Group members

Post-doctoral fellows

image Jalal Kazempour

“Stochastic complementarity models within energy markets”

Mentor: Pierre Pinson (DTU Elektro)

Supported by the Danish projects PROAIN and “5s” (Danish Strategic Research Council)

The newly established stochastic complementarity optimization tools are properly compatible with most decision-making problems within energy markets, since they appropriately model the markets functioning while representing the potential uncertainties plaguing markets. Accordingly, the applicability of stochastic complementarity tools into the following research works related to the energy markets are expected to be investigated:
• Strategic ramp offering of a producer in a renewable-integrated electricity market
• Optimal long-term strategy of a trader in a renewable-integrated electricity market
• Optimal decisions of a distribution operator considering diverse distribution energy resources
• Optimal network simplification of a renewable-integrated power system
• Strategic offering of a gas-fired electricity producer considering interactions between electricity and gas markets
• Strategic wind producer in a two-settlement electricity market.”

Link to personal webpage: Jalal Kazempour

Publications (selection):

A. Nasri, S. J. Kazempour, A. J. Conejo, and M. Ghandhari (2015). Network-constrained AC unit commitment under uncertainty: A Benders’ decomposition approach. IEEE Transactions on Power Systems, to be published.

S. J. Kazempour, A. J. Conejo, and C. Ruiz (2015). Strategic bidding for a large consumer. IEEE Transactions on Power Systems 30: 848-856.

S. J. Kazempour and H. Zareipour (2014). Equilibria in an oligopolistic market with wind power production. IEEE Transactions on Power Systems 29: 686-697

S. J. Kazempour, A. J. Conejo, and C. Ruiz (2013). Generation investment equilibria with strategic producers – Part I: formulation. IEEE Transactions on Power Systems 28: 2613-2622.

image Athanasios Papakonstantinou

“Auctions with probabilistic offering in electricity markets”

Mentor: Pierre Pinson (DTU Elektro)

Supported by the Danish project “5s” (Danish Strategic Research Council)

Electricity markets consist of a day-ahead market permitting to settle on supply, consumption and prices, and a real-time market for settling on deviations from the day-ahead schedule. Such mechanisms were designed in a context where dispatchable generators were dominating, with demand mostly seen as inflexible. These mechanisms are challenged by the advent of renewable energy, which brought substantial variability and lack of predictability, complemented by more pro-active demand.

The central objective is to rethink the design of electricity markets by embedding the uncertain nature of renewable energy supply and consumption in the market mechanism itself. Offers will take a probabilistic form, while market clearing will rely on stochastic optimization. A proposal of a conceptual market framework will be initially tested in a toy model based on NordPool and later will be generalised for various forms of electricity markets.

Addressing effectively the uncertainty in electricity markets will benefit power system operations, and increase social welfare by introducing more efficient market designs.

Link to personal webpage: Athanasios Papakonstantinou

Publications (selection):

A Papakonstantinou, P Bogetoft (2013). DEA based auctions simulations. European Journal of Operational Research 231: 507-511

image Emil Banning Iversen

“Space-time probabilistic forecasts of renewable energy generation”

Mentor: Pierre Pinson (DTU Elektro)

Supported by the Electric Power Research Institute, USA

Distributed generation from renewable sources such as wind and solar are becoming increasingly popular. This has the potential to curb emissions and to alleviate the dependence on fossil fuels. However, generation of wind and solar power is notoriously uncertain. Furthermore this uncertainty is correlated across space and time. Thus in order to adequately model the uncertainty in a large generation system the spatio-temporal interdependence in renewable generation needs to be taken into account. This is important for many operational aspects of the electrical grid ranging for grid stability and management to trading electricity on energy markets. In this project we address these issues by developing novel tools and models for forecasting and simulating distributed renewable energy generation. These models and tools will result in an open-source simulation tool for distributed renewable generation.

Publications (selection):

EB Iversen, JK Møller, JM Morales, H Madsen. Short-term Probabilistic Forecasting of Wind Speed using Stochastic Differential Equations. International Journal of Forecasting, in print, 2015.

EB Iversen, JK Møller, JM Morales, H Madsen. Probabilistic Forecasts of Solar Irradiance using Stochastic Differential Equations. Environmetrics, 25, pp. 152-164, 2014.

EB Iversen, JM Morales, H Madsen. Optimal Charging of an Electric Vehicle using a Markov Decision Process. Applied Energy, 123, pp. 1-12, 2014.

PhD students

image Tue Vissing Jensen

“Exploring market models for a European electricity grid with a high penetration of renewable energy sources”

Supervisors: Pierre Pinson (DTU Elektro), Martin Greiner (Aarhus)

Supported by the Danish project “5s” (Danish Strategic Research Council)

Renewable energy sources such as wind and solar energy are taking an ever-increasing role in the European power systems. These have characteristics that are radically different from those of conventional generation, owing to their non-dispatchable nature, variability and lack of predictability. Current electricity markets were never designed for a context with large penetration of renewable energy sources. More appropriate market designs should account for the complex space-time dynamics of renewable energy generation and the uncertainty of their forecasts.

The PhD project focuses on various aspects of this challenge, as applied to the European power system. Firstly, a much-needed large-scale data set including a simplified electricity network as well as predicted and observed supply and demand for the whole European system is to be generated and made publicly available. Based on this dataset, a variety of existing and novel approaches to market design will be implemented, analysed and compared. Optimal bidding of renewable generation will thenbe examined using ideas from stochastic game theory.

It is expected that a better understanding of the interplay between renewables and electricity markets will lead to better use of renewables, as well as an increase in social welfare for both electricity suppliers and consumers.

Publications (selection):

image Emil Mahler Larsen

“Electricity market design for distributed energy resources and flexible demand”

Supervisors: Pierre Pinson (DTU Elektro), Yi Ding (Zhejiang University)

Supported by EcoGrid EU

Industrialized countries are marching forward to aggressive renewable energy targets, to reduce CO2 emissions and to reduce dependence on imported fossil fuels. The current market framework may not allow these ambitious targets to be met. One aspect in particular, electricity demand, is too rigid for system operators to compensate for fast variations in wind and solar production.

The EcoGrid EU project aims to design a new real-time electricity market and implement it on the Danish island of Bornholm. Electricity prices will be sent to customers every five minutes, allowing them to change their demand and reduce their electricity bill. New market rules, control methods and physical infrastructure will be needed, as well as further understanding of pricing problems and system reliability.

Countries will be able to meet their commitments of higher renewable energy integration, while transmission system operators will be able to maintain reliability.

Publications (selection):

Y Ding, S Pineda Morente, P Nyeng, J Østergaard, EM Larsen, Q Wu (2014). Real-time market concept architecture for EcoGrid EU — A prototype for European smart grids. IEEE Transactions on Smart Grid 4: 2006-2016

image Stefanos Delikaraoglou

“Modeling of market-based cross-border exchange of balancing power”

Supervisors: Pierre Pinson (DTU Elektro), Kai Heussen (DTU Elektro), Juan Miguel Morales (DTU Compute)

Supported by the BPES project (EERA/ForskEL)

Significant growth of renewable energy sources (RES) has been seen in recent years and further strong development is expected throughout Europe. Power generation from many RES, e.g., wind power, is largely variable and only partly predictable, which certainly affects the operation of the electric power system and it increases the need for balancing resources. Meanwhile, the wide penetration of RES challenges the efficiency of the existing electricity market. Both operation strategies and market rules have to be revisited in order to effectively accommodate a high proportion of RES in our energy supply.

This project focuses on the development and evaluation of predictive dispatch strategies for intra-hour balancing aiming to increase the flexibility capabilities of the power system and to avail new balancing resources in the markets, such as demand response. In addition, new market coupling mechanisms that promote the effective cooperation on balancing services between different power systems will be studied.

The results of this project will contribute towards an improved regulating market design and balancing operation of the power system with high shares of stochastic RES.

Publications (selection):

S Delikaraoglou, TK Boomsma N Juul (2013). Optimal charging of electric drive vehicles: A dynamic programming approach. in: Grid Integration of Electric Vehicles in Open Electricity Markets (ISBN: 978-1-118-44607-2): 113-128

image Chunyu Zhang

“Market design and network planning for active distribution grid”

Supervisor: Pierre Pinson (DTU Elektro), Yi Ding (Zhejiang University)

Supported by the Danish Research and Innovation Platform iPower

The common use of demand response, electric storage and an increment penetration of distributed generation (DG)based on renewable energy resources into existing distribution network brings high complexity of system operation and network planning. It also leads to many issues regarding new market design, regulation, congestion, reliability and etc. The conventional passive network management is not suitable in the new environment, which need to be changed into active network management. Active management of distribution network enables the distribution system operator (DSO) to dispatch network in an integrated efficient manner.

The objective of the PhD project is to propose a DSO market design for active distribution grid, which can motivate the efficient utilization of demand response and DG and reduce electricity price for end customers. The objective also includes the development of incentive market mechanism, which can mitigate the daily peak load and motivate grid upgrade for ensuring sufficient capacity.

The results of this project will contribute to develop a robust market design for active distribution grid, which will be the basic frame of normal operation. Moreover the necessary emergency functionalities needed in situations where operation is no longer normal must be developed for maintaining the reliable system operation.

Publications (selection):

Q Wang, C Zhang, Y Ding, G Xydis, J Wang, J Østergaard (2015). Review of real-time electricity markets for integrating distributed energy resources and demand response. Applied Energy, in press

S You, J Hu, K Heussen, C Zhang (2013). Analytical framework for market-oriented DSR flexibility integration and management. Energy and Power Engineering 5: 1367-1371

C Zhang, Y Ding, Q Wu, Q Wang, J Østergaard (2013). Distribution network expansion planning based on multi-objective PSO algorithm. Energy and Power Engineering 5: 975-979

image Qi Wang

“System-wide socio-economic and reliability impact of active management of distribution grids and distributed energy resources”

Supervisors: Pierre Pinson (DTU Elektro), Juan Miguel Morales (DTU Compute), Salvador Pineda (KU), Peter Meibom (Danish Energy Association)

Supported by the Danish Research and Innovation Platform iPower

It is foreseen that distributed energy resources and demand response programmes will increasingly affect the operations and planning of Danish distribution grids, while similar phenomena will also be observed in other countries. Such distribution grids are at the interface between the transmission system, where electricity is exchanged through a wholesale market, and the final consumers. Usual assumptions and simplifications made at the transmission level are there not valid anymore. There is a crucial need today to better model and optimize what is happening at the distribution grid level, to also further understand impacts on socio-economic and reliability aspects.

The objective of this Ph.D. project is to build a simulator “DG+” that emulates the efficient operation of local resources in a given distribution grid, while considering its interface to the wholesale energy market at the transmission level—represented here by WILMAR and OPTIBA, state of the art models developed by DTU and others through previous projects—at different time scales. DG+ is to account for the technical constraints imposed by distributed energy resources, while exploiting the flexibility of final consumers.

The PhD will develop such of model for distribution grid operations integrating distributed energy resources and demand response, based on cutting-edge stochastic and robust optimization techniques allowing a compromise between economic optimality and system reliability.

Publications (selection):

Q Wang, C Zhang, Y Ding, G Xydis, J Wang, J Østergaard (2015). Review of real-time electricity markets for integrating distributed energy resources and demand response. Applied Energy, in press

C Zhang, Y Ding, Q Wu, Q Wang, J Østergaard (2013). Distribution network expansion planning based on multi-objective PSO algorithm. Energy and Power Engineering 5: 975-979

image Tiago Soares

“Energy and ancillary services management considering high penetration of renewable energy sources”

Supervisors: Pierre Pinson (DTU Elektro), Hugo Morais (DTU Elektro)

Supported by the Technical University of Denmark, as well as the Danish project “5s” (Danish Strategic Research Council)

In a future power system with significant penetration of renewables, one is to rethink the way energy and ancillary services are exchanged through the market place. The main purpose of this PhD project is to develop and simulate energy and ancillary service management methodologies suitable at both transmission and distribution levels, considering the large penetration of renewable energy sources. The development of market mechanisms and strategies for the future electricity markets are one of the main aspects of this PhD project. In this scope, there are three essential goals that will be studied in this PhD project: the development of an energy and ancillary services joint market model based on stochastic optimization approaches; study of ancillary services coordination at different levels of operation and between areas/regions of the network; and the development of producers portfolio definition based on different behaviors and strategies to participate in the joint market.

Publications (selection):

P Faria, T Soares, Z Vale, H Morais (2014). Distributed generation and demand response dispatch for a virtual power player energy and reserve provision. Renewable Energy 66: 686-695

image Christos Ordoudis

“Towards market models for integrated energy systems”

Mentor: Pierre Pinson (DTU Elektro)

Supported by Center for IT-Intelligent Energy Systems (CITIES)

The deployment of renewable power production units in power systems is on the rise due to the worldwide environmental concern and economic incentives. This development has introduced significant variability and lack of predictability in power systems which as consequence have increased the requirement for system flexibility. Due to the intermittent nature of renewable energy generation and in order to acquire flexibility, it is important to couple various energy system infrastructures and networks (e.g. electricity, gas and heat) that are highly interconnected and compose the total energy system. This results in large-scale complex networks that need to be properly modeled in order to accommodate even higher shares of renewable energy.

The research is focused on the development of market designs and mechanisms for the operation of multi-carrier energy systems. Each system’s own complexities as well as interactions between them have to be taken into account in order to achieve the coupling of different markets and to propose new market structures for the efficient operation of integrated energy systems. Towards that goal, various optimization and decision making under uncertainty techniques are required as well as reliable forecasting in production and demand side. The development of such market structures will eventually result in the efficient coordination of various energy systems and will facilitate the integration of renewables in the energy system.

Publications (selection):
C. Ordoudis, P. Pinson, J.M. Morales, M. Zugno (2015). Stochastic unit commitment via progressive hedging – Extensive analysis of solution methods. IEEE PowerTech 2015, Eindhoven, The Netherlands

Guest PhD students

image Lazaros Exizidis (University of Mons)

“Wind power predictions and their use in energy markets”

Supervisors: Francois Vallee (University of Mons), Zacharie de Greve (University of Mons), Pierre Pinson (DTU Elektro), Jalal Kazempour (DTU Elektro)

Supported by the GREDOR project, funded by the Public Service of Wallonia.

The increasing environmental awareness along with the limited availability of fossil fuels has led to political decisions, in an international level, for the replacement of conventional energy sources by renewable ones. Wind energy, being not only efficient but also economically attractive, becomes a promising alternative which, however, comes with the drawback of wind’s stochastic nature. Wind power predictions are, therefore, mandatory in various time horizons serving different purposes such as operational planning, control, investment decisions, energy trading etc. On the other hand, the various system actors are characterized by contracting interests especially when it comes to wind power trading. A framework for defining these interactions and the use of wind power forecasting is the subject of this ongoing PhD research.

Publications (selection):

Z De Greve, C Daniels, L Exizidis, F Vallee, J Lobry (2014). A review of time series models for the long term modeling of wind speed in distribution network planning. CIRED Workshop 2014, Rome, Italy

L Exizidis, Z De Greve, F Vallee, J Lobry (2014). Construction and reduction of scenario trees for the day-ahead prediction of wind speed. IEEE Young Researchers Symposium 2014, Gent, Belgium

L Exizidis, Z De Greve, F Vallee, J Lobry (2014). Construction of aggregated scenario trees for the day-ahead prediction of wind speed. EWEA Annual Event 2014, Barcelona, Spain


image Guillaume Le Ray

“Big data analysis for energy systems”

Mentor: Pierre Pinson (DTU Elektro)

Supported by a blend of sponsors…

A recent trend in the development of energy system is the increasing availability of massive quantity of data at possibly high spatial and temporal resolutions. One may think of remote-sensing data for wind and solar energy applications (weather radars, lidars, sky imaging), or simply of smart meters measuring the consumption of residential households. There is much knowledge to be gained through these massive amounts of data for those able to extract relevant features, dissaggregate signals , etc. The overall purpose of Guillaume’s research is explore various ways to do so, being both methodologically sound and of practical relevance. In parallel, his project is to look at ways to produce space-time scenarios of renewable energy generation (both wind and solar power) in very high dimensions (say, for up to 40.000 sites and dozens of lead times)…