Workshop Programme

Proceedings: all participants have been given the link to the page from which they can download all presentations. In the near future the proceedings will also be available to all EWEA members via the EWEA Members’ Area.

Thursday 1 October 2015
Session 1: Users’ perspective and experience
Session 2: Numerical weather prediction models
Session 3: Wind power forecasting models and operational systems
Friday 2 October 2015
Session 4: Special applications
Session 5: Grid integration
Session 6: Crowdsourcing the future

Session 1: Users’ perspective and experience

Session chair: Lars Landberg, Director, Strategic Research & Innovation, Renewables, DNV GL

The session addressed end-user requirements at different time-frames and for different applications.

  • What are the particular needs of TSOs, DSOs, utilities, energy traders, wind farm operators, aggregators, etc.?
  • Seen from an end-user perspective what does the ideal system look like? What would be the one thing that would improve current systems?
  • Value and perceived value of forecasts, eg what would the value of a 1% reduction in uncertainty be?
  • Standardisation: the need for standardised products, tools, data, evaluation protocols, etc. in order to supports communities sharing knowledge.
  • Reviews of requirements and gaps related to forecasting in grid codes


Beyond MAE – the value of being right at the right time
Jon Collins
Global Head of Practice Forecasting
DNV GL, United Kingdom
Determining the Value of More Accurate Wind Power Forecasting in Global Electricity Markets
Pascal Storck
Global Manager of Energy Services
Vaisala Energy, USA
Benchmark of forecasting models: reviewing and improving the state of the art Daniel Cabezon
Head of Meteorological Models and Special Tasks, Energy Assessment,
EDP Renovaveis, Spain
What’s next for wind energy forecasting systems? – Current and future requirements seen from a utility’s point of view Tilman Koblitz
Optimization Continental Manager, Forecasting & Short-Term Optimization Continental
Business Area markets Vattenfall
Prediction of 40 GW of wind power
in the German power system – positive effects of the market option and increasing situational awareness in grid operation
Matthias Lange
Leading associate/Co-founder
Energy & Meteo Systems, Germany

Session 2: Numerical weather prediction models

Session chair: Thomas Haiden, Team Leader, Verification and Observation Monitoring, European Centre for Medium-Range Weather Forecasts

The session addressed:

  • NWP models: state-of-the-art in wind forecasting including future challenges and possible ways forward;
  • Value of limited area models and other downscaling methods for improving wind predictions;
  • Representation of uncertainty of near surface wind in NWP models.


Progress in wind forecasting in the ECMWF model
Thomas Haiden
Team Leader, Verification and Observation Monitoring
European Centre for Medium-Range Weather Forecasts
Improved weather forecasts for wind energy applications: Lessons learned and perspectives based on the ongoing German research project EWeLiNE Kristina Lundgren
Research Coordinator/Scientist
Research and Development
Deutscher Wetterdienst, Germany
US experience (WFIP II – Wind Forecasting Improvement Project)
James Wilczak
Team Lead for Wind Energy
NOAA Federal, United States
Forecasting wind (and radiation) for energy production and balancing of the network Claus Petersen
DMI-Danish Met Institute, Denmark

Session 3: Wind Power Forecasting models and operational systems

Session chair: George Kariniotakis, Professor, Centre for Processes, Renewable Energies and Energy Systems (PERSEE), MINES ParisTech, France

This session addressed:

This session presented advances in the development of models for short-term wind power forecasting. Focus was given on the prediction of extreme conditions and namely icing and on how these conditions affect the performance of the forecasting models. The session also presented advances on spatio-temporal forecasting where off-site measurements are used to improve very short-term predictions. Results from real-world test cases were presented. Several works on new modelling approaches were also presented in the related poster session.


Large-scale Very-short-term Spatio-temporal Forecasting by Spare Vector Autoregression
Jethro Dowell
PhD Student, Electronic and Electrical Engineering
University of Strathclyde
United Kingdom
Big data techniques applied to very
short -term wind power forecasting
Ricardo Bessa
Senior Researcher & Area Manager
Validation of New Model for Short-term Forecasting of Turbine Icing Beatrice Brailey
R&D Specialist
Forecasting DNV GL
Forecasting of icing for wind power applications Øyvind Byrkjedal
R&D Manager
Kjeller Vindteknikk

Session 4: Special applications

Session chair: Gregor Giebel, Seniort Scientist, DTU Wind Energy, Denmark

This session addressed:

  • Dedicated forecasting models for the time frames (very-short term, short, medium and longer term) of the various applications;
  • Advances in various forecast products needed in the applications (probabilistic forecasts, scenarios, ramps, extremes event forecasting, spatio-temporal forecasting…);
  • Quantification of the level of uncertainty in each part of the modelling chain (NWPs, downscaling, power curve modelling…)
  • Simulation of forecasts errors for long-term studies in a planning perspective.
  • Joint forecasting of wind power with other processes (i.e. with demand or PV generation at feeder level…).


Integrated weather risk assessment for operations & maintenance strategy optimization
Hajo Koopstra
TU Delft, Netherlands
Wind resource seasonal forecast
outcomes from SPECS project
Gil Lizcano
R&D Director
Evaluation of the level of prediction errors of PV and wind in a future with a large amount of renewable Robin Girard
Associate Professor, Energy and Processes
MINES ParisTech, France
Weather Information Decision Support Tool Tailored to the Needs of Renewable Energy Industry Isabel Metzinger
Research Scientist
Deutscher Wetterdienst, Germany

Session 5: Grid Integration

Session chair: Hannele Holttinen, Principal Scientist, Technical Research Centre of Finland, VTT, Finland

Forecasting will be needed in power system operation. This session presented forecasting use from a grid integration point of view: managing the power flows (congestion management) in both distribution and transmission system as well as managing the balancing task of system operators.


Dynamic Line Rating Probabilistic Forecasting and Evaluation
Romain Dupin
PhD Student
MINES ParisTech, France
Hotspot analysis for Dynamic Line Rating
Dirk Malda
R&D Energy metereology manager
Wind power and management of the transmission system in France
Emmanuel Neau
R&D Engineer
Relevance of forecasting on onshore wind units participating in R2 delivery. The perspective of a TSO
Jan Voet
Cross-Border & Long term balancing
market development manager
ELIA, Belgium

Session 6: Crowdsourcing the future

This interactive session addressed participants’ inputs on future issues related to wind power forecasting.

Session chair : Lars Landberg, Director, Strategic Research & Innovation, Renewables, DNV GL

During the 2 days of the workshop, participants were asked to submit their ideas for what the most important issues in wind power forecasting will be.Participants proposed 47 ideas which Lars Landberg grouped under 16 categories and presented these in the final session. Participants were asked to vote for the two that they thought most important.

It is hoped that this will inspire researchers and practitioners to steer their work in the directions identified. This valuable input will also be the starting point for EWEA’s third workshop on Wind Power Forecasting in 2017.