Embedding smart energy technology in built environments. A comparative study of four smart grid demonstration projects

Transcription

1 Embedding smart energy technology in built environments. A comparative study of four smart grid demonstration projects Tomas Moe Skjølsvold, tomas.skjolsvold@ntnu.no Marianne Ryghaug, marianne.ryghaug@ntnu.no This paper appeared in Indoor and built environment where it was published online before print on July 21, This is the paper as it appeared when it was accepted, but before copy editing by the publisher. Please cite as: Skjølsvold, Tomas Moe, and Marianne Ryghaug. "Embedding smart energy technology in built environments: A comparative study of four smart grid demonstration projects." Indoor and Built Environment (2015): doi: / X Full, formatted and copy edited text available here: Abstract The smart grid has become important in energy discussions. Inspired by science and technology studies (STS), this article compares four smart grid demonstration projects that engage households in Norway, where much activity has been triggered by a mandatory roll out of smart electricity meters by We ask how local actors across different sectors interpret the potential of smart electricity meters, and how different understandings lead to diverging innovation strategies. The result is four case studies illustrating how smart grid set-ups are constructed, and how new smart technologies are socialized into pre-existing localities or not. The four case studies highlight: (1) A constellation of healthcare and the electricity industry, resulting in a focus on simplicity and welfare. (2) A merger between the construction industry and the electricity industry, exploring how passive houses, PVsystems and smart technology influences collective practices. (3) How local authorities, research, and industry interpreted smart grids differently, resulting in a focus on greening, grid optimization and user flexibility. (4) How organizational and disciplinary conflicts led to alienation from the smart grid concept, hindering socialization. This diversity within a small country suggests that smart grids should be treated as situated technology rather than a catch-all silver bullet. Key words: Smart grid, demonstration projects, practice interventions, innovation, SCOT 1

2 Embedding smart energy technology in built environments. A comparative study of four smart grid demonstration projects Introduction Currently, buildings consume around 30% of all energy on a global level.(iea 2014) Thus, buildings represent a tremendous opportunity for reducing global energy consumption and in turn reducing climate gas emissions. The energy used for heating, cooling, operating household appliances, and other electronic devices inside buildings also represents critical challenges to vital societal infrastructure. Typically, energy use is unevenly distributed across the day, the week or the year. For example, the hectic domestic activities of the morning hours or during dinner lead households and their inhabitants, our main focus in this paper, to use much more energy in such periods than at night. Such clustered activities tend to form electricity consumption peaks. These peaks represent challenges to the electricity industry and the electricity grid. Peaks increase the costs of operating the grid, and they increase the need for back-up generation. This is often expensive, carbon intensive and strengthens fossil fuel lock-in dynamics. Ultimately, peaks in the grid might cause failure and blackouts. Thus, there are many good reasons to study and understand the energy consumption practices of households, and to gain knowledge about the dynamics of practice change. How could we change energy consumption patterns? Increasingly, actors in electricity intensive societies are advocating and researching the implementation of smart technologies connected to new, smart electricity grids as a possible solution.(ding et al. 2014, Fadaeenejad et al. 2014) Visions of what the new, smart energy future will look like typically contain technical elements like digital electricity meters, sensors, smart household appliances, new feedback technologies e.g. in-homedisplays and new types of home automation combined with new tariffs.(verbong, Beemsterboer, and Sengers 2013) Further, visions of smart energy futures often contain some sort of electricity storage, e.g. a battery, an electric vehicle or thermal storage, as well as new roles for householders as prosumers active participators and co-managers of the energy system, producing and selling electricity through the grid, e.g. through solar panels or other micro-renewable installations.(katzeff and Wangel 2015, Haunstrup Christensen, Gram-Hanssen, and Friis 2013) The combination of new information (feedback), new pricing schemes, and new technologies are typically expected to make energy users behave more rationally, which in turn would result in more efficient energy consumption. In practice, this is often an expectation of reduced consumption, or an expectation that consumption will shift to other periods of the day. This paper deals with the role of households in the Norwegian smart grid transition. The Norwegian interest in the smart grid is partly manifested by the regulation of AMI automated electricity metering infrastructure. In 2011, the Norwegian directorate of water resources and energy (NVE) passed a regulation demanding that all households should have advanced electricity meters installed by In Norwegian policy debates, this implementation has been considered a key stepping stone towards a full-blown smart grid implementation. The expectations amongst decision makers for reduced energy consumption, active user-participation and altered practices related to energy consumption are, and have been, quite high.(skjølsvold 2014, Frøysnes 2014, Ballo 2014) This dominant understanding, however, has not been without critics. Prior to the implementation of the 2

3 regulation, part of the electricity industry were vocal opponents to the development.(throndsen 2013) In this paper, we are interested in how implicated Norwegian actors prepare for the employment of smart, digital electricity meters in These meters allow for two-way communication between electricity customers and grid operators. Empirically, we study four Norwegian smart grid demonstration projects, where industry actors experiment with new arrangements based on smart electricity meters combined with other technologies under realistic conditions. In all cases, private households are at the centre of the experiment. Thus, we explore policy-driven innovation in industries that in Norway have traditionally been conservative, in the sense that they have tended to favor well proven concepts and materials at low cost, often developing new solutions and business models only when pushed in this direction through regulation.(ryghaug and Sørensen 2009) As called for by previous research, our study moves beyond an optimistic focus on the benefits of smart grids and their meager beginnings with the roll out of smart meters (Throndsen 2013, 1832), to focus on specific and localized opportunities and challenges in encounters between new, smart technologies, households, and various types of industrial actors. The question then becomes how the actors respond to, pick up, understand and make sense of the political push to install smart meters, and more specifically how they use the generic smart meter technology as a catalyst for different innovation and business strategies that span across a range of industries. Beyond implementing smart meters: some theoretical considerations In Norwegian policy discussions on smart meters, there has been a tendency amongst key decision makers to argue that the installation of smart meters will lead to changes in electricity consumption patterns. During the winter of 2002, for example, Norwegian electricity prices were higher than usual. Jens Stoltenberg, a social economist who has served several periods as prime minister of Norway for the labor party, was quoted in the newspaper Dagbladet, arguing that smart meters combined with new tariffs would solve the problem: Electricity becomes unnecessarily expensive, because everyone uses electricity at the same time. Covering the consumption peaks becomes expensive. It is important to award smart consumption. All you have to do, is install smart meters and make new tariffs (Ellingsen 2002) (our italics) While this is undoubtedly anecdotal, and competing voices have been raised(tjeldflåt and Kolbeinstveit 2006, Tjeldflåt and Vingås 2004), ample analysis has been carried out in the past to suggest that Norwegian policy actors have had somewhat naïve assumptions about what smart meters alone can do(ballo 2014, Frøysnes 2014, Skjølsvold 2014, Throndsen 2013, Löfström 2014), and how the smart meter roll-out is expected to result in practice changes to the benefit of society. Arguably, there has been a flavor of technological determinism in the debate. In its extreme form, this view depicts technology as an essentially autonomous entity, which develops according to an internal logic and in a direction of its own. Technological change is understood as beyond social influence, and implementation of new technology is perceived to change or mould the social realm in the image of the technology. A classic example of a successful critique of such a stance can be found in Raymond Williams famous discussions from the 1970s about whether or not television had changed the world.(williams 1974) 3

4 In this paper, our ambition is to contribute to the growing literature focusing on the social or sociotechnical aspects of smart grid development.(christensen et al. 2013, Goulden et al. 2014, Hargreaves, Nye, and Burgess 2010, 2013, Haunstrup Christensen, Gram-Hanssen, and Friis 2013, Naus et al. 2014, Strengers 2013, Verbong, Beemsterboer, and Sengers 2013, Wolsink 2012, Nyborg and Røpke 2011, Throndsen 2013) We shall study empirically how four Norwegian smart grid demonstration projects have used smart meters to build radically different approaches and strategies when it comes to the development of smart grid solutions for households. We find it fruitful to draw theoretical inspiration from Science and technology studies (STS).(Bijker, Hughes, and Pinch 1987, Jasanoff et al. 2001, Sismondo 2011) In particular, we build on the tradition describing the social construction of technology (SCOT).(Pinch and Bijker 1984, Bijker 1995, 2010, Pinch 1996) This line of thought describes how technology development takes place under specific local, social and historical contexts, and underscores that it is never given a priori how a piece of technology will be interpreted, how it will be used or understood, or the degree to which it will be successful. Instead, SCOT stresses that technologies are subject to interpretative flexibility. This means that different social groups involved in the technology development process can have radically different understandings of what the same technological artefact is, what it can do and how it should be used. Advanced electricity meters, for example, can take on different meanings for different social groups such as policy makers, electricity grid companies, the building industry and the users of the technology. Consequently, the fate of such meters is not yet clear, at least not in Norway. Interpretative flexibility is often strong and highly visible in new technologies, and has been identified in earlier studies of smart grid development.(nyborg and Røpke 2011) Once interpretative flexibility has been observed and accounted for, SCOT theorists tend to highlight how technology design and development is also concerned with processes of stabilization and closure. How can diverging views be brought together, controversies closed, and understandings of technologies stabilized? With this as a backdrop, successful technology development and implementation is seen as action involving much more than the mere making of technological artifacts. Technical qualities are never enough in themselves to ensure success. Instead, actors must work to embed, anchor or socialize technology into pre-existing socio-technical arrangements.(bijker and d Andrea 2009, Skjølsvold 2012, Sørensen 2013, Heidenreich 2014) In standard sociology or social psychology, the notion of socialization is used to denote the process whereby individuals find their place, and become embedded parts of collectives such as families, school classes or societies.(macions and Plummer 1997) This is a process where the entry of the individual into the group on the one hand changes the social group, while on the other hand becoming part of the group changes the individual. Thus, it is a transformative process, and a process of what would often be labelled co-production in the STSliterature.(Jasanoff 2004) In this article, we build on the same kind of reasoning while exploring the process of technology socialization and embedding into four different local settings. All four cases are examples of efforts to explore how actors in and around the electricity industry could move from politically forced encounters with generic smart electricity meters, to the production of some variant of what the involved actors believe a future smart grid will look like on the household level. Thus, we might observe the interpretative flexibility of multiple concepts and technologies (e.g. smart meters, smart grids, users, flexibility etc.), and efforts to introduce, embed or socialize specific interpretations of these concepts into different specific pre-existing settings. This means that we are interested in the transformative process, whereby the individual, isolated smart meter technology is made 4

5 social, and, through this process, is potentially stabilized as part of a smart grid solution. We are also interested in how project participants and key stakeholders imagine and interpret other key concepts of the smart grid discourse. How do they make sense of and imagine users and how do such imaginations influence the local smart grid strategies? In summary, we want to explore the development of the smart grid at household level, in a range of different local settings. Methodology This paper builds on qualitative studies undertaken at the four main Norwegian smart grid demonstration projects. They all involve users at household level. Our aim was to investigate the formulation of smart grids in different localities and contexts, and to study smart grid development, by different actors, with different meanings, for different reasons in different locations. To achieve this goal, we visited the four sites in question. We identified the key personnel involved in the demonstrations, and also tried to identify those who had been involved during the start-up of the projects. We carried out some initial interviews and preliminary discussions via and phone, before we conducted ten on-site in-depth interviews. Several of these were group interviews. In total we interviewed 16 persons. We were also given quite extensive guided tours so that we could experience some of the involved technology ourselves. All interviews were recorded digitally and transcribed in verbatim. We have consent from interviewed parties for using quotes from the interviews in research papers. Following this, the data was carefully analyzed. Since we wanted our analysis to be firmly anchored in the empirical data, we carried out a coding loosely inspired by grounded theory.(strauss and Corbin 1997) Coding here, refers to an iterative process where the data is read and re-read multiple times to identify patterns of themes and topics. These were then grouped, labelled, or coded along themes such as innovation, challenges, technology, user imaginaries/user rationality, involved actors etc., in order to build relatively coherent case studies. Thus, our analytical categories emerged from our interpretation of the data at hand, as opposed to a more deductive approach where we would impose categories and assumptions on the data a priori. Thus our approach is inductive.. For the sake of clarity we will now give a brief account of the political regulation which is the basis for many of the activities explored in this paper. Embedding smart energy technologies in Norwegian households through smart grid demonstration projects Norwegian electricity grid companies are responsible for electricity measurements. It is their task to implement advanced meters by The regulation requirements in terms of smart meter functionality are quite limited. The meters must be able to measure the electricity consumption every 15 minutes as a minimum. Further, the smart meter interface should be standardized to cater for communication with other equipment based on open standards. In terms of feedback to the users, the operators must present measurements online to allow users to compare prices, consumption and costs over time. The grid companies should facilitate the connection of in-home displays for interested customers. Thus, the regulation is relatively open ended, and the authorities have stressed that it is now up to the grid companies to make the most of the situation in terms of finding innovative and lucrative solutions to the benefit of the companies, the users and society. The underlying vision of the regulation has been based on socio economic assumptions about the increased rationality of the grid operation once the technology is in place. For households, the 5

6 argument has circled around user rationality, and the notion that what users are currently lacking to optimize their behavior is information and correct price signals. The authorities also foresee great potential for innovation and new business solutions in the interface between new industries such as information and communication technologies, electricity and construction.(skjølsvold 2014, Ballo 2014, Throndsen 2013) While the motivation and goals of the project participants involved in the four case studies differ, they all pointed out that they would not be involved in smart grid activities if it were not for the regulation. While the regulation was implemented in 2011, Norwegian electricity grid companies and have known about the regulation plans since Thus, they have been given sufficient time to experiment, and learn from these experiments. To us, the diversity of the following case studies illustrate that ample time is needed, both to find potential benefits and business solutions, feasible ways to embed the technology locally, and to come to terms with potential technological obstacles. Demo Lyse: Smart grids for simplicity and health A common trait for three out of our four case studies was that the smart grid trials were initiated when local actors picked up and acted on challenges or opportunities produced by specific local resources or existing social arrangements. This was clearly articulated by our respondents in the project called Demo Lyse. This project is located in Stavanger, a city on the western shore of Norway. Stavanger has rapidly expanded for many years, mainly because it hosts many of the off shore industries involved in Norwegian petroleum activities. This has generated wealth and many well-paid employment opportunities, resulting in a substantial population growth and an increased pressure on the electricity grid. Åshild Helland, who is the head of the AMI-project in the Lyse Grid Company explained: I think we are highly influenced as a company by being in this region. You know, we have always had to deal with rapid changes. It is a tremendously expansive region. For a long time we have been the region in the country with the highest number of new buildings per year. So we have always had to be pro-active and expand our grid. Thus, there was a perceived need in this company to do something to address this situation. As the quote indicates, the historically favored solution has been to expand the grid. However, expanding the electricity grid is not by default the same as developing smart grid solutions. In the Demo Lyse project the choice to move in the direction of a smart grid demonstration project was sparked by the observation that the region faced not only technical supply problems, but that there were also a set of regional-specific social issues that digitized energy technology combined with smart meters could help solve. Through innovation workshops and dialogue with several regional actors, amongst them a hospital and other health care organizations, an alliance was forged between representative of the electricity sector and healthcare, where project participants from the different sectors sought added value. Dagfinn Waage, who is the head of the Lyse innovation department, explained how they approached stakeholders in the healthcare sector to discuss how the potential of digitizing, automating and simplifying home services with basis in the smart meter and related equipment could help them: When we started up we had several sub-projects. And the welfare project, which more or less has defined our [smart grid] pilot, was based on real user stories from nurses who visited the homes of patients. We asked them, what are the biggest challenges you see? Write 6

7 down three relevant stories. And then we will build a pilot, use technological services and see if we can help solve them. Thus, the push from the authorities for finding relevant benefit in the smart meters led Lyse in the direction of health-related services. This might be interpreted as an early attempt of socializing and embedding ideas related to smart energy technologies in a setting where it is not necessarily intuitive that smart meters or smart grids can play a role. Waage explained that he believed that it was no coincidence that this particular development happened in the Stavanger region. Rather, he believed that the regions specific demographic challenges, and the related regional healthcare sector needs were vital in shaping Lyse s venture into smart grid demonstration and experimentation at the household level: We will have a boom of senior citizens in this region unlike any other region in Norway. And this will happen earlier in this region than in the rest of the country. Meanwhile the number of available hands [workers e.g. in institutions] will be on decline, the work force will be reduced. We saw this challenge [the coming senior boom ] already in For the healthcare sector, smart energy technologies represented not only an opportunity to reduce energy consumption and shift loads, it constituted a technological opportunity to make simple home control and automation technologies for the elderly, or for disabled persons. The idea, then, was that such technologies would make it easier for senior citizens and disabled persons to live in their own private homes, rather than in a healthcare facility. This was seen as beneficiary for life quality, but it was also argued that this would be economically rational, because retrofitting a house with smart technology was seen as substantially cheaper than permanently placing inhabitants under the formal care of hospitals, nursing homes or other health care facilities. The prospects of positive effects on the energy consumption were seen as a possible added value. Thus, the smart grid demonstration project emerged from observations about pressure on the electricity grid, as well as the chance to move into a new business area where the primary customers are elderly people and by extension, the healthcare sector (so-called welfare technology). These interests were combined with user oriented tests of home automation systems which were controllable with various smart devices such as phones, tablets, in-house displays or physical switches. All of this intended to shift consumption patterns and to avoid consumption peaks, and with the goal of increasing the ease of living a simple, comfortable and healthy life at home. In this way, the actual demonstration project might be interpreted as a localized technology designed to stabilize, or to close potential interpretations of what the smart grid could eventually become at household level in this particular local setting. An interesting aspect of this project was that the combination of an interest in welfare technology with ordinary electricity consumers influenced the perception of human rationality, and the link to electricity consumption, smart technology and ideas about what constituted good design. The interviewees highlighted that most people were not really interested in smart energy technologies; people just wanted simple, comfortable lives. This led the project towards a focus not only on flashy gadgets such as in-home displays and apps, but also on the development of very simple solutions. Dagfinn Waage explained: We want to help the end user. Make it easy for them. Help them become energy efficient. The elderly need this both for their health, their electricity bill and for simplicity s sake, but I need it, too! 7

8 Thus, Waage imagined the users in the Lyse project as quite far from the relatively common descriptions of homo economicus or resource man(strengers 2013), a character who eagerly awaits price signals to adjust behavior. No, he stressed, users were comfort creatures interested in simplicity. Hence, linking smart grids, smart homes and welfare technology led to the establishment of a new technological norm. The ideal now, was that what had previously been described as welfare technology should not be the strange cousin of ordinary technology. Rather, it was postulated that all technology should in principle be so simple that it could also work as welfare technology. One example of this was a set of very simple physical switches that could be programmed to control various aspects of a household s electricity consumption. For instance, one could simply set up a home or an away mode. Another opportunity was to create a night or day mode etc. It was argued that this made it easier to save electricity e.g. by getting rid of stand-by consumption and that it was simple enough to not alienate people unfamiliar with ICTs. Further, it was considered a safety measure for elderly people, who could rest assured that their household was safe for a good night s rest, simply through the push of one button. Figure 1: example of simple scenario switch Waage explained that the users of the project had received the strategy of designing for simplicity very positively: We can log what [technologies] our users use, and what they do in the project. So we know if people use their ipads or iphones, or if they use our simple, hardwired scenario-switches. And what we see is that people love the simple switches. Even the ipad is too complex. You have to pick it up, and push one or two buttons. And this is simple much simpler than [using] a PC! So when I hear these energy companies talking about making a website so that people can log on and check their consumption, I say come on, just forget it, people can t be bothered to do this [ ]. It s like universal design. It s a good thing. If it s simple enough for a disabled person or for the elderly, it s simple enough for me and the rest of us. In other words Demo Lyse was anchored in an implicit criticism of what was seen as common sense in much of the smart grid discourse, namely the degree to which technology users would be interested in engaging new technologies actively and thereby changing their energy consumption practices. This is a contrast to others who have shown that the electricity industry typically envisions 8

9 a techno-savvy, economically motivated gadget freak as their user: a resource man.(strengers 2013) Demo lyse became an exploration of the opposite assumption, namely that people are not very interested in their electricity consumption, and that the successful embedding and socialization of the technology at household level, requires a strategy which focuses on simplicity. Regionally, the socialization strategy seems to build on specific regional social and demographic challenges, creating synergies that are not necessarily intuitive. Demo Skarpnes: Passive houses and new energy practices The second case we have studied is located in the south of Norway, a few kilometers outside of the relatively small town Arendal. In this case, the interest in experimenting with smart energy technologies emerged from an alliance of a different kind. Interviewees from the electricity company Agder Energi explained that their close cooperation with the building industry had initially sparked their interest in experimenting with smart metering and other smart home technologies in ordinary households. The background for their interest was that the construction company Skanska was building a new neighborhood of so-called plus-houses. This was considered a pioneering project, partly conducted for the potential learning involved, but these houses were also to be sold on the ordinary retail market for a considerable premium price, targeting early adopters of new green technology. The houses should qualify as passive houses and they were equipped with solar panels, to allow for production of electricity for self-consumption and to sell electricity to the grid. Geothermal boreholes and solar collectors produced hot water for space heating and showering, as well to be used in in hot-fill washing machines and dishwashers. The houses were also equipped with smart meters and smart home technology, and state of the art ventilation systems. This was a very uncommon and novel technical setup for a Norwegian neighborhood, which means that it could eventually serve as a real-life laboratory or a sort of natural experiment where the relationship between new technologies and energy related practices were explored. So far so good, at least from the perspective of the construction company who was the entrepreneur standing behind the scheme. However, the construction of this neighborhood was not only a technical novelty, it raised numerous questions about how people would live inside these new types of buildings, and in turn what their lives would mean for the electricity consumption of the buildings. This was a particular challenge for the electricity grid and its owners in Agder energi. Their concern regarded how the new technologies and new building standards, combined with potential new emerging energy related practices would affect the electricity grid. Rolf Erlend Grundt, who works as an electricity grid developer in Agder energi explained how what he perceived as a uniformly positive discourse around the new smart technology neglected to take into account the actual grid: When people think of passive houses, plus houses and smart homes everyone thinks about saving energy. But what about the effect? Will the new houses allow us to reduce the dimension of the aluminum wires that we [the grid company] put into the house? Well, we don t know, but actually many things indicate that no, we cannot. You have all sort of new [and automated] equipment. An induction oven, new types of pumps that circulate air. And suddenly one day they might all start at the same time. Does this happen four days a year, or half the winter? We simply don t know, but we need to know! 9

10 To us, these reflections are interesting, because they suggest that the industry participants were involved in unpacking the socio-technical practices that constitute energy consumption. They wanted to reduce peaks. As we see here, however, the particular socio-technical practices that lead to peaks are not always well understood. The fear, in this case, was that technology intended to reduce or shift peaks and to change the energy related practices in buildings, might achieve this, but that they might also create undesired consequences such as new peaks. Much of the technical discourse surrounding smart grid solutions in Norway make jumps from relatively abstract concepts such as end-user flexibility to the idea of shifting or shaving peaks, without discussing the practices that make up electricity consumption, and thus also produce consumption peaks. For example, it is quite common to talk about flexibility as an input factor in a systems model. The interest in unpacking such phenomena has recently increased amongst social scientists.(buchanan, Russo, and Anderson 2015, Powells et al. 2014, Nicholls and Strengers 2014) Demo Skarpnes is an interesting case, in this respect, as it attempts to unpack what is referred to as end user flexibility in practice. In our view, this way of exploring and developing a particular smart grid variant for households is promising, because it looks at humans and technologies in a dynamic way. It underscores how parallel and tangential technology developments will impact the future of the smart grid. Electric Vehicles, for instance is increasingly thought of as a challenge to the capacity of the grid, and a potential storage unit in smart grids. How new technological configurations pan out in terms of energy practices is an empirical question as pointed out by Agder energy, and in our view it is essential to always treat it as such. One example of this type of experimentation in Demo Skarpens came with the combination of photovoltaics, new types of air pumps, ventilation and heating systems, and extensive home automation. While the setup was expected to reduce the overall electricity use, and to result in general load shifting, there were unexpected and uncharted consequences. Bjarne Tufte, who was head of innovation in Agder energi explained: There are many new construction technologies involved [in these buildings], such as circulation pumps [ ] and we want the solar panels to power those pumps. But then there might be a mismatch between production and need, and we are interested in that. But measurement-wise we are not interested in monthly readings here, we need to look at each individual minute, or each individual second even. Some of the equipment might be extremely fluctuating, turning on and off several times every second. And the results of this collectively, we simply don t know. In terms of technology embedding and socialization, it can again be fruitful to think about this on a regional and on a household scale. Regionally, the neighborhood could be constructed in this way partly because of natural conditions such as good solar measurements in southern Norway, and because of social elements such as a high demand for new homes. In actual households, prospective green technological frontrunners would serve to test how the technology would be used, and what kind of practices which emerged, or to use another metaphor how the new technologies would become domesticated.(sørensen 2005, Sørensen and Lie 1996) The intricate set-up became an arena to explore how the combination of renewables with fluctuating production, new construction practices and smart grid technology would influence electricity demand peaks, distinctly different 10

11 from the focus on user simplicity and welfare technology found in the previously mentioned case, Demo Lyse. Demo Hvaler: Greenness, grids and users In the two previous cases, smart grid solutions for households have been embedded in two quite different regional settings, as well as into quite different collectives with a very different outcome. The third case, Demo Hvaler, as we will see, also adds to this image of diversity in its own peculiar way. The Hvaler demo project illustrates how regional authorities could play an important role in crafting smart grid strategies. Here, the municipality has been central to the project development. Further, Hvaler is known as a summer vacation paradise for many people in Norway. Therefore, it also contains a special kind of built environment consisting of many second homes (so-called hytter) for the development of smart grids. The second homes are typically used most intensively during the summer months. Another specifically regional trait of Hvaler is that it can be very cold during a few days in the winter. Thus, there is usually great pressure on the electricity grid in a few intense periods of the year. The island also has favorable conditions for small scale and (distributed) local renewable energy production, both for small-scale solar and small-scale wind power. The roots of the demonstration project can be traced back to a regional networking dinner in the county Østfold, where by coincidence, the Mayor of Hvaler municipality was seated next to a representative from the research and industry network NCE Smart energy markets. NCE Smart is a cluster of research, innovation and business activities, which receives various types of governmental support to further the smart energy transition process. NCE Smart energy was planning to establish two smart grid demos, but had not yet decided on a location. However, during the dinner party discussion the NCE representative successfully managed to convince the Mayor of Hvaler about the positive traits of smart grids, and smart electricity consumption. It was informally decided that one of the projects would be based in Hvaler. In an interview the mayor, Eivind Normann Borge explained: We were talking about everything and nothing. And then he started talking about this demo project where two Norwegian municipalities would be pilots. And he asked me if Hvaler could be interested. He had already told me a lot, at least the main principles of it all [the smart grid], and I immediately though that this was a chance I could not let go. So I thought I ll say yes, and let the municipal board butcher me afterwards. I think about this as a way of building our reputation, you know, that we are a frontrunner when it comes to thinking about energy and sustainability. So I said to him; do not talk to anyone else, Hvaler will do it! One way to interpret this is to read the efforts of the representative from NCE as an early attempt at socializing the technology into Hvaler municipality, through interesting a key representative from the local authorities with a vision of what a smart grid could be. For Hvaler this was a welcomed approach, as they were already searching for some way to get involved in green issues. After making formal agreements, the demo project was set up with three primary participants: The NCE cluster, the electricity provider Fredrikstad energi and the local authority Hvaler municipality. The three participants had distinctly different interests in participating in the smart grid trials. There were different technological opportunities as well as a range of economic, political and cultural conditions, which were mobilized in the construction of both socio-technical visions for the future, and hands on smart grid demo development. 11

12 The electricity provider and owner of the grid interpreted the smart grid solutions in domestic buildings as a tool to strengthen the security of supply through the distribution grid. They had some concerns regarding capacity during summer peaks, but they were extremely concerned about capacity during a few hours of the coldest days of the year when the very old grid at Hvaler was (already) in trouble. Vidar Kristoffersen in Fredrikstad Energy explained: For us, we want better utilization of the distribution grid. Today, the load in the grid is extremely skewed. This load is tied to a few hours of the year when it is extremely cold. We use a lot of electricity for heating, you know Fredrikstad energy highlighted that they had two strategies to achieve better utilization and strengthening of the grid. The first was a long-term investment plan to increase the capacity of the distribution grid. This, however, was seen as achievable only within a timeframe of years, since the required technology was expensive. Thus, they wanted to mobilize the smart home technologies, home automation and advanced metering tested in the Hvaler demo to stimulate electricity users to provide their flexibility in the short run. Vidar Kristoffersen at Fredrikstad energy explained: We want better utilization of the capacity in the grid, and we are very interested in seeing how the end-users can influence this whole game. What is this end-user flexibility really? Do we need new technology, Smart house controllers? Do we need new incentives through a new pricing structure and can this help us balance [the consumption]? For us these types of investments are very short-term. If we think smart houses, maybe 4-5 years, gadgets for wireless communication, that kind of thing Thus, the electricity company clearly interpreted the potential implementation of a range of smart services and technologies as an intervention meant to change electricity consumption practices, stimulate flexibility provision and thereby optimizing grid utilization. In this sense, they imagined the role of users differently from what we have seen in the other projects mentioned above. They hoped that the users would be active, and that they would not only be active consumers of electricity, but that they would use the buildings actively to provide flexibility. The three main demonstration participants had different, but complementary goals. Hvaler municipality wanted the benefits from a green reputation, but they also welcomed new forms of commerce and economic activity. Further, as they learned more about the technology they also saw opportunities in the production of services, such as health and welfare technology, quite similar to what we found in the Demo Lyse project. Fredrikstad energy was primarily interested in the distribution grid, but in the short run, this meant a focus on users. Finally, NCE smart energy was primarily focused on the users. Since the actors could interpret the technologies into their own frames of reference and use them as means to pursue different individual goals, the socialization of smart grid solutions for households in this setting was relatively successful. As mentioned, the key concern regarding users in this project dealt with the notion of flexibility. It was an outspoken goal of the project to achieve flexibility through active user participation. This meant that the technology installed was considered an intervention, which was meant to stimulate practice change, also changing the role of the users and their buildings. Bernt Bremdal, who is a professor associated with NCE Smart explained: 12

13 We are not only concerned with the technical stuff. We are interested in user engagement. How do you recruit your users? On what conditions? We compare our work to many demos around Europe. And many of them have gone very far to get users involved. And once the demo is over, the users are fed up. They cannot be bothered to pay attention [ ]. And in reality, what we care about are concrete results. And the question is how you can understand these things so that you can relatively quickly develop something that is sustainable both economically and environmentally in the longer run in the sense that you integrate flexibility into future plans. If you do a demo and show that some people are extremely eager at this point in time, you are in trouble if you are not 100% certain that they will also act the same way in a real operating situation Thus, the actors involved were quite reflexive regarding their own project, and how it related to actual flexibility provision outside a demo setting. Our interviews with project participants in demo Hvaler also illustrate that the concept of user flexibility is in itself subject to significant interpretative flexibility. Currently, the project is recruiting new users who are interested both in providing flexibility by changing their consumption as well as taking on a prosumer role through installing photovoltaics. The target group is a community of second-home owners, buildings that are vacant for much of the year. Bernt Bremdal discussed the notion of flexibility, by rhetorically asking what the interest of people was: What do you think they [people] value? Lifestyle, comfort, convenience [ ]. The grid industry always wonders; how much should they have to pay for flexibility? But the debate is completely flawed, because they don t really consider that flexibility as valuable. They think that people have to participate This is interesting because it questions what is often taken for granted, at least in much of the Norwegian smart grid discourse, namely that people will want to offer their flexibility in return for economic gains. However, the project participants at Hvaler looked at their own community, and found that one particular social group already did many of the things typically associated with flexibility. These were the owners of second homes. Many of the owners of second homes were already users of a service called My second home (Hytta mi), which was a digital communication infrastructure providing services quite similar to what is often envisioned when it comes to the smart grid. Bremdal explained: They [the company Hytta mi ] are already inside the second home. And they provide two main services that they market extensively. The first is surveillance of the second home while you are not there. They will make sure that the temperature is ok and that nothing freezes. This has to do with securing what you own, your values. The second aspect is having the second home heated before you arrive, so that you don t have to carry firewood and freeze for the first hour [ ]. And people are willing to invest around kroner in this system, plus an additional 1000 kroner a year for the service. And in a sense, this is exactly what we want to do, but with other gadgets. They have customers and they make money. So we use that as a reference, and we also use the same equipment to see if we can build on top of an existing business model. The idea, then, was to use the vacant second homes to balance the grid through automation and remote control. Through this, one envisioned that the need to ask the inhabitants of more ordinary 13

14 domestic houses at Hvaler to change their practices would be reduced. This strategy could be read as an attempt to make the implementation and socialization of the smart grid/smart house technology smoother in the eyes of prospective users. By configuring the ordinary users (the residing electricity customers) as potentially negative towards changing their practices, focusing on the second homes was seen as a promising strategy. Demo Hvaler clearly illustrates the interpretative flexibility of smart grid solutions for households (and second homes). Ideas about what problem the technology should solve, how it would solve it and who would be involved differed between project partners. One could imagine that this degree of flexibility could result in controversies, and that lack of closure and stabilization would make socialization of the technology difficult. In this case, however, interpretative flexibility seems to have provided a quite productive leeway. It allowed different project participants to pursue different goals and construct different visions of the future, to learn different things and to formulate diverging strategies. In turn, the visions fed into the design of solutions, e.g. for using second homes as flexible load. This confirms what a long-standing tradition of studying the performativity of future expectations and visions have argued for years namely that visions and imaginations of futures can influence present-day actions(borup et al. 2006, Brown and Michael 2003, van Lente 2012, Skjølsvold 2014) Demo Steinkjer: Business meets science The fourth demonstration project focusing on smart grid solutions for Norwegian households is yet another story. The project is based in Steinkjer, a small town in the middle of Norway. The town has close relations to the main Norwegian technical university in Trondheim, the Norwegian University of Science and Technology (NTNU). It was the informal links to research and development that sparked what would eventually become Demo Steinkjer. Jan Foosnæs, advisor at Nord-Trøndelag elektrisitetsverk (NTE) explained how it all started: That is actually an interesting story. I was head of the grid department at that time. And we were going to run a very small trial with smart meters [ ] just install smart meters with a few customers to get a little experience. And then there were some people at NTNU who heard about this [ ]. And together with the head of strategy and business development, they decided to do something more with this. So NTNU got involved, there was teaching and research and stuff like that. It was a pretty big thing the first year Thus, the electricity grid company, NTE originally had quite modest ambitions. University researchers picked up this idea, and were able convince the NTE leadership to significantly up-scale and increase the ambition of the project. Thus, the dynamics and conditions for technology embedding were quite different in this case; from in the three we have so far seen. As already pointed out, the original intention of NTE was simply to gain some experience with smart meters. Therese Troseth, the current leader of Demo Steinkjer, pointed out that it had not been trivial that the initiative of establishing the demonstration project had come from outside the electricity company s own organization. Within NTE, many employees did not identify with what was considered as a development imposed on them by the external actor NTNU. What needs did NTE have that such a demonstration project could help solve? In Troseth s words: 14

15 Well, the organization of this demo has been up for debate many times. It is very different from the other Norwegian demos. There, the demos have emerged from concrete needs in the grid company In terms of organization, the demonstration project was established as a holding company outside the other departments of NTE. Thus, where other demos had emerged as grid companies responded to local socio-technical particularities, and found ways to grow out of the electricity grid and production companies, Demo Steinkjer had to find a way to grow into the organization. Troseth explained this as follows: It was not exactly clear how it would work. The department of strategy- and business development hired many of their own people to work with this, and many people with the wrong competence. So there was a lot of frustration, and eventually people there started quitting, and it fell apart Thus, there were clear challenges of embedding and socializing the demonstration of household smart grid activities into this setting. On the one hand, this underscores the importance of organizational anchoring of demonstration projects. On the other hand, it also points towards not underestimating ownership to processes and alienation from technology development projects. This could also serve as a word of caution in the years ahead of the national smart meter roll out. Finding strategies that genuinely provides the electricity grid companies with ownership tp this process, seems an important stepping stone towards successful smart grid strategies for households. The question of ownership and alienation in this demo was not only an organizational challenge, but also a challenge of interdisciplinary work. Jan Foosnæs explained: Well, in the infrastructure department people tend to have more of a power engineering type of background. While the people there [the strategy and business department] probably had backgrounds more in the direction of economy and that type of thing Thus, built on the interest from university researchers, this demo was established in Steinkjer. It created an initial buzz, but many in the electricity grid company felt alienated from the project according to our interviewees. As the company economy tightened, funding to the smart grid demonstration project was more or less cut-off, and the project crumbled. The Demo Steinkjer project was eventually re-started, but now in a more cautious manner. As this case demonstrates, the technology might be ready-made and simple enough. However, its interpretation and possible stabilization, the embedding and socialization of the technology is no trivial matter. In this case, many in the electricity grid company saw the smart grid demonstration as a spearhead in what was also a disciplinary and organizational battle. As we have seen in this and the three demo projects outlined above, technology appropriation is a complex and political process that is influenced by the initial roles, commitments, identities, knowledge and expectations of a range of groups and individuals.(russell and Williams 2002) Success requires interactions across different divides in work environments: management and workforce, professional and occupational groups, functional divisions in an organisation and often across different organisations. 15