Science Day
Valuable expert feedback at Erasmus Energy Forum Science Day
Valuable feedback for developing research projects, and ideas to improve the chances of those ideas being implemented in practice came from the audience of experts during Science Day at the two-day Erasmus Energy Forum in June. This academic part of the Erasmus Energy Forum presented in-depth scientific work about using energy informatics and management to encourage energy efficiency in city and port environments. It was held in the Erasmus Paviljoen at Rotterdam School of Management, Erasmus University.
The expert international audience included not only academics but also practitioners from companies such as Capgemini, Toyota, Vattenfall, Hewlett Packard, Deloitte, Enexis, Stedin and Greenchoice among others. Science day was moderated by Dr Ksenia Koroleva and Dr Arthur Carvalho of RSM.
Science Day was lively compared to other conferences of its type, commented Dr Ksenia Koroleva. “There was a lot of interaction between research and practice and interesting discussions with both,” she said. Science Day was, by comparison with the preceding Business Day, characterised by a high level of discussion from speakers and audience, all of whom are experts in the field. Presenting authors gained valuable feedback to develop their papers for conferences and journals, as well as ideas to improve the chances of their ideas being implemented in practice.
Erasmus Energy Science award
The Erasmus Energy Science Award was presented for the paper, which in the opinion of the judges, showed the most scientific rigour, potential for innovation and applicability to the real world. Winner of the award, and a cheque for € 500, were Serkan Özdemir and Rainer Unland for their paper The broker strategies of a winner agent in Power TAC. Runner up were Frits de Nijs, Matthijs T. J. Spaan and Mathijs de Weerdt for their paper Best-response planning of thermostatically controlled loads under power constraints.
Digitising capital and energy efficiency
Keynote speaker Professor Richard Watson from the Terry College of Business at the University of Georgia spoke about energy efficiency and the digitization of capital. Creating a sustainable society requires the input of information into the energy equation and the digitisation of capital and capital conversion processes, he said; types of capital include natural, economic, human, organisational, social and symbolic. The digitization of capital is an evolution in the role of information systems in enabling society to maximize energy efficiency and the use of capital. Each type of capital can be enhanced through digitization, he explained.
Science Day concluded with an academic panel discussion entitled Efficiency or flexibility: What is more important in the future grid?
Energy grids will benefit from increased efficiency and flexibility – making the most of availability – thanks to information systems, and there are currently plenty of examples of such strategies being worked on by researchers as well as practitioners. Flexibility seems to become more important with the increasing amount of renewable energy in the network. But energy efficiency might also decrease flexibility in supply. This, in turn, might lead to problems with grid stability during times when large amounts of renewable energy are being generated. The panel debating the conundrum were: moderator Prof. Gilbert Fridgen from the University of Bayreuth; Prof. Eric van Heck from RSM; Prof. Richard Watson from Terry College of Business, University of Georgia; Prof. Vedran Podobnik from the University of Zagreb and Peter Reffeltrath from the Energy Research Centre of the Netherlands, ECN.
The panel’s ideas included:
- In order to enable efficiency and flexibility, consumers need to change their behaviour; this will probably be influenced by informational or financial incentives
- Market economies cannot expect central co-ordination, but will have to find mechanisms that incentivise decentral optimisation such as micro grids and energy co-operatives.
- New technologies must tackle both questions – efficiency and flexibility. It’s not a question of efficiency-versus-flexibility, but rather a matter of developing new technology and changing the behaviour of users.
- Skills from management, computer science, operations, optimization and behavioural sciences are needed – as well as the need for interdisciplinary specialists – in order to tackle these questions.
Abstracts from Science Day
Research papers were presented in four sessions, structured according to topic. A total of 12 papers were presented during Science Day, and were followed by questions from the audience. The papers are published in full on the Erasmus Energy Forum webpages, here; each one is described from its abstract below:
Paper session 1: energy policy and behavioural aspects
Shaping the future energy market – making energy demand more flexible
Johannes Schmidt and Lutz Kolbe
Demand response (DR) has experienced only moderate expansion in Germany, mainly due to regulatory requirements. Given the rapid expansion of renewable energies, however, it is important that in future smaller, flexible consumers are also be able to participate in DR and respond to the intermittent supply of energy. Schmidt and Kolbe conduct a case study with a fleet operator that applied DR programmes for its fleet of electric vehicles and used the results to derive recommendations for redesigning the energy market in order to make more use of flexibility options on the demand side. Their findings indicate there are several parameters that policymakers can use to lower the entry barriers to energy markets.
The capacity challenge: a perspective on the potential of energy information systems to drive policy
Edward Hunt and Richard Orwig
Hunt and Orwig restate the problems facing the electricity industry as the result of the misalignment of incentives for consumers seeking to minimise electricity costs and a regulated system obliged to deliver security of electricity of supply. They call this the ‘capacity challenge’. They advocate that the information systems (IS) discipline plugs a gap in information available to regulators and other policy makers, which would enable them to understand how to encourage consumers to provide price signals which will permit economically efficient outcomes. Watson’s manifesto of 2010, about the role of IS in sustainable development, proposed as his research question 8: “What data should be reported by an energy information system to inform governments’ energy policies?” The authors hope they provide, if not the units of measurement that the data should have, then at least a refined suggestion for the role that the data should play.
Bridging social innovation and entrepreneurship: business models for local renewable energy
Antonia Proka, Matthijs Hisschemoller and Derk Loorbach
With local renewable energy initiatives booming throughout Europe and beyond, and with the advent of an empowering technology that is expected to make the electricity grid smarter, the question arising is whether these bottom-up initiatives have the potential to radically change the energy system. Will new social practices and lifestyles used by these social innovations succeed in diffusing and consequently changing the energy landscape? The value creation mechanism of these initiatives is crucial for making them more attractive to the broader public. To help understand and evaluate this mechanism, and with the parallel ambition to develop ideas on how to improve it, the paper from Proka, Hisschemoller and Loorbach merges two linked conceptual issues: First, the concept of niche, and second, the concept of business model. They look into business models from Dutch renewable energy initiatives as a vehicle and subject of innovation. Drawing upon cases in the Netherlands, they study the way these initiatives develop and emerge, and whether and how business models may contribute to their growth and expansion.
Paper session 2: economics of energy markets
Energy co-operatives as an application of micro grids: multi-criteria decision support for investment decisions
Gilbert Fridgen, Anna Gründler and Milos Rusic
The future of energy generation is expected to become increasingly decentralised. Many customers already act as more than demand units; they also act as energy producers (‘prosumers’) and are thus active participants in the energy market. The development of energy co-operatives in recent years emphasises -this observation. Information and communication technologies (ICT) enable management of energy co-operatives by incorporating data from smart meters and energy generation. This paper uses multi-attribute utility theory (MAUT) to provide support for decisions about investing in new supply units for energy co-operatives based on micro grids. Literature analysis shows energy co-operatives have economic, ecologic and social goals. However, each co-operative weights the identified goals differently according to its preferences. The researchers define measures for fulfilling goals, including available data, calculating and aggregating each criterion into an index reflecting the total utility of each new investment, for example the utility of an additional photovoltaic plant in the co-operative. Furthermore, Fridgen, Gründler and Rusic test their model with real-world data. The results indicate that the artefact provides useful support for decision-making for energy co-operatives.
Designing markets for congestion management on distribution grids
Rens Philipsen, Mathijs de Weerdt and Laurens de Vries
In an electricity market, controllable loads allow users to react to price signals. If the share of renewable electricity generation increases, the electricity price may not correlate with available network capacity. In such a scenario, the electricity price must be supplemented with a signal that indicates the availability of network capacity. This is especially problematic on a distribution grid, where the grid topology and limited edibility of load impede direct application of methods used on the transmission grid. Philipsen, de Weerdt and de Vries investigated the performance of congestion management mechanisms for distribution grids, paying specific attention to the presence of multiple electric vehicle (EV) aggregators. They define the problem of charging EVs in terms of a job-scheduling problem to explicitly account for the (inter)temporal constraints that distinguish such a problem. They describe the auction in terms of empirically evaluated market designs for optimistic valuation by aggregators, showing that forward markets perform adequately, while real-time markets are often infeasible. They plan a more refined analysis in future to include uncertainty in the arrival of EVs and in the environment.
The influence of the energy spot market on ‘on the micro grid markets’ (OMMs)
Verena Dorner, Gilbert Fridgen and Rebecca Trick
The transition towards distributed energy resources (DER) for electricity generation has engendered new applications for information systems that enable the development of innovative business models for grid control and operation. Information systems enable the merging of DERs and energy consumers into micro grids. Micro grids increase efficiency, for example through shorter transmission distances. Cost savings from gains in efficiency can be distributed among micro grid participants via an ‘on the micro grid market’ (OMM) platform. However, due to differences in power plant parks, and fluctuations in the amounts of renewable energy generated, energy prices on OMMs are likely to differ from those on the spot market. With varying prices on both markets, from time to time the supply or demand side of the OMM may prefer to trade on the spot market – and that could cause it to stop functioning. The researchers developed an economic model to analyse the conditions under which OMMs are a viable concept for energy trading, and their results indicate that OMMs can in indeed function in competitive energy markets.
Paper session 3: technical aspects of the smart grid
Best-response planning of thermostatically controlled loads under power constraints
Frits de Nijs, Matthijs T. J. Spaan and Mathijs de Weerdt
Renewable power sources such as wind and solar are inflexible; demand must quickly follow supply to maintain an energy balance. Heat buffers that use electricity to maintain a temperature at a set point promise controllable demand and thermostatically controlled loads (TCLs), which have been shown to follow a power curve using reactive control. De Nijs, Spaan and Weerdt investigate overcoming temporary imbalances with the use of planning under uncertainty to proactively control an aggregation of TCLs. They present a formal definition of the planning problem under consideration, modelled using the multi-agent Markov decision process (MMDP) framework. And because they consider hundreds of agents, solving these MMDPs directly is intractable. Instead, they proposed composing the problem by decoupling the interactions through arbitrage. Decomposition of the problem means relaxing the joint power consumption constraint, which means that joining the plans together can cause overconsumption. Arbitrage acts as a conflict resolution mechanism during policy execution, using the future expected value of policies to determine which TCLs should receive the available energy. The researchers compare several methods to plan with arbitrage experimentally, and conclude that a best response-like mechanism is a scalable approach that returns near-optimal solutions.
Virtualizing balancing power: an energy-aware load dispatcher for cloud computing
Gilbert Fridgen, Robert Keller, Markus Thimmel and Lars Wederhake
Balancing mechanisms assure grid stability, and large-scale storage facilities (SFs) are especially well-suited for balancing purposes. However, their potential is largely set by geographic realities. On a transnational level, offering that potential to regions in need of balancing power rarely appears economically viable – an issue that is frequently related to the construction of power lines. Fridgen, Keller, Thimmel and Wederhake illustrate an early version of a design artefact giving remote balancing mechanisms access to a local balancing power market without deploying power lines. Utilisation of such cloud-scale data centres is typically very low (30-40 per cent) so they represent a cheap source of demand flexibility. The researchers let one data centre participate in an existing balancing power market while tying a second to a remote balancing mechanism. By doing so, the design artefact enables both load and balancing power to flow seamlessly between distinct power markets, contributing to grid stability and efficient utilisation of balancing mechanisms. In their extended summary, the researchers perform a preliminary evaluation of the design artefact based on real world data.
The broker strategies of a winner agent in Power TAC
Serkan Özdemir and Rainer Unland
In the light of governments’ energy transformation policies, the future electricity grid will encourage new participants such as local producers, storage units and interruptible consumers to the electricity grid – along with the challenge of sustainability. The Power Trading Agent Competition (Power TAC) is a comprehensive simulation platform to enable and verify various smart grid studies from the perspective of sustainability, as well as an annual competition in which autonomous agents trade in energy markets and make profits. Agent UDE won the Power TAC 2014 final as the newest agent of the competition, using an adaptive and reactive agent. The paper from Özdemir and Unland details the basic retail and wholesale market strategies of Agent UDE, as well as tournament results.
Paper session 4: electric vehicle charging and energy storage
Economic benefits of smart parking lots
Jurica Babic, Arthur Carvalho, Wolfgang Ketter and Vedran Podobnik
The ever-increasing number of EVs on the road promotes the idea of sustainable transportation by reducing CO2 emissions. This means there is also a growing need for charging stations. A potential solution to the need for charging stations is to transform traditional parking lots into smart parking lots, providing not only parking services, but also the possibility of charging and discharging EVs. The inherently complex and dynamic environment is a potential obstacle to the complexity of estimating the profit of such a smart parking lot after transformation and for the time required to recover the cost of the initial investment for the lot's owner. Babic, Carvalho, Ketter and Podobnik propose a simulation-based approach to estimate the owner's profit over a certain period of time using real-life data from existing parking lots, charging stations and the wholesale electricity market, analysing a variety of different investment strategies. Using a set of the most relevant ‘what if’ scenarios, they discuss the potential impact of smart parking lots on energy business.
Private providers of climate-neutral electric vehicle charging services – enabling remote access to self-produced energy at other peoples' charging points
Friedrich Chasin, Martin Matzner and Fritz Rettberg
Unresolved technological and societal challenges hamper the widespread practical adoption of EVs. Despite calls for systemic approaches, research tends to focus on specific obstacles in separation. This article reports two publicly funded research projects – both dedicated to developing a specific EV service offering – to co-operate and work on bundling the two services. The two offerings are providing public access to private charging spots, and transferring self-produced energy through public energy grids to other peoples’ charging spots. Chasin, Matzner and Rettberg argue that the value proposition of the single services can be elevated via bundling. This conceptual research describes use cases, processes, and supportive IT infrastructure needed to bring the service bundle into action. The contribution of this work lies, first, in discussing and outlining the bundling approach, which merges the outcomes of the two research projects. This approach can make a small step towards achieving the socially accepted goal of propelling EVs as a more sustainable means of transport. Second, the paper is novel for that it introduces a set of models, methods, and tools taken from the academic discipline of service science into the strands of research related to EVs.
Electric vehicles for energy storage – evaluating a business model
Tobias Brandt, Sebastian Wagner and Dirk Neumann
Aggregating the storage capabilities of EVs is generally considered to be a promising method of supporting the integration of volatile renewable energy sources into the power grid. Brandt, Wagner and Neumann analyse this concept from an economic perspective by developing and evaluating a business model for EV aggregation. Specifically they investigate the case of parking garage operators using the EVs located at their facilities to provide reserve energy for frequency regulation. The researchers evaluate revenues and cost structures using extensive data sets from the German market for frequency regulation, on battery states of charge at different times of day, and on occupancy rates of parking facilities. They find that given current market conditions, possible revenues are far inferior to investment costs for charging and IT infrastructure. Even if the operator installs these features to enable customers to charge their vehicles for a fee, it would be more profitable to focus on this service and charge vehicles immediately when they enter the garage instead of delaying charging for frequency regulation. Overall, the researchers conclude that a large-scale adoption of electric mobility may produce greater challenges to the power grid than is currently projected.
Poster session
A speed poster session allowed further research to be presented to the audience in addition to the research papers discussed in depth on the main stage. Researchers were allowed two minutes to make their presentations and the audience was invited to vote for the best one. The poster session award went to K. Valogianni, W. Ketter and J Collins for their presentation of the paper A multi-agent approach to variable-rate electric vehicle charging co-ordination.
Those taking part in the speed energy poster session were:
- Energy agents – foundation for open future energy grids
Christian Derksen and Rainer Unland - Real-time virtual power plant decision support
M. Kahlen and W Ketter - Developing a model predictive control scheme for building and district energy systems to optimize energy demand and consumption
Thomas Lawrence, Richard Watson, Maric Boudreau, Jason Perry, Jordan Coleman and Michael Karsten - The economic value of energy co-operatives and their role in the transition towards decentralized grids
A. Rieger, R. Thummert, G. Fridgen, M. Kahlen, and W Ketter. - TamagoCAR: exploring the effectiveness of pricing mechanisms to balance energy demand with a smartphone app
K. Koroleva, M. Kahlen, F. Lanz, and W Ketter. - Delft energy standards – providing society with knowledge on energy storage
G.A. Dorp van MSc, C.B. Davis PhD (RUG), Prof. G.P.J. Dijkema (RUG), Prof. F.M. Mulder (DUT) - A multi-agent approach to variable-rate electric vehicle charging co-ordination
K. Valogianni, W. Ketter and J Collins. - SMARTICITY – an energy feedback system for household consumers based on energy disaggregation
Carl Heckmann, Diaa Abdelaziz, Uemit Tepe and Alexander Maedche - To share a car or not: effects of psychological, mode specific and trip specific characteristics on transportation mode
Joshua Paundra, Jan van Dalen, Laurens Rook and Wolfgang Ketter.
