art methods.However, decision-makers of the energy sector need to understand how decision-aid tools construct their outputs from the data. AI-based tools are often seen as black-box models and this penalizes their acceptability by end-users (traders, power system operators a.o.). The lack of interpretability of AI tools is a major challenge for the wider adoption of AI in the energy sector and a fundamental requirement to better support humans in the decision-aid process. Agents of energy systems expect very high levels of reliability for the various services they provide. As energy systems are impacted by multiple uncertainty sources (e.g. available power of RES plants, climate change, market conditions), developed AI tools should not only be performant on average situations but be able to guarantee robust solutions in the case of an extreme event.Therefore, our research focuses on understandable representations of data-driven decision-aid models for human operators in the energy sector. In order to enhance the interpretability of the AI models, a technique borrowed from the computer science domain is explored and further developed. Genetic programming and more precisely Symbolic Regression is used to derive a symbolic representation for the data-driven model that can take the form of a single equation. This equation results according to a specific reward function. The optimal solutions are selected naturally mimicking the biological theory of survival of the fittest.The main outcome is the production of symbolic representations of the AI models that require minimum changes when applied to different case studies. In this presentation a real-world use-case is considered, to demonstrate the added value of the proposed tools for decision making when trading the production of wind and solar power plants to the day-ahead market. An annual period of data is considered to train and test the proposed model. The typical modeling chain involves as many as 12 models for forecasting RES production and market quantities together with stochastic optimization to derive trading decisions. This complex chain is here replaced by a single AI-based model. Such simplification is a significant enhancement to the modeling chain interpretability and facilitates trust to the human decision-maker.This work is carried out in part in the frame of the European project Smart4RES supported by the H2020 Framework Program and in part in the frame the Marie-Curie COFUND project Ai4theSciences

International audience; The Horizon2020 Smart4RES project develops forecasting and optimization solutions for the operation of Renewable Energy Sources (RES) and their application in electricity market trading and grid management. As a follow-up of the Poster presented in 2021, this poster presents the latest solutions developed in Smart4RES that intend to cover a large spectrum of prediction horizons and spatial scales in order to maximize the interest for the wind power industry.At the scale of a wind farm, the fluctuations of wind conditions at very-short-term horizons (seconds to minutes ahead) impose significant variations in the structural load of the turbine and its power output. These variations are challenging for a precise control of wind turbines. The forecasting models developed by Smart4RES make use of new data sources unexploited by traditional models such as 2-beam and 4-beam nacelle-mounted LIDAR. Results obtained on an operating turbine equipped with LIDAR show significant improvement in the forecasting error of both structural load and active power output for the next 20 seconds ahead. At the horizon of the next minutes, the power output of a wind farm is impacted by weather variability but also by wake losses that may vary as a function of turbine curtailment following a system operator request. This is why Smart4RES proposes a dynamic Machine Learning (ML) prediction model based on Transfer Learning which provides adaptive forecast that beat state-of-art approaches including a similar ML model trained only in batch mode. Aggregations of renewable power plants are key players for renewable-based provision of services to the grid and optimized management of distribution grids. In this context, a coherent forecast over the entire hierarchy of the aggregation is essential in order to take decisions that are feasible considering local constraints in the various levels of the hierarchy. A main challenge in such hierarchies is to produce a forecast even if data is missing, which can occur frequently at different periods and levels in the hierarchy. Whereas existing approaches tend to discard periods with missing data, which can drastically reduce the amount of available data for training and the applicability of the forecasting model in real conditions, Smart4RES proposes an end-to-end learning approach that is able to derive coherent and precise hierarchical forecasts even in the presence of missing values at different levels of the hierarchy.

International audience; Spatially aggregating renewable power plants is beneficial when participating in electricity markets. In this context, a substantial number of features is available from various data sources. In machine learning, feature selection is common so as to relieve the curse of dimensionality and avoid overfitting. However, there is no guarantee that the selected features result in reliable forecasts and post-processing can therefore be valuable. In this study, we combine model agnostic feature selection with linear and nonlinear probabilistic forecast combination techniques. Moreover, the filters automatically compute the weights for our analog ensemble (AnEn) forecast model that does not require training. We verify our model chain by generating intra-day forecasts of the aggregated output of 60 wind turbines and 20 photovoltaic power plants using 831 input features in total. We show that automatic feature selection improves forecast accuracy while forecast combination improves reliability.

International audience; Spatially aggregating renewable power plants is beneficial when participating in electricity markets. In this context, a substantial number of features is available from various data sources. In machine learning, feature selection is common so as to relieve the curse of dimensionality and avoid overfitting. However, there is no guarantee that the selected features result in reliable forecasts and post-processing can therefore be valuable. In this study, we combine model agnostic feature selection with linear and nonlinear probabilistic forecast combination techniques. Moreover, the filters automatically compute the weights for our analog ensemble (AnEn) forecast model. We verify our model chain by generating intra-day forecasts of the aggregated output of 20 photovoltaic power plants using 831 input features in total. We show that the collection of filters selects a heterogeneous feature set but that each individual AnEn-filter combination results in underdispersed forecasts, which is efficiently remedied by the forecast combination techniques.

International audience; Power systems feature an inherent hierarchical structure. Ensuring that forecasts across a hierarchy are coherent presents an important challenge in energy forecasting. In this context, proposed reconciliation or end-to-end learning approaches assume coherent historical observations by construction; this assumption, however, is often violated in practice due to equipment failures. This work proposes an end-to-end learning approach for hierarchical forecasting that directly handles missing values. First, we show that a class of off-the-shelf machine learning algorithms already leads to coherent hierarchical forecasts. Next, we describe a conditional stochastic optimization approach based on prescriptive trees for end-to-end learning with missing values, without imputation or disregarding of quality observations. We validate the proposed approach in two case studies comprising 60 wind turbines and 20 photovoltaic parks, respectively. The empirical results show that end-to-end learning outperforms twostep reconciliation approaches and that the proposed solution mitigates the adverse effect of missing values.

National audience

International audience; High-quality data obtained from three-dimensional Particle Tracking Velocimetry (3D PTV) is pivotal for indoor environment engineering when designing ventilation strategies or monitoring airborne pollutants dispersion in inhabited spaces. A new method is proposed to link multiple 3D PTV systems, positioned side by side so that the entire measuring volume can be covered. An algorithm is developed to establish a link between the particles' trajectories calculated by each 3D PTV system. To evaluate the validity and robustness of the multi-PTV algorithm, synthetic particles were created, which follow different motions implemented in Matlab, including two analytical solutions for incompressible Navier-Stokes equations, namely Kovasznay and Beltrami flows, as well as a linear motion. Two different tracking algorithms were used to obtain 3D particle positions. The numerical results reveal that the proposed method is capable of connecting adjacent multiple PTV systems with reasonable accuracy and therefore can be considered as a promising method for indoor airflow measurements.

International audience; High temporal resolution intra-day and day-ahead photovoltaic (PV) power forecasts are important to maximize the value of PV systems because they enable stakeholders to participate in both the energy and ancillary services markets. Whereas most day-ahead electricity markets feature an hourly temporal resolution, intra-day markets may require forecasts at 5-minute resolution. In addition, battery integration can improve power system management in isolated grids with high PV power penetration, but battery control requires high temporal resolution forecasts. We propose an efficient method based on pattern matching to generate multivariate probabilistic forecasts, approximated by trajectories, at high temporal resolution and without the need to separately forecast the marginals and estimate the covariance matrix. We compare the proposed method against quantile regression forests in combination with copula theory and show that our method reduces the forecast time by approximately 98% and simplifies the modeling chain while incurring a minor performance penalty.

International audience

International audience

International audience; The combination of renewable energy sources into the power system network has been growing rapidly in recent decades. As a consequence, there have been thoughtful concerns over the consistent and acceptable working of the electrical power systems. One of the elucidations being recommended to expand the dependability and functioning of this power system network is to join together with energy storage devices. The battery storage device may possibly be used for increasing the profit margin of solar or wind farm proprietors. This chapter discusses the present state of battery energy storage technology and its economic viability which impacts the power system network. Further, a discussion on the integration of the battery storage technology to the grid-tied photovoltaic (PV) is made.

We present the properties of magnetic fields (B-fields) in two clumps (clump 1 and clump 2), located at the waist of the bipolar H II region Sh2-201, based on JCMT SCUBA-2/POL-2 observations of 850 $\mu$m polarized dust emission. We find that B-fields in the direction of the clumps are bent and compressed, showing bow-like morphologies, which we attribute to the feedback effect of the H II region on the surface of the clumps. Using the modified Davis-Chandrasekhar-Fermi method we estimate B-fields strengths of 266 $\mu$G and 65 $\mu$G for clump 1 and clump 2, respectively. From virial analyses and critical mass ratio estimates, we argue that clump 1 is gravitationally bound and could be undergoing collapse, whereas clump 2 is unbound and stable. We hypothesize that the interplay between thermal pressure imparted by the H II region, B-field morphologies, and the various internal pressures of the clumps (such as magnetic, turbulent, and gas thermal pressure), has the following consequences: (a) formation of clumps at the waist of the H II region; (b) progressive compression and enhancement of the B-fields in the clumps; (c) stronger B-fields will shield the clumps from erosion by the H II region and cause pressure equilibrium between the clumps and the H II region, thereby allowing expanding I-fronts to blow away from the filament ridge, forming bipolar H II regions; and (d) stronger B-fields and turbulence will be able to stabilize the clumps. A study of a larger sample of bipolar H II regions would help to determine whether our hypotheses are widely applicable., Comment: 25 pages, 10 figures, and 4 tables. Accepted for publication in ApJ

International audience; As the world’s population grows, current agricultural activities are putting enormous strain on the climate, including deforestation, excessive water use, and pesticide runoffs. These stresses are expected to increase in the coming future as more food products are needed. One promising method to reduce the adverse impact on the environment is greenhouse-based farming, which can enhance land-use efficiency and water consumption. The major limitation of the greenhouse is the high energy demand in a constrained space. Solar energy has shown a promising approach for integration with agricultural greenhouses in recent years. The developments of greenhouses integrated with various solar energy technologies including photovoltaic (PV), photovoltaic-thermal (PVT), and solar thermal collectors are discussed in this chapter. The results indicated that PV modules installed on greenhouse roofs or walls cause shading and, in some cases, have affected the growth trend of cultivated crops inside. The PVT modules are more efficient at producing both heat and electricity, while higher performance values are reported for greenhouses integrated with solar thermal collectors especially in moderate climate conditions. Additionally, storing thermal energy has shown an improvement in the profitability of solar greenhouses by extending the period of solar heat availability. Also, the most commonly used mathematical models describing the thermal behavior of solar greenhouses as well as their economics are explored and discussed. In the end, various commercial solar greenhouse implemented around the world are introduced.

In this presentation we detail highlight results obtained from the research work within the European Horizon 2020 project Smart4RES (http://www.smart4res.eu). The project, which started in 2019 and runs until 2023, aims at a better modelling and forecasting of weather variables necessary to optimise the integration of weather-dependent renewable energy (RES) production (i.e. wind, solar, run-of-the-river hydro) into power systems and electricity markets. Smart4RES gathers experts from several disciplines, from meteorology and renewable generation to market- and grid-integration. It aims to contribute to the pathway towards energy systems with very high RES penetrations by 2030 and beyond, through thematic objectives including:Improvement of weather and RES forecasting, Streamlined extraction of optimal value through new forecasting products, data market places, and novel business models; New data-driven optimization and decision-aid tools for market and grid management applications; Validation of new models in living labs and assessment of forecasting value vs costly remedies to hedge uncertainties (i.e. storage). In this presentation we will focus on our results on models that permit to improve forecasting of weather variables with focus on extreme situations and also through innovative measuring settings (i.e. a network of sky cameras). Also results will be presented on the development of seamless approach able to couple outputs from different ensemble numerical weather prediction (NWP) models with different temporal resolutions. Advances on the contribution of ultra-high resolution NWPs based on Large Eddy Simulation will be presented with evaluation results on real case studies like the Rhodes island in Greece.When it comes to forecasting the power output of RES plants, mainly wind and solar, the focus is on improving predictability using multiple sources of data. The proposed modelling approaches aim to efficiently combine highly dimensionally input (various types of satellite images, numerical weather predictions, spatially distributed measurements etc.). A priority has been to propose models that permit to generate probabilistic forecasts for multiple time frames in a seamless way. Thus, the objective is not only to improve accuracy and uncertainty estimations, but also to simplify complex forecasting modelling chains for applications that use forecasts at different time frames (i.e. a virtual power plant - VPP- with or without storage that participates in multiple markets). Our results show that the proposed seamless models permit to reach these performance objectives. Results will be presented also on how these approaches can be extended to aggregations of RES plants which is relevant for forecasting VPP production.

International audience; In many countries, electricity prices on day-ahead auction markets result from a market clearing designed to maximize social welfare. For each hour of the day, the market price can be represented as the intersection of a supply and demand curve. Structural market models reflect this price formation mechanism and are widely used in prospective studies guiding long-term decisions (e.g. investments and market design). However, simulating the supply curve in these models proves challenging since estimating the sell orders it comprises (i.e. offer prices and corresponding quantities) typically requires formulating numerous techno-economic hypotheses about power system assets and the behaviors of market participants. Due to imperfect competition, real market prices differ from the theoretical optimum, but modeling this difference is not straightforward. The objective of this work is to propose a model to simulate prices on day-ahead markets that account for the optimal economic dispatch of generation units, while also making use of historical day-ahead market prices. Inferring from historical data is especially important when not all information is made public (e.g. bidding strategies) or due to difficulty in accurately accounting for qualitative notions in quantitative models (e.g. market power). In this paper we propose a method for the parametrization of sell orders associated with production units. The estimation algorithm for this parametrization makes it possible to mitigate the requirement for analytic formulation of all of the above-mentioned aspects and to take advantage of the ever-increasing volume of available data on power systems (e.g. technical and market data). Parametrized orders also offer the possibility to account for various factors in a modular fashion, such as the strategic behavior of market participants. The proposed approach is validated using data related to the French day-ahead market and power system, for the period from 2015 to 2018.

International audience; Standard practices of decision-making in energy systems are dynamic, non-linear, complex, and chaotic processes in nature. Trading the power produced by solar photovoltaic (PV) plants in electricity markets is an important decisionmaking problem which receives increasing attention in the past few decades. The main objective of this paper is to build an interpretable data-driven decision aid model for the case study of a solar power plant with the objective to minimize imbalance costs and thus maximise the revenue, using Symbolic Regression (SR) through Genetic Programming. The use of SR in the experiments and analysis developed in this paper show numerous advantages. SR evolves linear combinations of nonlinear functions of the input variables. Three penalty metrics are introduced to enhance the interpretability of the final solutions. SR shows robust results, especially in the case study.

International audience; Bio-based materials such as straw are becoming a promising alternative to improve the building energy performance and to reduce its carbon footprint. When compared to common building construction materials, bio-based materials control the temperature and the relative humidity variation to ameliorate the indoor comfort with a low embodied energy and CO2 emission. This paper presents a comprehensive review of the thermal and mechanical properties of straw-based materials and buildings. The objective is to synthesis the work that has been carried out by the research community and to compare the results. The paper first introduces straw bale as a construction material from a historical viewpoint and in the context of the current building sector. The second part focuses on the available chemical and microstructural data of the straw fiber. The third part refers to the thermophysical and mechanical properties of the bales. The fourth part reviews the numerical and experimental studies done at the wall scale. The fifth part describes straw bale construction methods considering the regulation, structure requirements, and life cycle assessment data. Last, a critical analysis of the currently available data on straw as a building material is carried out and pending research issues are discussed. It was found that, despite abundant literature on structural and thermal properties of straw bale constructions, there is still a lack of some information. At a fiber scale, more research should be done to compare straw fibers to other natural and synthetic fibers. At a bale scale, further pH-related research is needed because it affects the material's interior conditions and durability. In addition, a thermal conductivity model for straw should be developed. On a bigger scale, the hygrothermal characteristics of various types of walls must be measured and computed experimentally and theoretically under various exterior and internal situations. More research is needed to improve the sound resistance of the straw wall by adding new layers capable of absorbing acoustic waves. Studies on the energy behavior, cost analysis, and how interior air moisture is self-regulated in straw buildings are needed at the building size. Therefore, a lack of consistent data among the different studies was noted depending on the straw characteristics.

National audience

International audience; Energy forecasting models deployed in industrial applications face uncertainty w.r.t. data availability, due to network latency, equipment malfunctions or data-integrity attacks. In particular, the case when a subset of features that has been used for model training becomes unavailable when the model is used operationally, poses a major challenge to forecasting performance. In this work, we present a principled approach to introducing resilience against missing features in energy forecasting applications via robust optimization. Specifically, we formulate a robust regression model that is optimally resilient against missing features at test time, considering both point and probabilistic forecasting. We develop three solution methods for the proposed robust formulation, all leading to Linear Programming problems, with varying degrees of tractability and conservativeness. We provide an extensive empirical validation of the proposed methods in prevalent applications, namely, electricity price, load, wind production, and solar production, forecasting, and we further compare against well-established benchmark models and methods of dealing with missing features, i.e., imputation and retraining. Our results demonstrate that the proposed robust optimization approach outperforms imputationbased models and exhibits similar performance to retraining without the missing features, while also maintaining practicality. To the best of our knowledge, this is the first work that introduces resilience against missing features into energy forecasting.