ORIGINAL ARTICLE

Constructing users in the smart grid—insightsfrom the Danish eFlex project

Sophie Nyborg & Inge Røpke

Received: 4 July 2012 /Accepted: 21 November 2012# Springer Science+Business Media Dordrecht 2013

Abstract The smart grid is promoted as one of thekey elements in a low-carbon transition in many coun-tries. In Denmark, the dominant framing of the smartgrid emphasises the challenge of integrating muchmore wind power into the electricity system and usingelectricity for heating (heat pumps) and transport(electric cars). In the process of radically transformingthe electricity system, strategic system builders needto align many forces, including consumers, who playan important role in the functioning of such largenetworked systems. System builders need to explore,for instance, whether and how users can be motivatedto be flexible in relation to moving electricity con-sumption over time. This paper reports on one of thefirst smart-grid-related projects in Denmark in whichconsumer aspects have been central and where poten-tials for flexible electricity consumption have beentested. The aim of the paper is to explore what canbe learned from such experiments and which rolesthey play in the construction of the smart grid. In thiscontext, the concept of the ‘aligned user’ is introduced.

Keywords Smart grid . Flexible electricityconsumption . User-oriented innovation . Alignment .

Low carbon transition

Introduction

Facing the great challenge of transforming the energysystem to a system based on renewable energy, manyactors—governments, business, local authorities, in-ternational organisations—promote the ‘smart grid’as an important element in the solution. In generalterms, the smart grid implies the application of infor-mation and communication technologies to make theelectricity system ‘intelligent’ in an effort to ease theintegration of intermittent energy sources like windand solar power and to improve energy efficiency. Inmore specific terms, the meaning of the smart griddiffers between countries, depending on the specificcombination of energy technologies and the organisationalset-up of the energy systems; and sometimes, differentnational actors frame the smart grid in different ways, inrelation to their particular perspectives. This paper dealswith the smart grid issue in a Danish context, where thedominant framing presently emphasises the challenge ofintegrating much more wind power into the electricitysystem, and using electricity for heating (heat pumps)and transport (electric cars).

When new provision systems are constructed andwhen old systems are radically transformed, manyforces have to be aligned. To some extent, the align-ment can emerge without active and conscious coor-dination, but usually some actors play the role ofstrategic system builders. As Hughes (1983) has de-scribed, the early system builders of the power systemwere aware of the need to interconnect a large number

Energy EfficiencyDOI 10.1007/s12053-013-9210-1

S. Nyborg (*) : I. RøpkeCenter for Design, Innovation and Sustainable Transition,Department of Development and Planning,Aalborg University, Copenhagen, Denmarke-mail: sony@plan.aau.dk

of diverse components into a seamless web—physicalartefacts, utility companies, manufacturing firms, aca-demic research and development laboratories, invest-ment banks, regulatory authorities, consumers,etc.—to realise their visions. Surely, this is also thecase today in relation to the smart grid. During the lastfew years, many actors within business, research, andgovernment have made a considerable effort to pro-mote the idea of a smart grid in Denmark, culminatingin the political decision to finalise a plan before theend of 2012 on how to implement a smart grid. Untilnow, system actors have concentrated on providing re-ports on the advantages and challenges related to thesmart grid, and on initiating technical research anddemonstration projects (Energinet.dk, Dansk Energi2010, Ea Energy Analyses 2010). In addition to techni-cal and economic issues, the reports emphasise the needfor exploring how electricity consumers can be motivat-ed to play their potential roles in the envisioned systemand how business models can be constructed that com-bine consumer motivation with the business interest inearning a profit. In only a few cases do the research anddemonstration projects include studies on consumeraspects, but this is about to change, since more studiesinvolving users are planned for the near future. In thispaper, we report on the eFlex project, one of the firstsmart-grid-related projects in Denmark in which con-sumer aspects are central, and discuss what can belearned from such projects and which roles they playin the wider construction of the smart grid.

It can be argued that all products and services, andthus all innovations, are integrated into larger systemsof provision and use, but this is particularly evidentwhen it comes to innovations in relation to largenetworked provision systems like the electricity sys-tem, where consumers play an important role for thefunctioning of the system. In the formulation by Shoveand Chappells, householders are not just consumersof, for instance, electricity—they own ‘the sensitivefingertips of existing infrastructure’, such as wires andcontacts, ground fault circuit interrupters, routers andswitches, and all the electric appliances in the home,which are literally part of the infrastructure itself;householders are thus co-managers who are implicatedin the routine functioning of the system as a whole(Shove and Chappells 2001, p. 57). Already in theearly history of the electricity system, it was a chal-lenge for providers to encourage consumers to behavein ways that fit better with system demands. The

nearly exclusive use of electricity for lighting causedhigh peaks and long periods with low demand, imply-ing ineffective utilisation of generating capacity. Toeven out demand, utilities thus actively promoted theuse of electric appliances. The period of manufacturingdemand was followed by a period of meeting demand:After the Second World War, electricity consumptiongrew rapidly along with economic growth, and thechallenge for utilities was seen to be ‘predict and pro-vide’. This gradually changed when the environmentalagenda emerged and questioned the societal rationalityof an ever-expanding system, which also required largeinvestments. Rather than just meeting demand, the con-cern for managing demand became more pressing. In aDanish context, the focus was on energy savings (e.g.labelling appliances and prohibiting installation of elec-tric heating in new houses) and promotion of combinedheat and power production (involving a duty to con-nect). The interest in reducing or shifting the peakthrough demand management came much later inDenmark than in many other countries because thetransmission and distribution grids had been constructedwith considerable excess capacity. But now, the time hascome to focusmuchmore on demandmanagement as anintegral part of preparing the functioning electricitysystem for the future.

The present interest in exploring and defining userroles in relation to the establishment of a smart gridhas emerged within a contemporary context in whichuser involvement and user-oriented innovation is pop-ular. Studies or direct involvement of users in relationto innovation projects can be motivated by variousconsiderations. When a new product or service isoffered, it is obviously important to assess whether itwill have a market, whether consumers can be seg-mented in different groups, and what price thesegroups are willing to pay. It is typically a task formarketing departments or consultants to consider suchquestions, which may also include the desirability ofvarious product features. There is a long tradition forsuch marketing studies, where methods usually in-clude questionnaires and focus groups. Users may alsoplay a direct role in the innovation process by testingprototypes or providing ideas for product developers.As emphasised in various strands of the literature,users have never been passive recipients of new tech-nologies, but contribute to their development through‘learning by doing’, ‘learning by using’, and ‘learningby interaction’, which provides more qualified

Energy Efficiency

feedback on user requirements than market transac-tions, and processes of domestication. However, feed-back from early users has often been part of thediffusion process—that is, after the initial productlaunch. Increasingly, such feedback is organised to beincluded earlier in the development process throughcooperation with lead users (Von Hippel 1988) orthrough experiments, tests and discussions with ordinaryusers (Heiskanen et al. 2005, 2010). In the case of leadusers, they are expected to have a personal interest in thedevelopment of a particular product and to be competentadvisors for developers, able to point out fundamentalproblems and assist with technical knowledge. Somealso argue that lead users represent needs that are ‘aheadof the market’ and thus anticipate the demand of futureconsumers later in the diffusion process. The interactionwith more ordinary users is expected to support thealignment of user practices and design options: Userinterests cannot be identified in advance, but have tobe constructed in relation to the technical options.

Sometimes, product developers cooperate directlywith users, while in other cases, professional media-tors are involved in the process. Schot and de laBruheze emphasise the importance of mediators inthe process of mutual articulation of demand andsupply and introduce the concept of mediation junc-tion as ‘the place at which consumers, mediators, andproducers meet to negotiate, articulate, and align spe-cific technical choices and user needs’ (Schot and de laBruheze 2003, p. 234). Their conception of mediatorsis broad and includes all sorts of associations related tospecific product groups (e.g. auto clubs), more generalconsumer organisations, marketing and testing agen-cies, retailers and groups of academics and consultantswho claim to provide knowledge on consumer inter-ests. Mediators take on the task of bringing ‘represent-ed users’ into the process, in addition to the ‘projectedusers’ whom innovators and designers try to inscribein their products, and the ‘real users’ who becomeinvolved sooner or later. In relation to user-orientedinnovation, it has become popular to include anthro-pologists as mediators, since they are considered tohave particular skills in accessing and interpreting theworld of users (Suchman 2013). The case study forthis paper exemplifies this trend, since anthropologistsare involved as mediators.

The case study is based on participatory observa-tion. The first author cooperated with the anthropolo-gists in the eFlex project, carried out some of the field

work and participated in analysing the empirical data.This provides the basis for both reporting on keyresults of the eFlex project and for broader reflectionson the role of such a project. Informed by the literatureon the role of users in systems and in innovationprocesses, the following research questions areexplored:

& What kinds of learning can the system buildersobtain from such experiments?

& Do the experiments serve other purposes thanlearning?

& Which roles are assigned to the users in theexperiments?

& Which roles do the mediators play in the process?& In which ways can such a project contribute to a

low carbon future? Can the project be counterpro-ductive in relation to this aim?

In the following, we provide first a little more back-ground on the Danish context for smart grid experi-ments, and then we present the case study: the purposeof the eFlex project, the methods applied in the projectand the results. In the final section, we discuss theresearch questions in relation to the learning achievedthrough the project, other roles of the project and theimplications related to a low carbon future.

The Danish context: two interacting systemtransformations

The smart grid transformation of the electricity systemdevelops in interaction with the process of liberalisationof the electricity trade, which was set in motion by EU’smarket directives in the 1990s. In Denmark, consumershave been able to choose their electricity supplier sinceJanuary 2003. Production and trade in electricity havebeen separated from transmission and distribution,which are natural monopolies. The transmission net isrun by the state-owned company Energinet.dk, whichhas the overall system responsibility, while the distribu-tion nets are owned by about 70 grid companies (2012),regulated by The Danish Energy Regulatory Authority(DERA). About 75 % of the wholesale trade in electric-ity takes place on the Nord Pool Exchange (owned bythe Nordic companies with overall system responsibility),where producers from all the Nordic countries (andGermany, UK and Estonia) sell electricity to traders andto firms using more than 100,000 kW h/year. The retail

Energy Efficiency

market is still nationally organised, but a Nordic retailmarket is being pursued.

Although nearly 30 traders sell electricity to Danishconsumers and smaller firms, competition in the retailmarket is considered to be limited (Konkurrence-ogForbrugerstyrelsen 2011). With liberalisation, each ofthe former local monopolies was transformed into agroup composed of a regulated grid company and acommercial trader, and the trader was assigned theobligation to supply customers in its old area for afixed price set by DERA if customers did not activelychoose another supplier. Nearly 10 years later, 83 % ofthe small customers still buy price-regulated electricity(p. 14); traders with the duty to supply usually haveabout 90 % of the customers in their own area (p. 17),and only 6 % of consumers have changed supplierduring the period 2008–2010 (p. 22). One reason isprobably that there is little to save by changing. Due totaxes and tariffs, the supplier’s electricity price consti-tutes a small part of the electricity price for consumers.Furthermore, the scope for competition is limited bythe fact that small customers pay a fixed price perkilowatt hour, independent of the consumption profileduring day and night (p. 15). Suppliers are thus notallowed to offer hourly payment that would give con-sumers the opportunity to save money by shiftingconsumption over time. A third reason relates to in-voicing: When customers buy electricity from a traderaffiliated with a grid company, they only get oneinvoice. In other cases, customers receive separateinvoices from the trader and the grid company, unlessthe trader makes cumbersome arrangements with thegrid company (p. 15). Finally, electricity is a low-interest area, and due to the supply obligation, con-sumers do not even have to bother with delivery.

These conditions are about to change due to thecombined pressure from continued market liberalisationand the smart grid challenge. A new model for payment(called the wholesale model) will be implemented, prob-ably in 2014, in which consumers will only be cus-tomers of the trader and no longer the grid company.The grid company then becomes a wholesaler sellingthe distribution service to the trader, and the consumerwill receive only one invoice. Change of supplier is alsoexpected to be eased in March 2013 by the establish-ment of a so-called DataHub, where all data on electric-ity consumption will be gathered to simplifycommunication between market actors and reduce entrybarriers. The process towards hourly payment is also on

the way. The first condition is the installation of remotemeter reading on an hourly basis. More than half ofDanish households either have or will have remotemetering in the near future, installed by the grid compa-nies, and political intervention can be expected to bringthis up to 100 % within a few years. Remote meteringand payment on an hourly basis combined with thepossibility of managing demand—smart metering—isthe key to involving consumers in the smart grid.

The smart grid appears both in relation to the over-all management of an energy system based on inter-mittent energy sources and in relation to demandmanagement at the household level (also at the busi-ness level, which is not the focus here). The need fordemand management occurs for several reasons. First,the expected increased use of heat pumps and electriccars may add to the traditional peaks in electricityconsumption and thus create local overload in thedistribution grids. Since extensions of the grid are veryexpensive, cost–benefit analyses indicate that consid-erable savings can be achieved by investing in a smartgrid that enables demand management (Energinet.dk,Dansk Energi 2010). Second, storage capacity inhouseholds may provide outlets for abundant windpower, and third, households may provide short-termregulatory services. The most immediate concern isthe avoidance of increased peak loads—an issue that isparticularly important for the grid companies, whichare responsible for strengthening the grid.

The interest in electricity consumers in relation tothe smart grid and smart meters thus emerges fromvarious perspectives. For instance, while grid compa-nies focus on the avoidance of peak loads, electricitytraders face the challenge of profiling themselves inrelation to customers by developing brands and differ-entiated services in spite of selling exactly the samebasic product. Simultaneously, there is a societal in-terest in encouraging energy savings and changing themix of energy sources. This variety of interests isworth keeping in mind when studying concrete exam-ples of smart grid projects involving users.

The eFlex project

Purpose

DONG Energy (in the following DE) is one of theleading energy groups in Northern Europe with

Energy Efficiency

headquarters in Denmark. The company procures, pro-duces, distributes and trades energy. Since the finan-cial crisis put a stop to the process of privatisation, thegroup is still a public limited company with 76 % ofthe shares owned by the Danish state. The eFlexproject is based in the ‘Sales & Distribution’ division,which consists of two separate companies—‘Sales’and ‘Distribution’—due to the liberalisation of theelectricity market. The project was commissionedand paid by ‘Distribution’, which owns and operatesthe distribution electricity grid in Copenhagen andNorthern Zealand, but daily management of the pro-ject resided in the Grid Strategy Department, a tech-nical unit under Sales & Distribution that was hired asconsultants by Distribution.

The electricity distribution companies are facedwith the scenario that Danes will have welcomed300,000 heat pumps and 600,000 electric or plug-inhybrid cars by 2025 (Energinet.dk, Dansk Energi2010). The eFlex project was initiated as an attemptto understand what it takes to make consumers movetheir electricity consumption to other times of day toavoid escalating peak loads and huge investments inexpanding the distribution grid. To ‘make consumersplay along’ and engage in a ‘partnership on peakshaving’, they have to be motivated, and since fewhouseholds take much interest in their electricity con-sumption, DE Distribution expected that consumers’price sensitivity might be relatively low. Therefore,DE Distribution aimed to explore what other incen-tives for flexibility were in play and how these couldbe mobilised in the change process. To investigatethis, DE Distribution hired antropologerne.com, asmall consultancy company working with user-oriented innovation, service design, organisationaldevelopment and communication. The consultantswere to conduct a user study that supported theother more technical part of the eFlex demonstra-tion project—the testing of new smart grid proto-type technologies for demand management ofelectric vehicles, heat pumps and domestic appli-ances in a number of households in DE’s distribu-tion area. Antropologerne.com was hired toinvestigate the assumption that customers’ pricesensitivity and their motivation for moving elec-tricity consumption—aided by the new smart homeequipment—would be strengthened if they devel-oped a new relationship to their electricity compa-ny and to electricity as a product.

The Grid Strategy Department is responsible for gridplanning with regard to improvements (automation),maintenance and investment planning. Their main inter-est in the eFlex project was to assess the ‘flexibilitypotential’ of the households, i.e. how much electricityconsumption could actually be moved away from peakhours—could they count on customers’ flexibility andcalculate when, where, how long and how much theywould be flexible. They were primarily interested incustomers who had either a water-to-water or air-to-water heat pump or an electric vehicle, as especiallythese two technologies are expected to contribute topeak loads in the future. The anthropological user studywas expected to identify the unknown ‘human’ param-eters, such as ‘flexibility-readiness’ and ‘acceptance ofsupply interruption’ in a complicated equation of thehousehold flexibility potential. This equation would,e.g. include knowledge on the m2 of the house, kilowatthour consumption, heat pump or not, insulation degreeand building year. The Grid Strategy Department wasalso concerned with how to categorise customers indifferent segments—i.e. what consumer types seem tobe very flexible and what characterises consumer seg-ments that are not. The hope was that these insightswould also provide knowledge on where DE distri-bution could geographically postpone investments inthe distribution grid and where not (Torntoft Jensen2011, p. 48). On the basis of these interests, theproject manager of eFlex and antropologerne.com,together, wrote a project design document describingantropologerne.com’s deliveries. These includedamong other the development of flexibility profilesof the users (segmentation), an evaluation of the testequipment, a mapping of ‘energy behaviour’ andappropriation of equipment, identification of motiva-tional factors for energy flexibility and an evaluationof the ‘flexibility promoting potentials’ of differentcommunication forms with the customers.

Method and framing of the eFlex project—the GWRequipment

One of the basic elements in the project design was thetesting of a home automation energy managementsystem (Fig. 1), which the company GreenWaveReality (GWR) had developed for DE based on aprevious user-oriented innovation study by theAlexandra Institute (Alexandra Instituttet 2010). Theequipment was included in the project on the

Energy Efficiency

assumption that visualisation of the customers’appliance-specific consumption and a new communi-cation interface with DE would provide the informa-tion, awareness and interest needed to encourageflexibility—as well as provide the ability to automatethe moving of consumption conveniently.

The kit that participants received and installed intheir houses consisted of a main unit called a ‘gate-way’, which is connected to the internet. The gatewaycommunicates wirelessly with a number of intelligentpower nodes that are controllable plugs with on/offcontrol. The users could control the gateway and thusthe power nodes via an online ‘portal’, which theyaccessed from either a computer or from an iPodTouch. Thus, if the users connected the power nodesto appliances around the house, they would be able tosee on the portal how much power each applianceconsumes throughout the day. They could turn themoff from the portal, or they could program certainpower nodes to turn off or on collectively at specifictimes of the day, and thus make e.g. an ‘out’ profile, or‘sleep’ profile’. Moreover, the participants agreed totransfer to hourly pricing, which follows the spot priceon the Nord Pool market, and they were also offeredvariable distribution grid tariffs (requiring special per-mission for this project). Accordingly, the next 24 h’dynamic prices, which were visible on the portal, andwhich the customers were priced after, were based ona combination of dynamic spot prices and variabletariffs and could differ between 1.50 kr (0.20 €) per

kWh to 4.30 kr (0.58 €) per kWh. Hence, the userscould utilise this information to construct certain pro-files or turn devices on/off individually at certain timesin periods when the price is low/high.

The project design included a group of householdswith a heat pump, a group of households with anelectric vehicle and a ‘control group’ of ‘ordinary’households without either. All three groups had theenergy management system just described. In the heatpump group, DE could reduce consumption—or‘optimise’—the heat pump externally, and this grouphad an extra feature on the portal they could use tofollow DE’s interaction with the heat pump. The par-ticipants had to indicate which minimum temperaturethey would accept in the home and where on a scalefrom ‘low’ to ‘high’ their ‘flexibility level’ was, whichmeant that the heat pump could be disconnected forperiods of 1–3 h. Likewise, the charging of the electriccar batteries was controlled externally by DE. Theusers had to specify through the portal at what timein the morning the battery should be ready and chargedand its minimum percentage level. Moreover, the usersalso had to tick off whether they prioritised consumingelectricity at the lowest price, or at a time when the shareof wind energywas highest in the grid’s energymix, or abalance of these alternatives. Based on the informationof minimum room temperature, the choice of flexibilityversus comfort, the prioritisation of price versus windand the daily price pattern (the el spot market is a day-ahead market), an algorithm calculated the period and

Fig. 1 The GWR equipment (image from antropologerne.com)

Energy Efficiency

time of heat pump interruption and downloaded theresults to each households’ gateway for in-housecontrol.

The user study was divided into three ‘loops’, eachincluding a round of fieldwork and preliminary analysis(Table 1).

Recruitment Participants for the study were recruitedthrough DE’s newsletter or were contacted directly,e.g. if they were already involved in another projectcalled ‘control your heat pump’ led by Energinet.dk.Antropologerne.com recruited the electric vehiclehouseholds through their own network and with thehelp of the Danish Electric Vehicle Committee.Participants were required to reside in DE’s distribu-tion area and consume more than 4,000 kW h/year.They were offered 1,000 DKK to participate in thetrial and were also allowed to keep the equipmentincluding the iPod after the trial had ended. DE orig-inally looked for approximately 50 households with anelectric vehicle, 75 households with a heat pump and30 households without. It turned out, however, to bevery hard to find that amount of electric vehicleowners in DE’s distribution area.

The eFlex pilots The eFlex participants’ homes weredistributed evenly over DE’s entire supply area, whichcovers a large part of mid- and northern Zealand. Ofthe 119 official test pilots—i.e. the main contact per-sons in the households—103 were men, mainly 40–59 years old, and a majority had families with two ormore children. It was emphasised that the projectincluded the household as a whole and not only thetest pilot. The participants were also among the morewell-to-do segment of the Danish population and oftenlived in detached or semi-detached houses. Theirdwellings were often large, i.e. 100–250 m2 and mostparticipants consumed 7,000–15,000 kW h/year; some

were also above 15,000 kW h. The eFlex pilots gen-erally had a high education, and many of them weretrained as engineers or economists or worked withinIT.

Fieldwork—empirical data

In total, the empirical data included 49 household visits,debates on PODIO, notes from different informationevents for the households, two user workshops, threeanalysis workshops with DE’s Team eFlex at the end ofeach loop, questionnaires on demographics and lifestyleissues and a ‘choose-a-profile’ exercise1 (Table 2).

Results

We include first a short summary of the technicalresults before presenting some examples of the expe-riences and observations from the householders’ ev-eryday life with the eFlex project and finally theconclusions from the user study.

Results from grid analysis

DE expects to publish a main report on both the techni-cal and behavioural findings from the eFlex study inautumn 2012. Results of the technical part of the projectoffered here are from the preliminary analysis presentedat a final conference for the eFlex project in March2012. The results only concern the heat pumps, as themeasurements from the electrical vehicle group werenot usable for any statistical purposes, since there wereonly eight participants (one resigned from the project

Table 1 Project design

Loop 1 Loop 2 Loop 3

Households included in thetrial (in total 119)

29 ordinary households 9 electric vehicle owners 55 heat pump owners26 heat pump owners

Number of households interviewed 16 ordinary households 9 electric vehicle users 15 heat pump owners6 heat pump owners 3 heat pump re-visits from loop 1

Field work period March, April, May 2011 September, October,November 2011

November, December 2011,January 2012

1 The first author of this article conducted 11 of the 49 house-hold visits and participated on PODIO, in workshops and ininternal meetings with subsequent analysis sessions. The finalwork with the results and writing of the report was done solelyby antropologerne.com.

Energy Efficiency

before time), and since it turned out that many of thevehicles were of a too-old model to be able to worktogether with the technical equipment. Optimisations ofthe heat pumps generally took place between 7 and 9AMor 5 and 7PM. Interruption of the heat pumps depended onthe flexibility choices (see earlier description), the house’sthermal behaviour and the household’s ‘living habits’.In periods with outdoor temperatures ranging from −15to −5 °C, 65 % of the heat pumps were still shut downafter 1 h, but this declined to 7 % after 2 h. With outdoortemperatures ranging from 5 to 15 °C, the same figureswere 85 and 35 %—i.e., the colder the weather the lessinterruption of heat pumps was possible.

Two interesting conclusions could be drawn fromthe technical analysis. The interruption period tends tobe too short compared to an average peak load in thegrid, which indicates that heat pumps may need to becascade controlled to ‘last’ an entire grid load.Secondly, the expected ‘kick back’ from releasing heatpumps to normal operation was not very clearly ob-served, although it was there. Since users’ ‘livinghabits’ in terms of using wood-burning stoves, open-ing windows, having guests, cooking, dressing, etc.had a significant influence on heat pump operationpatterns, the interpretation of results was difficult.

Finally, the ‘control’ household group, which onlyhad the GWR equipment, had on average saved approx-imately 10% on their kilowatt hour consumption duringthe project period March 2011 to February 2012. Thisfinding is consistent with other studies on the effect ofvisualisation of energy consumption and feedback,

which suggests that energy consumption can be reducedby 5–15 % depending on the kind of feedback provided(Darby 2006). A recent large study from the UK sug-gests that the top of this range may be far too optimisticand underlines that the savings that can be achievedthrough feedback vary greatly depending on period,customer group and type of energy (e.g. electricity orgas) (Raw and Ross 2011). The effect of feedbackthrough the portal was not calculated for the other twogroups, as DEwanted a ‘pure’ evaluation of the effect ofthe GWR equipment without the possible influencefrom also having a heat pump or an electrical vehicle.

Experiences and observations: everyday lifewith eFlex and use of GWR

In the final report, antropologerne.com organises some ofthe findings by applying the four phases of domesticationtheory: commodification, objectification, incorporationand conversion (Berker et al. 2006). The idea is to considerthe home automation equipment and the related services asa ‘wild animal’ that households try to domesticate. Wemust leave out many details in the description of thedomestication process here in order to concentrate on asmall selection of topics in two key phases. In some cases,we draw on examples from the first author’s interviews.

The first steps

In relation to the objectification process—the concretephysical placement of the GWR equipment—some

Table 2 Analysis was based on a broad range of field material

House-hold visits Each visit lasted four to five hours and included interviews, lunch or dinner with the families as well as a‘grand tour’ of the dwelling. Parts of the interviews and use situations were video recorded and photoswere taken of the user and the family, the dwelling, symbols of life style, the GWR equipment and theheat pump or electric vehicle, etc.

PODIO PODIO is a social media platform that combines text, images, video, etc. It functioned as a projectmanagement tool by which antropologerne.com, DE and the users could communicate and shareexperiences. Of the 119 eFlex pilots in the project, 114 of them registered on PODIO during the project

Arrangements Several events were held for the participants – among them a ‘futures night’, where the then CEO at DE,Anders Eldrup, came to talk to the participants about their important role in the future electricity system.Moreover, three information events before each loop were held, and a ‘question night’ in loop 1

Workshops A number of ‘analysis and user workshops’ were conducted after each loop in which preliminary findingswere discussed with DE as well as with the users. The users participated in two user workshops withfocus group exercises and interviews while three workshops were held with DE’s Team eFlex

Questionnaires and choosea profile

A questionnaire on lifestyle and demographics was answered by 97 of the 119 participants. Moreover, all119 users in the project were asked to place themselves in one of five user profiles that were constructedon the basis of the 49 household visits and a user workshop with 8 eFlex participants. 72 participantschose to do so, and the results were correlated with the results from the questionnaires

Energy Efficiency

participants faced integration problems. Setting up theequipment was a task for technical enthusiasts, and formany, it was a challenge to configure the power nodeson the portal. A specific problem at the project’s startwas that the power nodes had no ground connection,which meant that devices such as dishwashers, wash-ing machines and tumble dryers could not beconnected. This was unfortunate since these house-hold appliances consume much electricity and involvea form of consumption that actually could be moved.The problem with ground connection was later solved,but many kitchens also have in-built appliances withpower plugs that are not readily accessible anyway.

For some participants, the placement of the nodeswas also a challenge, and it was unclear whether thehome energy management system could really supportflexibility. One participant, for instance, walkedaround the house to find devices to plug into a powernode but concluded that most of the devices that couldbe plugged into the power nodes—e.g. lamps, TV’s,kitchen appliances, etc.—were devices you had to usewhen you had to use them. Similarly, some partici-pants argued that it would be too troublesome topostpone, e.g. vacuuming or working on house reno-vations with power tools to other times of the daywhen they might not have the energy, ‘inspiration’ ortime.

For heat pump owners, the set-up involved partic-ular decisions. A little more than half of the 81 heatpump users chose the price optimisation of heat pumpoperation, while the rest chose a combination of priceand wind content optimisation. Only one participantchose a pure wind content optimisation. However, thewind content optimisation resulted in heat pump inter-ruption that was difficult to understand. Because windenergy was often low at night, many heat pump inter-ruption programmes were set for interruptions at nightand not during price peak hours. Balance betweenprice and wind energy content was complicated tofind, and indeed to communicate to customers, andover time, more and more participants readjusted topure price optimisation. Many also chose the maxi-mum heat pump flexibility level, i.e. allowing DE toreduce electricity consumption for the longest time.The participants chose everything between 16 and20 °C as the lowest acceptable temperature in thedwelling, but the majority maintained the default valueof 17 °C as the minimum room temperature and acomfort temperature around 21–22 ° C.

Common applications

In the phase of incorporation, when the GWR equip-ment is incorporated into family practices, many partic-ipants concentrated on using the system in order to getan idea of their different devices’ consumption—e.g. thecost of brewing a cup of coffee. They were looking forthe greatest electricity consumers and were surprised athow much the children’s play station and fish tank ortheir hard disk recorder consumed. Consequently, theydeveloped a new habit of turning items off manually orthrough profiles when not in use, despite the hassle ofhaving to wait for the TV to warm up again, because itwas completely disconnected at the power node. Theequipment also showed when during the day or nightelectricity was cheap or ‘green’ and when to avoidexpensive tariffs. Accordingly, many participantsutilised their washing machine’s or dishwasher’s oldfashioned time delay mechanism and consistently post-poned doing their laundry until the weekends or nights,when tariffs were lower.

Moving laundry and dishwashing to nighttimeseemed to be the most frequently changed activityfor participants in the eFlex project. Most had alsoplugged in some appliances into the powernodes—usually lamps, TV sets, computers, internetrouters and gaming consoles. Moreover, some eFlexpilots plugged in chest freezers, refrigerators or elec-trical heating in bathroom floors and made profilesthat turned them off during certain periods of theday. Some participants explained that once they had‘played around’ experimenting with connecting andconfiguring the power nodes—or once they had givenup doing this—and started postponing, e.g. the wash-ing machine and dishwasher to nighttime, their com-mitment and activities related to the project decreased.

Flexibility in the family context

Domestication processes proceed differently dependingon family structures. This has implications for flexibil-ity. For instance, households consisting of singles orcouples with no children or pets are often more flexiblein their habits and electricity consumption and, e.g.more willing to compromise on their comfort, thanfamilies with (small) children or pets. As one retiredcouple explained, when the grandchildren were visiting,they were less prepared to keep a low temperature in thehouse and put on a sweater, especially because the

Energy Efficiency

smallest grandchildren crawl on the floor. Anothermother resisted the idea of washing at night becauseshe was concerned that the noise would wake up herteenage son, who had a room in the basement not veryfar from the washing machine. Yet another participantwas, e.g. upset that his aquarium fish had ‘been up allnight’: due to a malfunction the light in the fish tank hadnot been turned off by the GWR nodes. In general,families with children have a tightly coordinated every-day life, which means they have less flexibility poten-tial. As a father of two explained, it was hard to beflexible with, e.g. doing laundry, because the boys’soccer clothes had to be dry and ready for training andgames at certain times.

In some cases, the eFlex project gave rise to ten-sions within families. Several wives of enthusiasticeFlex pilots were irritated because their husbandsdemanded that they suddenly start running the dish-washer or washing machine at night. They wereconcerned that the clothes would be wrinkled fromlying in the machine all night and that the machinewould have to be emptied in the morning when theyhad to help the children get ready for school. Somewives also found the GWR equipment ugly; it did notfit very well into the interior decoration; they could notclean properly or they could suddenly not turn on theTVor brew coffee for their guests because their house-hold appliances were connected to the power nodes.These nodes would turn off at random or they did notknow how to control/manage them with the iPod. Inanother family, the teenage children were obviouslyannoyed with having their electricity consumptionsupervised and controlled by the father who wouldcheck on the iPod, which he brought to work, whetherthey played on the computer in the afternoon whenthey had promised to walk the dog or he would turnoff their computers through the iPod, if they playedgames past bedtime.

Complex user behaviour

The project gave rise to several findings that illustratethe complexity in the interplay between the companyand the users in relation to flexibility. In the context ofheat pumps, for instance, many owners also heatedtheir dwellings with other heat sources, like fireplacesor wood-burning stoves, which increased their flexi-bility potential. Some of the participants wanted theheat pump to be turned off for even longer than was

actually part of the project design because they couldjust use the fireplace if the temperature dropped no-ticeably. Thus, a participant with a large buffer tankand a fireplace was disappointed that he could not saveeven more from being flexible with the heat pumpthan the project actually allowed, and he explainedthat they never ‘felt’ the optimisations. In general,the participants experienced very little comfort lossfrom the optimisations.

In a few cases, participants were interested in takingcontrol themselves or redesigning the system. Somehad joined the project to ‘sniff’ at the smart griddevelopment, while others engaged in ‘system design’on a more micro-level. One of the electrical vehicleowners—a young man trained as an electrician—hadalso tampered with and rewired the power nodes him-self to give them a ground connection so he could plugin his refrigerator and freezer and start experimentingwith flexible profiles. Another participant proudlyproclaimed: ‘DONG control—I over-control’. He con-sistently turned off—or optimised—his heat pump forlonger times than DE. He set his heat pump to anindoor temperature of 27 ° C in periods with cheapestelectricity because he could store the excess heat pro-duced in his buffer tank and accordingly turn off thepump for long periods during the day, when electricitywas most expensive.

Conclusions—four main findings in eFlex

In the final user study report (antropologerne.com 2012),which did not include grid analysis, antropologerne.comsum up their results in four main conclusions.

Flexibility and innovation In a project like eFlex, thefuture is investigated at the same time as it is created.eFlex is thus not just a question of testing new tech-nologies and uses, but also a question of rehearsingnew relations and creating cooperation, learning andunderstanding with the customers. DE has not onlylearned about the customers, but also from and withthe customers.

Electricity is life Electricity is used to create life in thehome. It is used for both necessary household prac-tices like heating and for more ‘luxurious’ practicesthat are important for identity formation, such ashobbies and social life. Households seemed most will-ing to be flexible with regard to necessities, while they

Energy Efficiency

expressed a need to justify and legitimate hobbies, forinstance, by referring to environmental friendliness.

Motivation The user study resulted in several insightsconcerning motivation for flexibility:

(a) Flexibility potential: The user study identified thefour factors below (Fig. 2), which in practice play arole for household flexibility potential. Customers“must be reached differently through diverse mes-sages, concepts, products, services, interfaces andcommunication forms” (antropologerne.com 2012,p. 39) to realise the potential.

& Willingness to flexibility: The participants’willingness and motivation to move consump-tion is dependent on their general interests,attitudes, values and comfort habits, as wellas their relationship to technology, economyand the environment.

& Family composition: Flexibility depends onthe composition of the entire household. Carefor the wellbeing of, e.g. children or pets oftenleaves the participants less willing to compro-mise comfort or convenience or to changedaily habits and structures.

& Life situations: Interest in electricity consump-tion and flexibility can be triggered by changein life circumstances—e.g. moving from a flatto a house, acquiring a new electricity-consuming heat pump, refurbishment of thehouse, engagement in new hobbies or transi-tion from working to retirement.

& Household infrastructure and smart technologies:for example, degree of insulation, floor heating,buffer tanks and fireplaces in the home as well assmart home systems that can automate movingconsumption all influence flexibility potential.

(b) GWR equipment is a missing link and promotesflexibility: GWR equipment becomes a link thatconnects daily practices with the electricity con-sumption of the household. It creates a tangibleconnection to the electricity world and motivatesthe participants to be flexible, both by enhancingtheir interest in electricity consumption and pro-viding them with the tools to know what devicesto turn off or which practices to intervene in.

(c) Economy as motivating factor: The participantsare driven by multiple economic rationales,

including a ‘moral economy’ promoting the wishto do what is right and sensible—such asavoiding waste and optimising consumption.Variable prices and tariffs support this ‘feelingof doing right’. The participants do not necessar-ily know precisely how much they save whenmoving consumption; they just know that theyshould consume when the price is low—in effectmaking the price also a signal of right and wrong.

Five user profiles are identified The user studyresulted in five user profiles, which fall in the two maincategories of ‘the enthusiastic’ and ‘the interested’. Theparticipants were segmented according to their use ofthe GWR equipment in the home, their life values,professional background, knowledge of and relationshipto ‘the electricity world’ and their motivations for beinga part of eFlex. The different user profiles are motivatedto participate in the project and to provide flexibility fordifferent reasons. Moreover, different communicationstrategies apply to each profile. For example, ‘the tech-nical’ are greatly motivated to use the portal and PODIOto discuss technical issues with other users and DE,whereas the eFlex application for the iPod and informa-tion events are appealing to ‘the comfortable’.

Enthusiastic:

& The technical: technique-enthusiasts who are en-gaged in the ‘electricity world’; they have joinedeFlex to contribute to technological and societaldevelopment

& The economical: system-thinkers who have joinedeFlex to control and optimise the family’s energyconsumption

& The curious: people with an inquiring attitudetoward life; for them, eFlex is an opportunity tolearn new things about energy and electricity.

Interested:

& The participating: humanists who want to dosomething good for others; they are primarily inthe eFlex project for the sake of the environmentand the project’s greater cause

& The comfortable: appreciate comfort and con-venience in their homes; for them, eFlex is anopportunity to save money and do a good deedwithout compromising comfort or time for otherthings.

Energy Efficiency

Discussion and conclusion

Complex knowledge interests

On 7 March 2012, more than 150 people from thebusiness sector, research, media and policy making,including the Minister for Climate, Energy andBuildings, the Vice president in DE and the directorof the Danish Consumer Council, attended the confer-ence reporting on eFlex results.

Considering all the hype surrounding this event, theproject would seem to have been a great success, but acritical look at the course of the project reveals that thedesign and knowledge outcomes were challenged byhaving to accommodate many different actors’ knowl-edge interests. The many expected insights to bedelivered—the household’s flexibility potential, ‘ener-gy behaviour’, economy vs. new relations as motiva-tor for entering a partnership with DE, usability of theGWR equipment, etc.—resulted in an interview guidefor loop 1 that covered ten themes and a large numberof exercises to fulfil the in total seven deliveriesagreed upon. From a social science point of view, thisconferred a breadth to the interview guide that resultedin shallowness in the findings. Moreover, althoughDistribution’s main interest was in flexibility for heatpumps and electric vehicles, there was relatively littlefocus on these technologies in the field manual design.Relevant knowledge of the extent of householders’heat pump and electric vehicle flexibility wouldrequire thorough studies of heat comfort and mo-bility practices. Instead, the field manual focused a

great deal on communications and relations, atti-tudes, values and motivations for flexible electricityconsumption in general—which is not very surpris-ing, since DE formulated the project outcome to beknowledge on how to “attain a meaningful dialoguewith the customers about energy flexibility” (Ulk,unpublished working paper, 2011). Although ameaningful dialogue may be relevant to makingcustomers accept DE’s control of their heat pumpsand of charging of electric vehicles, the focus ofthe user study did not fit Distribution’s perspectivevery well. The final report on the user study re-vealed many interesting findings on participants’engagement in moving laundry and dishwashingfrom day to night and using the intelligent powernodes to survey and control the standby and powerconsumption of appliances such as TVs and com-puters. However, when considering Distribution’sexplicit disinterest in ordinary household applianceswhose relatively small electricity consumption wasconsidered irrelevant for any substantial load shed-ding, the relatively strong emphasis on these find-ings seems rather odd.2

2 The eFlex project started out as collaboration between DE Salesand Distribution, but the project was later completely separatedfromDESales due to the growing regulatory focus on reducing theadvantage of combining trade and distribution within the samecompany. Sales’ interests in developing new services and productsmay have influenced early design options in relation to the GWRsystem and the framing of the project—it should “be a commercialstepping stone in the attempt to handle the business models of thefuture” (Alexandra Instituttet 2010, p. 3).

Fig. 2 Several factors have influence on flexibility (image from antropologerne.com)

Energy Efficiency

Learning from eFlex

Despite these viewpoints, DE found the user studyuseful. On the technical side, for example, theyfound significant flexibility potential in the heatpumps, and the tests provided ideas on how toimprove the set-up—e.g. the optimisation algorithmneeded to be developed further, and the heat pumpportfolio should be controlled more intelligently. Onthe behavioural side, the household visits showedboth a relatively large willingness to be controlledand also fairly little comfort loss. In general, theuser study touched upon many of the questions thatDE wanted answered in the seven deliveries, such assegmentation (although without the desired ability forup-scaling), identification of motivations for flexibilityand development of communication principles. Thestudy had also confirmed two hypotheses: first, thathome automation can promote flexibility, and second,that “electricity is life” and people are not economical-ly rational in any simple sense—price is not theonly motivation (antropologerne.com 2012, p. 8).Actually, the project questions the traditional fram-ing of motivation which maintains that economicrationality (like savings) is opposed to moral con-siderations (doing the right thing). In this case,participants consider the price to be a ‘moral eco-nomic signal’ about what to do for the commongood.

Most of the findings in the user study confirminsights from existing literature about visualisationand feedback about electricity consumption (e.g.Darby 2006, 2010; Hargreaves et al. 2010), householdenergy consumption (e.g. Gram-Hanssen 2010, 2011)and family and ICT in everyday life studies (e.g.Røpke et al. 2010). However, since the study was aconsultancy job, the main aim was not to engage withthis literature but to make insights available for prac-tical use. The project team found it provided usefulinsights despite not generating all the results they hadhoped for. Much more time than anticipated was usedto dea l wi th t echn ica l p rob lems wi th theequipment—server break downs, malfunctioningpower nodes, etc. However, antropologerne.com ar-gued that the process itself was a result in its ownright. The involvement of households thus had aninfluence internally in DE—what the eFlex projectleader called ‘an unexpected side effect, but probablyone of the most important ones’. Due to their former

monopoly status, electricity companies tend to viewhouseholds as ‘loads’ rather than people (as the CEOof Danish Energy Association said at the eFlex con-ference), so the project rehearsed a more market-oriented relation to customers. The study was also atest-bed for new intra-organisational cooperation (alsoinvolving the IT department) and enforced a new typeof cooperation between the IT company producing theGWR equipment and DE. This is itself a relevantexercise for facing a possible smart-grid future, sincethe ‘old’ energy sector and the ‘new’ IT sector differconsiderably with regard to their relation to their prod-uct and customers.

For the engineers in DE, it was useful to considerhow the diversity of households’ comfort practicesmade the technical calculations difficult. One reasonfor these productive insights and the learning achievedamong all stakeholders in the project is related to theco-creation method that antropologerne.com appliedthrough user and analysis workshops with DE andthe methodology that emphasised photo and videorecordings during household visits. During the project,the varying backgrounds of the participants in theanalysis team—consisting of researchers, consultantsand engineers working in industry—meant that theactors had very different conceptions of what consti-tutes knowledge and interesting research questions.The more or less unstructured video recordings ofparts of the interviews, e.g. would normally not bethe chosen method in a sociological research project.However, the video recordings were not only meant asmaterial for analysis; they were also an important partof the co-construction process—a ‘translation’tool—and were an essential part of the delivery work-shops with DE, where selected video clips were oftenscreened.

Aligning users in networked system innovation

In addition to various kinds of learning, the eFlexproject’s benefits also related to branding and politicalimpact. Carrying out such an ambitious project con-tributes to DE’s positioning as a strategic systembuilder in current smart grid development. At theeFlex conference, the minister of Climate, Energyand Buildings could not emphasise enough what agreat success the eFlex project had been—and thathe expects to integrate the results of the eFlex projectin the planned Danish smart grid road map. The

Energy Efficiency

project can be said to develop a role for householdsthat fits the smart grid system demands. It has ex-plored and constructed a user image that can serve asa valid argument in the political battle to implement asmart grid—a system that integrates users who co-manage their electricity supply in a way that satisfieslarge centralised energy system actors’ demands.According to this image, the users themselves are inter-ested in demand management of their heat pumps, forexample, and this message is strategically communicat-ed through e.g. organisation of the eFlex conferencewith a press release that announced, ‘Just move myelectricity consumption’. This construction wassupported by the nursing of the participants: For exam-ple, as an expression of gratitude for inviting the anthro-pologists into their home, the participants received littlegifts from DE and many felt ‘pampered’ and importantfrom all the attention given them—e.g. when they wereinvited to events such as the Futures Night and addressedby the then CEO for DE, Anders Eldrup. As many of theparticipants were lead users, they would probably bepositive from the outset towards taking on the expectedrole, and if they were not, they would have to be ratherresistant to not ‘buy into the image’ through the process.

Based on these observations, we suggest that user-oriented innovation in relation to networked systemscalls for the supplementary concept of the aligned user.The eFlex project can be seen as a mediation junction(Schot and de la Bruheze 2003) where the projected DEusers and the designers of the GWR equipment meetboth some real users and the represented usersconstructed by antropologerne.com, and one outcomeof the mediation and negotiation process becomes thealigned user—the user that can serve as an argument inthe system-building process.3 In the design process oflarge networked provision systems, it is not only rele-vant to include users in order to develop a market for aspecific product or service; since users serve as impor-tant co-managers in the functioning of the system, it isalso important to construct an image of a user who needsthe system that the system builders aim to create.

Relating the discussion on the aligned user to the theo-retical discussions on lead users referred to in the intro-duction, we suggest that involvement of leadusers—like many of the eFlex pilots happened tobe—may serve a tactical purpose. Although the findingsmay not be generalisable and may not anticipate marketneeds, they may be useful as arguments.

The road to a low carbon future?

The question then remains whether the system that isbeing created, and the user roles related to it, actuallypromotes a low carbon future. Obviously, the thoughtbehind creating a smart grid is that it will enable theintegration of large amounts of renewable energysources and thus contribute to a low carbon future,but the development may also involve counter-trends.

Energy savings and counter-trends

The main intention of the eFlex project was to developan interest in peak shaving. The GWR system wasseen as a means to increase consumers’ interest inelectricity and to encourage learning about the patternsof electricity use—learning that was expected to leadto increased interest in load shedding. This aim wasachieved successfully, and in addition, considerableelectricity savings were realised. Some reduced elec-tricity consumption from heat pumps was ‘replaced’by other forms of energy (firewood), but this issue wasnot considered relevant to the project. As in othersmart grid projects, it was generally considered impor-tant that the results be achieved with little loss ofconsumer comfort, and it was not meant to challengehousehold expectations regarding indoor temperature,the ‘natural 21–22 °C’ (Shove 2003).

In the longer term, home automation may becomepart of a general trend towards the development of thesmart home, where ‘boring’ demand management ismade more attractive by bundling it with other ser-vices, which is described as “funwashing” (Nyborgand Røpke 2011). At the eFlex conference, it wasevident that many actors consider the smart grid pathas a way to develop new business and growth oppor-tunities for the ICT sector and that they emphasisecomfort and convenience for the ‘users of the system’.Such new services may add to electricity consumptionrather than savings. Moreover, the GWR system isdesigned to change certain practices and reduce

3 Just for clarification, the idea of the aligned user as a tacticalmove in relation to a wider strategic process of system buildingdiffers from Akrich’s (1995) discussion on alignment of userpositions. She focuses on the challenges designers face in thedesign process where different methods are applied to developand promote user representations in relation to a particularartefact and where a successful design depends on the reconcil-iation of the results of these methods.

Energy Efficiency

consumption, but as Geels and Smit (2000) writeconcerning the generative capacity of ICTs, perhapsthese technologies may also create new, unforeseen(energy consuming) practices.

Other systems could be constructed

The eFlex project is a strategic move toward activelynaturalising a specific future for energy provision—thesmart grid path—in the midst of a messy transition withcompetition between many possible pathways.Although the ‘hydrogen economy’, for instance, maybe an unlikely future alternative, other socio-technicalconfigurations for a sustainable energy system couldpossibly challenge the presented ideas about technolo-gies, ownership structures and practices (Walker andCass 2007). It may not be self-evident that a centralisedstructure is the only way forward: localised alternativesmay be preferable as suggested by research communi-ties concerned with renewable energy self-sufficiencyand the off-the-grid movement. Perhaps, a more activehousehold role as owners and managers of their energysupply in local communities might engage them farmore in participating in transforming the energy systemtowards sustainability (Späth and Rohracher 2010). Afew participants in eFlex, in any case, demonstratedtheir interest in actively engaging in independent devel-opment of new solutions.

There may also be reason to discuss the currentinstitutional and infrastructural lock-in related to atransport and heating system based on electrical vehi-cles and heat pumps, although the two technologies dohave wide-ranging potential for ‘greening the grid’.Attention should at least be paid to co-evolution be-tween systems and practices, such as increasing mo-bility or higher comfort expectations (Strengers 2008).One electrical vehicle owner in the eFlex study, afather of four, explained how he and his older childrenloved to drive the electric car because it was so easy,comfortable and silent. In addition, because of itsconvenient small size, it had replaced many of thefamily’s bicycle tours, e.g. for grocery shopping; orthe youngsters would use it when going to soccertraining, which the father would never allow with thegasoline car. The low operating costs of running anelectric car were not mentioned, but probably contrib-ute to the change of practices. Another study hasshown increasing mobility in relation to introductionof electrical vehicles, despite the simultaneous

introduction of measures aimed at shifting mobilitybehaviour, such as promotion of car-sharing or publictransport (see, e.g. Hoogma and Schot 2001).Likewise, studies of heat comfort practices in connec-tion to the introduction of heat pumps in Denmarkshow that heat pumps create new norms for higherindoor temperature in winter (Christensen et al. 2011).

Concluding remarks

The eFlex project has been one of the first smart griddemonstration projects in Denmark where the ‘consum-er side’ has played an essential part, and DE is planningto integrate the results in several others of its smart-grid-related projects. The project has been a first step intrying to open up the ‘black box’ of households, whichseems to be a core concern for many smart grid actors inDenmark. In the ‘old’ electricity system, these weremerely ‘loads’ that were predictable and ‘calculable’,but they have become a new ‘unruly factor’.

In the current struggle to develop a more sustainableenergy system, however, we also call attention to the‘aligned user’ and the political nature of this type of userstudies. They construct and naturalise certain futuresthat fit the agenda of the strategic system builders—forinstance, just by verbalising them as ‘the way to go’.Communicating that certain users are willing to acceptthe solutions can be an important part of the marketing,including the political marketing, of a new and unfamil-iar product. In addition to studies such as eFlex, we alsorecommend more studies of households and their role insustainable transitions, which challenge to a higher de-gree current institutions, systems and practices.

Acknowledgements The research for this paper was partlyfunded by the Danish strategic research program on SustainableEnergy and Environment (DSF 2104-09-0085). We are gratefulto antropologerne.com for letting the first author of this papertake part in their project, and we thank Eva Heiskanen and twoanonymous reviewers for very useful comments.

References

Akrich, M. (1995). User representations: Practices, methods andsociology. In A. Rip, T. J. Misa, & J. Schot (Eds.), Managingtechnology in society: the approach of constructive technologyassessment (pp. 167–184). London: Pinter Publishers.

Alexandra Instituttet (2010). eFlex B2C rapport. Aarhus:Alexandra Instituttet.

Energy Efficiency

antropologerne.com (2012), eFlex :: Fleksibelt elforbrug. En kvalitativundersøgelse af, hvordan deltagerne i projekt eFlex opleverhverdagen med energifleksibilitet, DONG energy andantropologerne.com.

Berker, T., Hartmann, M., Punie, Y., & Ward, K. (Eds.). (2006).Domestication of media and technology. London: OpenUniversity Press.

Christensen, T. H., Gram-Hanssen, K., Petersen, P., Larsen, T.F., Gudbjerg, E., Rasmussen, L. S., & Munther, P. (2011).“Air-to-air heat pumps: A wolf in sheep's clothing?”, en-ergy efficiency first (pp. 1963–1974). Stockholm: TheFoundations of a Low Carbon Society, ECEEE.

Darby, S. (2006). The effectiveness of feedback on energy con-sumption. A review for DEFRA of the literature onmetering, billing and direct displays, EnvironmentalChange Institute, University of Oxford.

Darby, S. (2010). Smart metering: What potential for house-holder engagement? Building Research and Information,38(5), 442–457.

Ea Energy Analyses (2010). Intelligent energy systems. A whitepaper with Danish perspectives. Copenhagen: Ea EnergyAnalyses.

Energinet.dk, & Dansk Energi. (2010). Smart Grid i Danmark.Fredericia: Energinet.dk.

Geels, F. W., & Smit, W. A. (2000). Failed technology futures:Pitfalls and lessons from a historical survey. Futures, 32,867–885.

Gram-Hanssen, K. (2010). Residential heat comfort practices:Understanding users. Building Research and Information,38(2), 175–186.

Gram-Hanssen, K. (2011). Understanding change and continuity inresidential energy consumption. Journal of Consumer Culture,11(1), 61–78.

Hargreaves, T., Nye, M., & Burgess, J. (2010). Making energyvisible: A qualitative field study of how householdersinteract with feedback from smart energy monitors. EnergyPolicy, 38(10), 6111–6119.

Heiskanen, E., Kasanen, P. & Timonen, P. (2005). “Consumerparticipation in sustainable technology development”, In-ternational Journal of Consumer Studies, 29(2), 98–107.

Heiskanen, E., Hyysalo, S., Kotro, T. & Repo, P. (2010).Constructing innovative users and user-inclusive innova-tion communities, Technology Analysis and Strategic Man-agement, 22(4), 495–511.

Hoogma, R. & Schot, J. (2001). How innovative are users? Acritique of learning-by-doing and -using. In R. Coombs, K.Green, A. Richards, & V. Walsh (Eds.), Technology and the

market: Demand, users and innovation (pp. 216–233).Cheltenham: Edward Elgar.

Hughes, T. P. (1983). Networks of power. Electrification inWestern society, 1880–1930. Baltimore: The John HopkinsUniversity Press.

Konkurrence- og Forbrugerstyrelsen. (2011). Detailmarkedetfor elektricitet. Konkurrence- og forbrugeranalyse.København: Konkurrence- og Forbrugerstyrelsen.

Nyborg, S. & Røpke, I. (2011). “Energy impacts of the smarthome - conflicting visions”, ECEEE 2011 Summer Study:Energy efficiency first: The foundation of a low-carbonsociety ECEEE, , 6–11 June 2011.

Raw, G., & Ross, D. (2011). Energy demand research project:Final analysis. St. Albans: AECOM.

Røpke, I., Christensen, T. H., & Jensen, J. O. (2010). Informationand communication technologies—A new round of house-hold electrification. Energy Policy, 38(4), 1764–1773.

Schot, J., & de la Bruheze, A. A. (2003). “The mediated designof products, consumption, and consumers in the twentiethcentury” in How Users Matter. In N. Oudshoorn & T. Pinch(Eds.), The co-construction of users and technology (pp.229–245). Cambridge: The MIT Press.

Shove, E. (2003). Converging conventions of comfort, cleanlinessand convenience. Journal of Consumer Policy, 26(4), 395–418.

Shove, E., & Chappells, H. (2001). Ordinary Consumption andextraordinary relationships: Utilities and their users. In J.Gronow & A. Warde (Eds.), Ordinary consumption: Studiesin consumption and markets (pp. 45–59). London: Routledge.

Späth, P., & Rohracher, H. (2010). ‘Energy regions’: The trans-formative power of regional discourses on socio-technicalfutures. Research Policy, 39(4), 449–458.

Strengers, Y. (2008). Comfort expectations: The impact ofdemand-management strategies in Australia. Building Re-search and Information, 36(4), 381–391.

Suchman, L. (2013). Consuming anthropology. In A. Barry &G. Born (Eds.), Interdisciplinarity: Reconfigurations of thesocial and natural sciences. London: Routledge.

Torntoft Jensen, L. (2011). EFLEX - en udforskning af, hvordander skabes og deles viden i et tværfagligt udviklingsprojekt,Digital Design og Kommunikation. København: ITUniversitetet.

Von Hippel, E. (1988). The Sources of Innovation. Oxford:Oxford University Press.

Walker, G., & Cass, N. (2007). Carbon reduction, ‘the public’and renewable energy: engaging with socio-technical con-figurations. Area, 39(4), 458–469.

Energy Efficiency