Version classiqueVersion mobile
OpenEdition Books


Madeleine Akrich
Yannick Barthe
Fabian Muniesa
et al.

Material disruptions in electricity systems: can wind power fit in the existing electricity system?

Peter Karnøe

Texte intégral


  • 1 I have developed this argument in small steps starting from an initial idea through many drafts, P (...)

1Policies designed to combat climate change and secure energy independence have facilitated the integration of renewable energy sources into power systems.1 However, adding intermittent energy sources like wind to an existing configuration of electricity generation disrupts its smooth functioning, and has spillover effects on the institutional arrangements for pricing electricity. I propose that the reason for this combination of material disruption and disruption of the institutional arrangement for valuation lies in the materially-grounded logics related to generating and balancing electrical loads. I have empirically studied the Danish configuration of electricity generation, which has been transformed over the years to include more and more wind power. As a percentage of total energy, wind power increased from 2 % in 1990 to 20 % in 2006, and is projected to be at 40-50 % by 2030 according to national energy plans.

2Electricity is a real time charge of electrons flowing at a constant frequency. The term real time refers to the fact that since there is no form of perfect storage for electricity, it must be consumed as it is produced to preserve the frequency or load balance. Such real time interdependency of production and consumption is what separates electricity from other technologies and goods like automobiles driven by either gasoline or electricity. For electricity the infrastructure cannot in the same way be separated from the consumption.

  • 2 The term material grounding of valuation emerged from conversations with Fabian Muniesa on the top (...)

3I offer the concept of material grounding of valuation2 to analytically highlight the ways in which the material configuration of electricity generation directly shapes cost structures and economic pricing of the associated energy technologies in current institutional arrangements related to market pricing. Recent sociological and institutional approaches to understanding the economy have provided important insights on how social networks, culture, institutional rules and regulations affect markets and economic actors. However, these new approaches under-specify the role played by material arrangements in making calculations, in framing economic thinking and action of actors, and in price-making mechanisms.

Sociomateriality in technologies and markets

4Coase (1988) may be used as a stepping stone towards an understanding of markets and prices, by emphasizing that markets are social institutions (and physical structures) that facilitate exchange and price making, with the notable characteristic that externalities are left out of market prices. But Coase did not as Weber see prices and the institutions producing them as outcomes of struggles of interests (Swedberg, 2003), took for granted the existence of agents capable of articulating ‘issues of concern’ and negotiate, and presupposed that property rights are distributed to make efficient solutions (Callon, 1998: 265). Such conceptions contributed analytically to the work of deleting the ongoing, but varying linkages between politics, society and economics, almost like in mainstream economics (Dobbin, 2004).

5The expression ‘material grounding of valuation’ adds nuances to the concepts of ‘framing/overflowing’ (Callon, 1998) that is used to denote the socio-technical operations involved in making an object economic and ready to transaction or fit for economic calculative and (e)valuative operations. Material grounding of valuation builds upon Dewey’s theory of valuation that stated that the ‘realm of human values’ and rational ‘world of facts-based’ value cannot be separated (Dewey, 1939: 64; Stark, 2009). Rather, to assign value (quality) to any object is an operation of appraisal – and this is captured in the double meaning of appraisal which is ‘to prize’ an object in the sense ‘I like/dislike the object’, and to price in the more strict economic sense ‘this is the price of object x’. To render an object economic is to assign it qualities that allow it to be legitimately priced and consumed. These qualities involve the intertwined process of ‘liking/ disliking’ and of putting an economic price on an object: “Prices are done, but they can do things too. Prices can be expressed in a quantitative fashion (‘105.30 USD’, ‘Euribor plus 20 basic points’) or in a rather qualitative one (‘cheap’)” (Muniesa, 2007: 378). Prices are neither naturally occurring phenomena coming out from natural markets, nor only artifacts resulting from being embedded in social elements. Prices are material entities and their existence is tied to – or grounded in – concrete arrangements and associations, and prices are but one particular quality that involved actors use as the basis for their judgment of the value of an object.

6Three examples from the sociology of pricing illustrate the relevance of the material grounding of valuation. In the 1990s the new electronic Paris Bourse witnessed a controversy about the appropriateness of closing price (Muniesa, 2007). The stock exchange is one particular form of price making and some traders did not find the closing prices in the stock exchange fair. Three different pricing configurations were considered – the weighted mean, the last trade in continuous trading system, and the closure call auction – that produced different aggregates and designated the value of a stock differently. The study shows that for the prices to ‘work’ they must rest upon a legitimate representation of buying-selling, because the prices serve as “input for market agents to consider the sign (price) as a legitimate input to further calculations and considerations, based, for instance, on general financial valuation codes” (Muniesa, 2007: 387). Accordingly, the ‘quality’ of prices, assigning if they ‘work well’, is a symbol that expresses a convention on the value of the stock. The process of ‘fixing’ of the inappropriate closing prices so they close at the ‘right’ level was an important role of market engineers and architects involved in re-making the pricing arrangement of the Paris Bourse.

7What is the role of pricing formulas in producing prices? Financial market studies show how material theoretical equipment matters (MacKenzie, 2003). The BlackScholes-Merton equation for option pricing was not a close empirical description of the world of option pricing. Rather, it was a new pricing model that proposed a new economic reality for traders. An economic actor equipped with the BlackScholes-Merton option-pricing model will think and act differently from an actor without the model. Further, when they embark upon arbitrage, traders discuss and assign value to the theory, i.e. they must convince themselves that the theory on which the arbitrage rests is at least plausible enough to be the basis of practical action. They will often also want to or need to convince others. Thus, the success of the Black-Scholes-Merton option-pricing model was not caused by an a priori fit between ‘theory’ and ‘reality’, but by contrast that fit improved rapidly after the theory’s practical use because it performed a new reality (see also Garud & Beunza, 2007).

8A third example is the problem of externalities and ‘fair’ prices on fossil fuels in existing price-making arrangements in the markets of production and consumption of fossil fuels. Coase’s ‘problem of social cost’ and the tragedy of the commons are often referred to, but there is more interesting pricing formulas within the economic profession itself. For example, economic theory established a certain convention that accepted that oil and coal was so abundant around the 1920s that the ‘theory of economics’ was developed without the assumption that it is a scarce non-renewable resource (Mitchell, 2009). Proponents of the view that oil and coal were scarce resources, like Alfred Marshall, claimed they should be priced differently, but this view by did not have lasting effects on theory building nor on political regulations. Another important economic convention linked to externalities is the invention of the term Gross National Product.

9Nobel Laureate Kuznets invented the term ‘GDP’ in the 1930s to be able to account for th ‘wholeness’ of economic activity, and since then the massive statistical and calculative apparatus of GDP has defined how we calculate the ‘wholeness’. Despite critiques for not including externalities it has been the dominant calculative and evaluative metric for measuring if the economy was developing in ‘good’ or ‘bad’ ways. However, even the economic convention of GDP may eventually change. For example the Sarkozy Commission led by Nobel Laureate Stiglitz concluded (once again) that GDP is a poor metric for economic activity, because it among other things misses ‘pollution’, ‘health’, etc. (Stiglitz et al., 2009). It is against this backdrop of historical development of energy consumption and ‘economic theory’ that we must understand the conclusions from the acclaimed Stern Report in 2006. The major conclusion is that the exclusion of the externalities from energy cost was the biggest market failure of the last century. The failure to internalize the ‘pollution cost’ in the fossil fuel prices means that these energy sources have been priced too low, and if the prices had been higher it is almost certain that the economic law of supply and demand would lead to lower consumption and more investment in more energy efficient technologies. Further, instead of renewable energy being too expensive it is the other way round: those fossil fuels are priced too low and implicitly subsidized because the tax payers or future generations must bear the environmental cost of using fossil fuels. Why is such unfair prices not fixed? Well, according to the Externe study by the European Commission in 2003 it was for political reasons impossible to put more taxes on fossil fuels, and therefore renewable energy technologies must be subsidized (and suffer from the identity-mark of ‘being subsidy dependent’ while the fossil fuel technologies are competing on fair grounds).

  • 3 Associations constitute a temporal interlocking of sociomaterial elements that may have a high deg (...)

10Following a ‘pragmatics of valuation’ perspective, prices are not right or wrong, subjective or objective, false or true. Prices are outcomes of and tied to specific material arrangements, and constructing a truth of ‘natural prices’ that make certain prices appear as objective is a costly and difficult activity of objectification and naturalization (Muniesa, 2007: 390). A critical insight from these examples (and from the pragmatics of valuation) is that the ‘economy’ can be seen as the establishment of more or less stable, competing valuation networks (Muniesa, 2007: 381). These valuation networks are not innocent or neutral, but make prices in particular ways, and consequently the term ‘material grounding’ may draw upon the classic STS theme that ‘artifacts have politics’ (Winner, 1980). The socio-technical arrangement for making prices does shape and reshape social action (how actors act and think about the economy), rather than simply making existing forms of social action more efficient. This argument is similar to the way Meyer & Rowan (1977) and Oliver (1990) showed how entrepreneurs inscribed their goals into the institutional regulations, which provide durability or carrying capacity (Scott, 2003) to a network of valuation. This critical insight is connected to Fligstein’s idea that architectures of markets are political, i.e. they carry the inscriptions of interests stemming from the actors who, with visible hands, worked to influence the architecture of a network of valuation in a distant past.3

11In this sense, the economic agencies and valued objects that constitute ‘the economy’ are not pre-existing, but outcomes of socio-technical accomplishments that generate different epistemic cultures of calculation and valuation – or networks of valuation. This phenomenon can be seen in the current economic climate, as the black boxes of calculative practices that led to the financial crisis are opened. These practical capacities of valuation are not constant, but are accomplishments that can be studied sociologically as they evolve and stabilize with the potential to transform.

12With the expression ‘material grounding of valuation’ I want analytically to highlight the particular material arrangement that are implicated in the operation of articulating qualities for appraising and for economic pricing in order to facilitate judgmental actions at the discursive level. Following Dewey (1939), humans are constantly involved in valuations of things, that sometimes lead to the formulating of new desires and sometimes ‘just’ following established habits. But shifts may happen, and Dewey stresses that when valuations that exist at a given time are found to lack the support they have previously been supposed to have, they exist in a context that is highly adverse to their continued maintenance. Thus, this brings to attention the potential competition between valuation networks which I will analyze in the study of how wind power disrupts the smooth operation of the Edison/Tesla configuration of electricity generation, and the associated valuations.

13The analytical focus on the material is not to undermine the role of human endeavors and agencies, but to point to the crucial linkage between ‘smoke and fire’. There is rarely smoke without fire, and following a STS approach it is rarely possible to claim anything about the qualities of an object without references can be made to the material arrangement of instruments, concepts and conventions that was involved in making such claims possible. (At the same time the path from fire to desires is not direct, as it involves reflexive operations that only humans are capable of doing.)

The Edison/Tesla configuration of framing electrons and economics

  • 4 Hughes (1983) and Granovetter and McGuire (1998) published foundational studies detailing how the (...)

14This framing perspective assumes that, at the outset, neither a product’s properties nor the capacities to produce, use, value or regulate it exist alone. Electricity is particularly interesting in this regard because the material nature of electricity makes it impossible to be produced, circulated and consumed in large quantities without machinery and cables. The Edison/Tesla model of electricity production has become the dominant model in all industrialized countries, with centralized thermal power plants generating electricity, a high voltage transmission grid, and a low-voltage distribution grid to consumers. The Edison/Tesla model has become so successful and naturalized that its origins have been black-boxed, disappearing into a rational ‘best way’ of thinking, which is problematic when configurations of electricity generation face transformation challenges.4 Current attempts to transform the existing Edison/Tesla configuration may be informed by the renewed understanding of the organizing processes that paved the way for this dominance.

The operational logic of load balancing in the Edison/Tesla configuration

  • 5 This is standard terminology today (Hoogwijk et al., 2007; Østergaard, 2007), but is increasingly (...)

15The packaging of electrons in the Edison/Tesla configuration resulted from a complex set of intertwined framing operations (and inscriptions) that ended with a sociomaterial configuration designed to send large quantities of ‘load’ from centralized power plants to distributed end-users. Today it is established knowledge that there is a critical interdependency between power generators with respect to their ability to adjust to shifts in demanded load, as well as being able to serve the highest load at any point in time. However, organizational and operational routines are not pre-existing. They involve learning as they are performed. The specific practices and terminology (routines) of operating the new technology emerged from experience with problems. Critical management terms like ‘load factor’, ‘base-load’ or ‘peak-load’ were invented while experimenting with ways to handle affordances in the techno-economic optimization of the electricity system (Hughes, 1983).5 Also the highly critical function of ‘load dispatcher’ was invented, and gave rise to a new actor with the sole role to preserve the load balance (frequency) in the new interconnected system of electricity generation and consumption.

  • 6 Source: Kai Heussen, personal communication, March 30, 2010.

16The special operational logic of load generation and load balancing became one of ‘turning controllable power plants up and down’ in a way that ‘generation follows demand’.6 In that sense the real time interdependency of production and consumption was solved in the Edison/Tesla configuration by making power plants load-active and by making consumption load-passive. In this configuration centralized power plants served base loads and were complemented with reserve capacity that respond quickly in order to meet ‘minute-to-minute-load-fluctuation’, equipment break-downs as well as to meet expected peak-load demands. It was thus not an inevitable way to handle the load balancing problem. It rather was a particular configuration based upon technologies with certain properties (i.e. power sources that can be controlled in order to generate electricity at relatively constant rates).

The evolution and possible demise of the Danish Edison/Tesla configuration

17In 1907, the Edison/Tesla configuration had already been proposed in Denmark by an electrical engineer (van der Vleuten, 1998). However, it took about 50 years before resistance from the competing networks of valuation in the Danish patchwork based on local generation and distribution (which included private and municipal ownership and their special interests) could be overcome by the formation of new interests and coalitions (Akrich, Callon & Latour, 2002). In a step-by-step process, proponents made their claims legitimate and interesting to opponents, with techno-economic calculations and demonstrations playing an important role in these translations (van der Vleuten, 1998). Denmark ended up with an Edison/Tesla configuration with centralized power plants generating 96 % of electricity in 1970 and 98 % in 1980. However, Denmark developed a specific version of the Edison/Tesla configuration with two characteristics. First, since the 1950s centralized power plants were transformed into or designed as co-generation plants that used the cooling water to supply heating to homes via a grid of water tubes. That increased the energy efficiency as well as provided a solution to cool down the cooling water before sending it back to the river/sea. Secondly, from the 1980s onwards a larger number of small decentral combined heat and power plants were built and corresponded to 17 % of electricity production in 2000 (Karnøe & Buchhorn, 2008). It is in this specific material configuration where electricity and heat production technically is tied together that wind power enters.

18From 1976 to 1990, about 1500 small 30-200kw wind turbines were installed in windy areas and typically connected to transformer stations at the ‘thin’ end of the grid. Technically, wind turbines could only be connected to the grid because they had a fixed rotor speed design using an asynchronous generator, which allowed the kinetic energy in the wind to be transformed to rotations in the generator equaling the grid frequency (i.e., the rotational speed of other generators). However, this technical design does not allow wind turbines to contribute to “dynamic reactive power and voltage control” and “they may disconnect at disturbances with excessive voltage and frequency sags in the power grid” (Akhmatov & Knudsen, 2007: 1229). Technological developments allowed turbine size to increase to about 200kw and improved aerodynamics (kwh production/sq. meter rotor) as well as the efficiency of material structures (weight/kwh) and machinery (Karnøe, 1991). By 1990, 2 % of the Danish electricity production was generated by wind power.

19The general valuation of wind power by the official electrical utilities organizations was based upon the existing Edison/Tesla configuration. Below is a summary of elements of the valuation represented in various reports from the 1980s:

  • Wind turbines only generate power when there is wind, and without a cheap and loss-efficient storage solution, it cannot work well.
  • Often, there is either too much or too little power generated to meet demand, and other energy technologies must be able to compensate in order to meet normal load in addition to peak load.
  • Within a given regional boundary, wind power creates overcapacity or pure over-production in relation to traditional electricity production – it has no net capacity effect. Indeed, the capacity factor of wind energy will decrease with increased penetration in the existing configuration of electricity generation.
  • Wind power will squeeze out conventional power production which will be required to run below techno-economic optimum levels. Even low-cost production may be squeezed out.
  • Wind power is costly and uneconomical.

20Based on a few installations of small wind turbines, electric utilities and the system operators framed load balancing as an economic problem; hosting and integrating a bunch of small wind turbines would increase the total cost of production of the existing configuration. Wind power was an overflow to this configuration. During the 1990s, climate policies and a new generation of larger and more efficient 600kw wind turbines made wind power a highly attractive investment, and installations increased to a 12 % share of Danish electricity production in 2000, and 20 % by 2008 (KarnØe & Buchhorn, 2008). Most of the capacity came from on-shore installations, but plans for off-shore installations were being developed quickly. The increased material presence of this ‘large-scale’ penetration of wind power in the electrical grid aggravated the overflow in terms of critical material imbalances in load balancing. The imbalances happened as wind turbine owners could not control wind speeds, and other heat and power plants had to be used in winter since it was technically impossible to separate heat and power production. As a consequence, the Danish Energy Agency asked the involved stakeholders – among them the wind turbine owners, electric utilities, TSOs, the Organization for Renewable Energy, and experts – to analyze the problem and come up with solutions to electricity overruns, ‘el-overløb’ in Danish (Energistyrelsen, 2001).

21The report invented the term ‘critical power overrun’: this is something which happens when “the electrical load cannot be consumed within a given area – and other power plants have difficulty in reducing further, therefore export is the only option to avoid critical system failure” (Energistyrelsen, 2001). Following the proposal form the report the political majority was able to make an important shift in regulation in 2004, when the decentral combined heat and power plants now shifted role in the ‘dispatch hierarchy of power generators’. In order to prevent ‘power overruns’ these plants were now required to stop production to give space for wind power generated loads in the system.

22Another result from these overflows was that, in 2005, IDA – the Association of Danish Engineers – announced 2006 to be an ‘Energy Year’. IDA’s ambition has for long been to influence the development of challenging jobs for engineers in research and industry. Since 1980, the new engineering and research jobs related to the energy and environmental sectors had been the strongest growth. IDA initiated a Creative Innovative Process orchestrated by professor in energy planning H. Lund and colleagues. 1600 engineers in Denmark participated in 40 workshops in 2005-2006. The IDA results were based upon the EnergyPlan model, that due to the funding from IDA and input from many workshops became updated (model-wise and data-wise). The IDA/EnergyPlan model works with a step-wise deviation away from the logic of the existing Danish Edison/ Tesla system. As such the IDA/EnergyPlan model is completely opposite of the Hoogwijk study (Hoogwijk et al., 2007), that was based upon preserving the existing Edison/Tesla system. The calculative machinery in the IDA/EnergyPlan study introduces, in a step-wise manner, new material technologies with new combinations in the supply and demand side, and the demand side is made more and more load-responsive (compared to a load-passive demand in the Edison/Tesla configuration). The report demonstrated that it was physically and economically possible to transform the Danish energy system to accommodate 50 % wind power.

23Even if it is not a direct consequence of the first IDA/EnergyPlan Report (Lund & Mathiesen 2009: 524) it was a conceptual and valuation revolution when the Danish energy system operator, TSO, published a report (the report) in 2009 projecting that it is technically and economically feasible that Denmark can increase the share of wind power from currently 20 % to 50 % share of wind power generated electricity by 2030. As a consequence of the political priorities the Danish TSO allowed a group of engineers and economists to respond to the political ambitions to increase wind power from 3000Mw to 6000Mw or more. Traditionally only strong proponents of wind power have mobilized such growth curves as new ‘utopian’ realities for wind power in the energy configuration of the system.

Figure 1: The TSO view on new material elements in the future Danish electricity/energy system

Figure 1: The TSO view on new material elements in the future Danish electricity/energy system

24Figure 1 shows the new material elements in the TSO model. One important step in this ‘material fit’ is to make ‘consumption’ active in power load balancing (this is what is known as a ‘Smart Grid’). By framing consumption as an active part of load balancing the volatility of wind power can be even more absorbed, and increases the value of wind power. But this is only one element. The material arrangements for heat production are expected to happen via individual and collective installations, when oil and gas-fired boilers are replaced with individual and collective heat pumps with water tanks. ‘Turning’ the water temperature up/ down becomes a new ‘semi-storage’ medium for electricity. Further, electrical battery driven cars can function as semi-storage of wind power as they can be charged or de-charged according to the use-profile of the user (see results of study by Pillia & Heussen, 2009). Furthermore, by increasing grid capacity and connections, load areas are distributed and this increases the value of wind power. Also, the increased geographical distribution of wind power farms is expected to even out grid loads and exploit the ‘west-to-east flow of dominant wind directions’.

A revolution in the Danish configuration of electricity generation and wind power valuation?

  • 7 Neither the IDA/EnergyPlan model nor the TSO report demonstrates a precise set of instruments and (...)

25This is a remarkable shift for wind power: from being unwanted and disregarded 1970s, to possibly becoming the primary energy source in the 2030s. Indeed, a revolution is under way in the Danish energy system. The TSO report is important because it is the first report from a central actor in the established Edison/Tesla system that demonstrates a new possible world. It may become a device that evokes futures and at the same time articulates uncertainties in ways that afford action rather than close courses of action. In this way it may contribute to solidify the transformation of the Edison/Tesla configuration. Indeed, the TSO report is already seen as so authoritative that a few months after its publication the Danish Association of Wind Turbines Owners and the Danish Association of Wind Turbine Industry made a joint publication (FDM & DWP, 2009). Here they strongly advocated for ‘a right way to expand wind power in the configuration of the energy system’, and the authoritative reference in the report was the 2009 TSO report. In this sense, the series of energy plans have gradually shifted the associations in the energy sector in the sense that the discourse, the interests, and the allies in the coalition that is backing ‘more renewable energy’ have become stronger as they can refer to these demonstrations.7

26Here there is new discursive smoke because references to a new fire can be made. From a pragmatics of valuation perspective, these reports offer competing networks of valuation of wind power in different material groundings, and produce conceptual material that allow understandings of the potential value of wind power to be articulated and contested in discussions. This is consistent with a performativity perspective as the IDA/EnergyPlan and TSO represent not only an alternative network of valuation, but also serve to stimulate action that can create a fit between ‘theory’ and ‘reality’, in the sense that the reports lay out a foundation to be enacted by utilities, regulators and other actors. The reports may be seen as ‘appraisal-valuation models in vitro’ of how to prize and price wind power, but performativity-wise they may help to bring into existence an ‘in vivo’ reality, that fits the theory. There is, however, a heated debate in electricity journals about the problems and value of wind power (DeCarolis & Keith, 2005; Bathurst et al., 2002; Meibom et al., 2009; Strbac & Kirschen, 2000; and Sapio, 2004). But how does this possible large-scale penetration of wind power influence the price-making mechanisms in the electricity markets?

The Nord Pool Spot Market: a price-making convention in an Edison/Tesla configuration

27The Nordic Power Exchange is a complex socio-technical market arrangement. The Nord Pool Spot Market was created in Norway in 1993; joined by Sweden in 1996, Finland in 1998, and Denmark in 1999. An extremely complex information technology infrastructure is required for bidding and price-making in the market. Demand and supply bids are collected and handled within each market type (spot market, and hourly balancing market) and a special mathematical price-making formula transforms them into an aggregate system price for Nordic countries. There are detailed rules describing appropriate behavior as well as sanctions for rule violations (i.e., attempting to affect prices).

  • 8 Hoogwijk et al. (2007) explains the bidding and price-making procedure in great detail. They elabo (...)

28The bidding procedure, which has a 24-hour lead time, is central to price-making in the spot market. The bidding procedure is a convention, that has been used for decades in order to give TSO time to appropriately plan and allocate expected production to different power plants in a given supply area (Hoogwijk et al., 2007).8 In the Nordic region, this technique began as a basis for price-making in Norway in the 1970s when hydropower was traded between different regions based on the

  • 9 Source: Ole Jess Olsen (professor in Nordic energy markets), personal communication, 2008.

29amount of water in their reservoirs9. In this way, the 24-hour day ahead bidding procedure in the Nord Pool Spot Market is not some neutral mechanism for price-making, but is a materially-grounded inscription of the existing planning cycle for production time allocation in the Edison/Tesla configuration.

  • 10 Textbook concepts may very well be formulated based on what happens within established technologie (...)

30Central to the bidding procedure is the ‘merit order curve’, a supply curve that reflects the cost and capacities of power generators ranked from least expensive to most expensive. Marginal cost differences are calculated based on technology and fuel cost. Normally, power sources are ranked in the following order on the curve: wind, nuclear, coal combined heat and power (CHP), condensing plants, gas turbines (Hoogwijk et al., 2007; EWEA, 2010). Hydropower is normally not included since it is a strategic resource that depends upon rain and water levels in reservoirs. The merit order curve is a convention for price-making based upon the textbook concept of increased marginal cost in economics.10

31Although the day-ahead bidding procedure is taken for granted as a convention for price-making, it is not neutral as most economic theories and ideologies would predict. On the contrary, it allows for appropriate planning and response, given the material properties and technical designs of the old Edison/Tesla configuration. Material load balancing problems of integration are expected to increase even further as wind power is installed in very large (>200Mw) off-shore wind farms. Although it will contribute more load, more wind power will also lead to more load fluctuations and result in a need for further compensation. This will put further pressure on load-following capabilities and will potentially create negative pricing consequences for wind farm owners.

Material disruptions of pricing architectures in the Nord Pool Spot Market

32The increased penetration of wind power has resulted in other unusual effects on pricing in the Nord Pool Spot Market, with prices going to zero and even negative. However, these problems are minor compared to wind power effects that may disrupt the merit order curve in spot markets. The term ‘merit order effect’ has been coined by economists who study the consequences of large-scale wind power penetration on spot and wholesale market prices. In April, 2010, the European Wind Energy Association published a document written by expert economists from the consulting firm Pöyry describing the new phenomena. The literature review presents widely divergent predictions based on specific material properties of the different configurations of energy systems, including “the assumed wind penetration level, the power generation mix and the marginal cost of the replaced conventional technology” (EWEA, 2010: 16). Thus, the material properties of existing configurations of energy systems and the specific price-making arrangements in liberalized power markets directly influence whether these effects occur, and to what extent.

33All of the reviewed studies show that the merit order effect occurs when a large amount of wind power (a zero marginal cost energy technology) shifts the supply curve to the right, lowering the price when demand is constant (EWEA, 2010). As prices fall, the profitability of the rest of the technology mix is negatively affected. These effects happen with a low to moderate wind power penetration rate. What happens with a very large penetration of wind power in the configuration of electricity generation?

34A simulation study by Bode and Groscurth (2008) addressed this question before most even thought about it. The authors concluded that the central pricemaking mechanism, the convention of the merit order curve, breaks down when wind power penetration reaches 40-50 % (Bode & Groscurth, 2008). Renewable energy technologies (e.g. solar, wind, hydro) have zero marginal costs since plant investments include the fuel, so to speak. Since adding zero marginal cost technologies causes prices to decrease, if renewable energy technologies can deliver all the energy demanded at a point in time, the resulting price is close to zero. This is framed as a failure in the market design since the resulting price is insufficient to cover the investment costs of wind power and the other power generators that are part of the flexible energy system. Consequently, it seems that the large-scale material penetration of wind power in the configuration of electricity generation requires the invention of new price models. Indeed, a central question becomes which actors, economic theories and criteria will be involved in the creation of efficient markets?


35The idea of a material grounding of valuation suggests a new research agenda. Empirical analysis of the evolution of the Edison/Tesla configuration shows that valuation of wind power is co-dependent upon the material arrangements in which it is integrated. The sociomaterial packaging of electrons can be delivered by different configurations of electricity generation and load balancing. This study shows that material grounding can impose a positive or a negative framing of the value of wind power.

36The operating logic of the existing electricity system, the Edison/Tesla logic, became a logic of ‘turning-controllable-power-plants-up-and-down’ with ‘loadpassive-consumption,’ as more/less fuel or more/fewer power plants can be activated. Adding volatile, uncontrollable energy technologies like wind power to the Edison/Tesla configuration disrupts the logic of operation. Wind power only generates electricity when there is wind, and wind power generation cannot ask for more or less wind. And that means that the compensation and regulation needed to accomplish load balancing must be supplied by other ‘controllable and predictable’ generators (or by reducing demanded load). Given the sources of storage this physical load balancing sets up a material condition for the interdependency of the ‘load components’ in the energy system, and it is possible to talk about a ‘material fit’ to a dominant logic of load generation and balancing. What is at stake is whether the added marginal energy technology is convergent to or diverging from the dominant logic of generating and balancing electrical loads. This puts constraints on the conventional part of the electricity system in terms of compensatory regulation capabilities – and it spills over in valuation and in affecting the design of the market architecture. Each logic represents a particular framing of how fossil fuel or wind power is transformed into some electromagnetic stuff that ‘obeys’ the law of the electrical load balancing, which avoids black-outs (preserves the volts and frequencies) and at any time delivers the electricity to consumers whenever a device is plugged in. But there are signs that a new logic of operation is possible; a logic which we label ‘volatility/load active consumption’ logic.

37The expression ‘material grounding of valuation’ makes it analytically possible to show that, besides from institutional rules and regulations, the very material configuration of the electricity system is central to the valuation of wind power and other energy technologies, and that the institutionalized rules, beliefs, and regulations of the electricity sector rests upon translations of specific material configuration of equipment, as well as associated calculative conventions. The effects of material grounding on valuation is clear from three sources here: 1) the load balancing’s overflows in the existing Edison/Tesla configuration, and 2) the overflows in the price-making mechanisms in the Nord Pool pricing arrangement, and 3) in the IDA and TSO reports that depict a new value for wind power in a transformed configuration of material technologies and of active consumption in load balancing. Seen from this perspective, and putting load balancing in the center, wind power does not fit smoothly into the existing physical configurations of electricity generation and their associated price/valuation mechanisms.

38By contrast, the conflicting logics of load generation and load balancing turn the basic form of the Edison/Tesla configuration into a ‘hostile territory’, where it is costly to make wind power ‘fit’.

39Once it has been shown that the Edison/Tesla configuration presents a ‘hostile territory’ for volatile intermittent energy sources, the potential for transformation raises two important questions. How should techno-economic problems associated with the current transformation towards more hospitable territories be framed? Should integration problems and costs be attributed to new technologies, or to the existing materially hostile Edison/Tesla configuration?

40There is a critical new research agenda in the invention of new price models for electricity markets, and a central question is which actors, economic theories and criteria will be involved in the creation of price-making arrangements that can be seen legitimately as creating efficient markets. With increasing policies to combat the ‘carbon crisis’ the time of fossil fuel based electricity production is running out, but the speed and directions of transformations are not clear. These depend upon constantly addressing the beliefs, identities, competencies, regulations and conventions of calculation associated with the existing Edison/ Tesla configuration that over time have become silenced, taken for granted and black-boxed or naturalized. It is theoretically and politically important to increase the analytical sensitivity to the material grounding of valuation in order to account for the existing or propose new material arrangements represented in the calculative technologies which are used to make and stabilize (or to break) the claims, discourses, and calculative conventions inscribed in market arrangements.



Akhmatov, V. & H. Knudsen (2007), “Large Penetration of Wind and Dispersed Generation into Danish Power Grid”, Electric Power Systems Research, 77, p.1228-1238.

Akrich, M., M. Callon & B. Latour (2002), “The Key to Success in Innovation, Part I: The Art of Interessement”, International Journal of Innovation, 6 (2), p. 187-206.

Bathurst, G., J.Weatherill & G. Strbac (2002), “Trading Wind Generation in Short Term Energy Markets”, IEEE Transactions on Power Systems, 17 (3), p. 782-789.

Beunza, D. & R. Garud (2007), “Calculators, Lemmings or Frame-makers? The Intermediary Role of Securities Analysts”, in M. Callon, Y. Millo & F. Muniesa (eds), Market Devices, Oxford: Blackwell, p. 13-39.

Bode, S. & H.-M. Groscurth (2008), “Incentives to Invest in Electricity Production from Renewable Energy under Different Support Schemes”, working paper, Arrhenius Institute for Energy and Climate Policy, Hamburg, Germany.

Callon, M. (1998), “Introduction: The Embeddedness of Economic Markets in Economics”, in M. Callon (ed.), The Laws of the Markets, Oxford: Blackwell,p. 1-57.

Coase, R. (1988), The Firm, the Market and the Law, Chicago: University of Chicago Press.

DeCarolis, J. & D. Keith (2005), “The Cost of Wind’s Variability: Is There a Threshold?”, The Electricity Journal, 18 (1), p. 69-77.

Dewey, J. (1939), Theory of Valuation, International Encyclopedia of Unified Science, Vol. II (4), Chicago: The University of Chicago Press.

Dobbin, F. (2004), “The Sociological View of the Economy”, in F. Dobbin (ed.), The New Economic Sociology: A Reader, Princeton, NJ: Princeton University Press, p. 1-48.

European Wind Energy Agency (2010), Wind Energy and Electricity Prices: Exploring the ‘Merit Order Effect’, Brussels: EWEA.

European Wind Energy Agency (2006), Wind Power Technology: Operational, Commercial Developments, Wind Projects, Grid Distribution, Brussels: EWEA.

Energistyrelsen (2001), Eloverløbsrapport: Rapport fra arbejdsgruppen om kraftvarme- og VE-elektricitet. Copenhagen: Energistyrelsen.

Energy Crossroads (2010), Changing The Game: Background Material, Copenhagen: Energy Crossroads.

FDM & DWP (2009), Den rigtige vindkraftudbygning: Anbefaling fra Danmarks Vindmølleforening og Vindmølleindustrien, Aarhus: Danmarks

Vindmølleforening, and Frederiksberg: Vindmølleindustrien, [available online at and].

Garud, R., A. Kumaraswamy & P. Karnøe (2010), “Path Dependence or Path Creation?”, Journal of Management Studies, 47 (4), p. 760-774.

Garud, R., J. Gehman & P. Karnøe (forthcoming), “Categorical Performances: Nuclear Power in the Energy Field”, in W. D. Sine & R. David (eds), Research in the Sociology of Work, Vol. 21: Institutions and Entrepreneurship, Greenwich, CT: JAI Press.

Granovetter, M. & P. McGuire (1998), “The Making of an Industry: Electricity in the United States”, in M. Callon (ed.), The Laws of the Markets, Oxford: Blackwell, p. 147-173.

Hoogwijk, M., D. van Vuuren, B. de Vries & W. Turkenburg (2007), “Exploring the Impact on Cost and Electricity Production of High Penetration Levels of Intermittent Electricity in OECD Europe and the USA, Results for Wind Energy”, Energy, 32 (8), p. 1381-1402.

Hughes, T. (1983), Networks of Power: Electrification In Western Society 1880-1930, Baltimore: John Hopkins University Press.

Karnøe, P. & A. Buchhorn (2008), “Denmark: Path Creation Dynamics and Winds of Change”, in W.M. Lafferty & A. Ruud (eds), Promoting Sustainable Electricity in Europe, EE Press, p. 73-101.

Karnøe, P. (1991), Dansk vindmølleindustri: En overraskende international succes, Frederiksberg: Samfundslitteratur.

Lund, H. & B. V. Mathiesen (2009), “Energy System Analysis of 100 % Renewable Energy Systems: The case of Denmark in Years 2030 and 2050”, Energy, 34, p. 524-531.

MacKenzie, D. (2003), “An Equation and its Worlds: Bricolage, Exemplars, Disunity and Performativity in Financial Economics”, Social Studies of Science, 33 (6), p. 831-868.

Meibom, P., R. Barth, H. Brand & C. Weber (2009), “Operational Costs Induced by Fluctuating Wind Power Production in Germany and Scandinavia”, IET Renewable Energy Generation, 3 (1), p. 75-83.

Meyer, J. W. & B. Rowan (1977), “Institutionalized Organizations: Formal Structure as Myth and Ceremony”, American Journal of Sociology, 83, p. 340-363.

Mitchell, T. (2008), “Re-thinking Economy”, Geoforum, 39, p. 1116-1121.

Mitchell, T. (2009), “Carbon Democracy”, Economy and Society, 38 (3), p. 39-432.

Muniesa, F. (2007), “Market Technologies and the Pragmatics of Prices”, Economy and Society, 36 (3), p. 377-395.

Oliver, C. (1991), “Strategic Responses to Institutional Processes”, Academy of Management Review, 16, p. 145-179.

Østergaard, P. (2007), Electricity Systems for Energy Planners, 2nd edition, Aalborg: Department of Energy Planning, Aalborg University.

Pillai, J. & K. Heussen (2009), “Bornholm as a Model for 100 % Renewable Energy Scenarios in Denmark”, working paper, Department of Energy Technology, Aalborg University.

Sapio, S. (2004), “Market Design, Bidding Rules, and Long Memory in Electricity Prices”, working paper, Laboratory of Economics and Management, Sant’Anna School of Advanced Studies, Pisa, Italy.

Scott, W.R. (2003), “Institutional Carriers: Reviewing Modes of Transporting Ideas Over Time and Space and Considering their Consequences”, Industrial and Corporate Change, 12, p. 879-894.

Strbac, G. & D. S. Kirschen (2000), “Who Should Pay for Reserve?”, The Electricity Journal, 13 (8), p. 32-37.

Stark, D. (2009), The Sense of Dissonance: Accounts of Worth in Economic Life, Princeton, NJ: Princeton University Press.

Stiglitz, J. E., A. Sen, J.-P. Fitoussi (2009), “Report by the Commission on the Measurement of Economic Performance and Social Progress”, Paris: CMEPSP [available online at].

Swedberg, R. (2003), Principles of Economic Sociology, Princeton, NJ: Princeton University Press.

Togeby, M., H. H. Lindboe & T. Pedersen (2007), “Steps for Improved Congestion Management and Cost Allocation for Transit”, TemaNord, 537, [available online at].

Van der Vleuten, E. (1998), “Electrifying Denmark: A Symmetrical History of Central and Decentral Electricity Supply until 1970”, PhD thesis, Aarhus University, History of Science Department.

Yacubovitch, V., P. McGuire & M. Granovetter (2005), “Electrical Charges: The Social Construction of the Rate Systems”, Theory and Society, 34, p. 579-612.

Winner, L. (1980), “Do Artifacts Have Politics?”, Daedalus, 109 (1), p. 121-36.


1 I have developed this argument in small steps starting from an initial idea through many drafts, PowerPoint presentations and conversations. I am thankful for comments from the CEESA group, Martin Kornberger, and participants from the Energizing Markets Conference (Copenhagen, October 2008), the EGOS Energizing Energy Markets subtheme (Barcelona, July 2009), the Constitution of Future Markets seminar (Technical University, Berlin, October 2009), and the Gain seminar series (Saïd Business School, February 2010). I am also grateful for conversations with Kai Heussen and David Stark in January 2010 and comments from Michel Callon in March 2010. This research has been funded partly by the Danish Research Council for Social Sciences, and partly by the Strategic Research Council.

2 The term material grounding of valuation emerged from conversations with Fabian Muniesa on the topic of this article.

3 Associations constitute a temporal interlocking of sociomaterial elements that may have a high degree of irreversibility, and it requires skilled work to re-constitute and modify such interlockings (Callon, 1991). However, the fact that it is costly or difficult to transform does not mean that it is impossible, indeed from a path creation perspective it is a matter of competing agencies and how alternative agencies can gain strength vis-à-vis agencies of entrenched actors (Callon, 2007; Garud, Kumaraswamy & Karnøe, 2010).

4 Hughes (1983) and Granovetter and McGuire (1998) published foundational studies detailing how the Edison/Tesla system became dominant in the US electricity industry from the 1880s to the 1920s. Edison and Tesla did not act in a vacuum; as reported by Granovetter and McGuire (1998), there were some 1,500 different electrical arc and incandescent configurations operating in homes and factories between 1878 and 1885. By the end of the 1920s the Edison/ Tesla configuration was completely dominant in the US and well on its way in Denmark and many other places (ven der Vleuten, 1998; Hughes, 1983). The forces that Granovetter and McGuire refer to are mainly sociopolitical, such as developed shared understandings, social connections, organizational conditions. Despite its richness this approach underspecifies how the material presence of Edison’s and Tesla’s patents and technical equipment afforded, but not determined, understandings and sharedness, as well as building the collective agency and its distributed work. Mitchell’s study (2008) emphasizes the role of calculative devices in making the socio-technical connections that made the US electricity industry. He demonstrates that “The economic calculus was not a market of rational agents outside the system. It was an apparatus of calculation that brought elements from other projects – generator cost, transmission losses, data on gas lightning, population densities – into Edison’s workshops at Menlo Park, New Jersey, where it could be combined with other instruments of measurement and calculation in the construction of the electrical system.” (Mitchell, 2008: 1118).

5 This is standard terminology today (Hoogwijk et al., 2007; Østergaard, 2007), but is increasingly criticized for belonging to an old world of technology and operational principles (Energy Crossroads, 2010).

6 Source: Kai Heussen, personal communication, March 30, 2010.

7 Neither the IDA/EnergyPlan model nor the TSO report demonstrates a precise set of instruments and policies to realize this new technical vision, but it is nevertheless a new platform. Source: Frede Hvelplund (Danish alternative energy researcher), personal communication.

8 Hoogwijk et al. (2007) explains the bidding and price-making procedure in great detail. They elaborate on the 24-hour planning cycle and bidding process. They also discuss the concept of merit order strategy, where power plants are operated in order of variable cost. When this strategy is applied, capital intensive plants with low operating costs, such as nuclear, wind and solar are operated for as many hours as possible to cover base load requirements. Intermediate plants using coal, oil or natural gas are designed to handle shoulder load fluctuations. Plants with low capital costs but high variable costs such as gas turbines and diesel engines are used in peak load situations where demand is extremely high since they have quick-start capabilities.

9 Source: Ole Jess Olsen (professor in Nordic energy markets), personal communication, 2008.

10 Textbook concepts may very well be formulated based on what happens within established technologies (e.g., the electrical and cement industry), thereby naturalizing their application, like the linear correlation between economic growth and energy consumption.

Table des illustrations

Titre Figure 1: The TSO view on new material elements in the future Danish electricity/energy system
Fichier image/, 65k

© Presses des Mines, 2010

Conditions d’utilisation :