Uncertainty of various kinds has always played a significant role in energy systems: physically there are uncertain supply and demands, possible failures can propagate in a highly unpredictable fashion, and prices in energy markets can be highly volatile (and can become negative for extended periods of time). Physical safety margins and operational plans are commonly made using robust (worst case) analysis; and strategic decisions are often chosen using the principle of minimising worst regret over scenarios.
The European Union is currently working towards its target of a 20% renewable energy system by 2020, and 97% renewable electricity by 2050. Recent research (Jacobson et al. (2015)) supports the hypothesis that 100% renewable energy systems are achievable by 2050. In order to attain such potential goals, many mathematical problems need to be solved.
The goal of this outreach day is to show a variety of particular problems in this emerging field of research to the Dutch Stochastics community. The workshop is open to other participants as well.
This workshop will be held at the Amsterdam Science Park Conference Halls.
Six speakers (from both academia and industry) have confirmed their participation.
Registration is open to all, and possible until May 18. Participation is free, but please complete the registration form for catering purposes.
Travel directions can be found on the Science Park Congress Centre website. For further information, please contact the organiser
Bert ZwartChris Dent
Power system planning under uncertainty: models and real systems
Mathematical and computer modelling is widely used for decision support in planning of energy systems. This presentation will describe three topics in the use of modelling for decision support:
Yannig Goude
Advances and perspectives in electric load forecasting
Electricity load forecasting faces rising challenges due to the advent of innovating technologies such as smart grids, electric cars and renewable energy production. For utilities, a good knowledge of the future electricity consumption stands as a central point for the reliability of the network, investment strategies, energy trading, optimizing the production etc. Many statistical models have been investigated recently at EDF (Electricité de France) to forecast electricity consumption at different geographical scale and at temporal horizon for point and probabilistic forecasts. Among them stand regression on functional data, additive models, spatio-temporal modelling, ensemble method and on-line aggregation of experts. We will dress a panorama of these studies, focusing on real data applications and suggest some research perspectives.
Sean Meyn
Distributed Control Design for Balancing the Grid Using Flexible Loads
A distributed control architecture is presented that is intended to make a collection of heterogeneous loads appear to the grid operator as a nearly perfect battery. Local control is based on randomized decision rules advocated in prior research, and extended in this paper to any load with a discrete number of power states. Additional linear filtering at the load ensures that the input-output dynamics of the aggregate has a nearly flat input-output response: the behavior of an ideal, multi-GW battery system. Joint work with Ana Busic and Joel Mathias.
Louis Wehenkel
Distributed Control Design for Balancing the Grid Using Flexible Loads
During the last four years, the European FP7 funded project GARPUR (http://www.garpur-project.eu) has developed new probabilistic approaches for reliability management of the pan-European transmission system. The project covers in a consistent way decision making under uncertainty for grid development, asset management and system operation. It defines a mathematical framework for formulating the broad class of reliability management problems addressed by European TSOs, and for evaluating the socioeconomic impact of moving away from the current N-1 reliability management criterion towards a full-fledged probabilistic approach. The project has also developed several algorithmic approximations of the resulting probabilistic reliability assessment and control problems. Pilot tests are currently being carried to determine how to progressively implement them in practice, with a particular attention on data requirements and interpretation by human experts. In this talk, we will focus on the mathematical formulation of the GARPUR Reliability Management Approaches and Criteria (RMACs) in the form of chance-constrained multi-stage stochastic programming problems over multiple time-scales. These hard problems offer many opportunities for further research to find out clever ways of leveraging recent progress in machine learning and optimization. Some ongoing work along these lines will be illustrated.