the power of grid parity: a discursive approach

Technological Forecasting & Social Change xxx (2014) xxx–xxx

TFS-17909; No of Pages 12

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

The power of grid parity: A discursive approach

L.A. Hurtado Munoz⁎, J.C.C.M. Huijben, B. Verhees, G.P.J. VerbongEindhoven University of Technology, School of Innovation Sciences, IPO Building, P.O. Box 513, 5600 MB Eindhoven, The Netherlands

a r t i c l e i n f o

⁎ Corresponding author at: Eindhoven University ofment of Electrical Engineering, Corona Building, P.OEindhoven, The Netherlands. Tel.: +31 648750733.

E-mail address: [email protected] (L.A.H. M

0040-1625/$ – see front matter © 2013 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.techfore.2013.12.012

Please cite this article as: L.A.H. Munoz, et(2014), http://dx.doi.org/10.1016/j.techfor

a b s t r a c t

Article history:Received 29 January 2013Received in revised form 26 June 2013Accepted 13 December 2013Available online xxxx

In the debate around solar photovoltaic (PV), the concept of ‘grid parity’ has emerged as thedominant benchmark for competitiveness, while some even argue that it will determine thepoint in time after which the PV industry will boom. But more recently, others have called intoquestion the usefulness of the grid parity concept. Yet despite its pervasive use and increasingcontestation, the grid parity concept has not been systematically interrogated to date. Thispaper makes two contributions towards that: first, to show how the grid parity conceptemerged and how it is calculated and second, to explore the role of the grid parity debate inthe solar PV field. The first contribution takes the form of a literature study of grid paritystudies. To arrive at a meaningful estimation of the grid parity point, assumptions made ineach step of the calculations have to be articulated and carefully evaluated. Nevertheless, thisis almost never done: the grid parity studies, presentations and reports we reviewed invariablyused the simplest representation available. We argue that their authors chose a simplifiedmodel for strategic reasons, e.g. to obtain (material and/or non-material) resources. Thisassessment leads to our second contribution: a discourse analysis of the grid parity debate. Wedistinguished ten key storylines and six discourse coalitions, comprised of actors who share aspecific set of these storylines. Analyzing these storylines and coalitions, we show that whilethese actors share a common goal of PV up-scaling, they can have drastically different ideasabout strategies to achieve this goal. Opening the black box of grid parity thus reveals tensionsabout preferred strategies in an otherwise seemingly homogeneous PV discourse.

© 2013 Elsevier Inc. All rights reserved.

Keywords:PhotovoltaicGrid parityCalculation methodologyDiscourse analysisUp-scaling strategies

1. Introduction: the role of grid parity

Over the last four decades, a growing number of peoplehave been advocating a change in the current energy systembecause of its unsustainable nature. At the current growthrate of energy consumption and of technology innovation,the world's capacity of meeting the needs of the future isinsufficient. Geopolitical instability, fossil fuel prices, CO2

emissions, global warming and energy sources depletion arejust some of the issues these actors use to argue that atransition to a more sustainable energy system is needed.Alongside efficiency increases, renewable energy sources like

Technology, Depart-. Box 513, 5600 MB

unoz).

ll rights reserved.

al., The power of grid pe.2013.12.012

PV, wind, hydro and biomass should replace fossil fuel firedpower stations. PV technologies are often argued to be anoption with very high potential because the sun provides10,000 times the planet's energy demand every day [1].But despite this huge hypothetical potential and decades ofpolicy support for the development of the PV technologies, theyare not yet part of the existing energy regime [2]. This is oftenexplained by the high production costs being the main barrierfor market diffusion, but many additional barriers have alsobeen identified: market barriers (e.g. market failures anddistortions), institutional barriers (e.g. lack of legal frameworksand institutions), technical barriers (e.g. lack of codes, standardsand skilled people) and social barriers (e.g. user acceptance andawareness) [3,4].

Nevertheless, in renewable energy literature, the debatearound the development of PV has focused on its high up-frontinvestment costs, the issue ofwhen PVwill become competitive,

arity: A discursive approach, Technol. Forecast. Soc. Change

http://dx.doi.org/10.1016/j.techfore.2013.12.012

mailto:[email protected]


http://www.sciencedirect.com/science/journal/00401625


2 L.A.H. Munoz et al. / Technological Forecasting & Social Change xxx (2014) xxx–xxx

and what should be done to facilitate this (e.g. what technologyto support or what kinds of support mechanisms to use). In thisdebate, the concept of “grid parity” has emerged as a keycompetitiveness indicator. Broadly speaking, grid parity refers tointersection of the price of the electricity generated by a PVsystem and the price of conventional electricity production.

This concept has become widely accepted as a “milestone”,and sometimes even “holy grail” of the PV industry, eventhough it relates exclusively to economic and financial barriers.It is used by many actors for obtaining resources, managingexpectations and steering activities within the PV sector, e.g. inthe form of maps such as in Fig. 1, which shows a seeminglyinevitable ‘front’ of grid parity washing over Europe from thesouth where high irradiation levels lead to relatively low PVgeneration costs. As PV generation costs decrease over timecountries with lower levels of irradiation also reach grid parity.

The use of the concept has been increasing over time,starting around 2007 (Fig. 2). This figure shows the numberof web search of the term “grid parity” in Google. The earliestsource relating to grid parity was written in 2005. An articlefor the magazine “Frontiers, the BP magazine of technology andinnovation” related grid parity with making solar PV compet-itive [6]. More recently however, the concept of grid parityhas come to be contested by a growing number of actors whocriticize its relevance.

Yet despite its pervasive use by the solar PV community,and the increasing contestation, the concept of grid parity hasnot been critically interrogated to date. This paper makes twocontributions to this general aim by answering the followingtwo questions:

1. How did the grid parity concept emerge and how is thegrid parity point calculated?

2. What is the role of the grid parity debate in the PV field?

Fig. 1. The grid parity front moving over Europe. From top to bottom:

Please cite this article as: L.A.H. Munoz, et al., The power of grid p(2014), http://dx.doi.org/10.1016/j.techfore.2013.12.012

In Section 2, we articulate a conceptual framework andmethodology for doing so: a discourse analysis. Section 3addresses the first question by describing the roots of gridparity and critically evaluating its dominant calculationmethod.Section 4 then addresses the second question by examining thevarious storylines, actors and positions in the grid parity debate.Section 5 provides conclusions and discussion.

2. Conceptual framework and methodology

In everyday life, people engage in debates around topics inspecific social contexts. Such debates can be driven by personalbeliefs and interests that are usually shared by some andopposed by others. In such debates, actors express their beliefs,using language strategically in an attempt to convince others [7].Among actors within the same social subsystem — such as thesolar PV field— the language used tends to be similar: actors usea similar way of talking about and expressing their views of theworld. Such a ‘way of talking’ is referred to as a ‘discourse’ —e.g. a PV discourse. Discourses give meaning to the physicalworld, and this meaning depends on the social and culturalcontext in which the discourse is embedded. Discourse thuscontributes to constructing reality, in the sense that things andconcepts only gain meaning through discourse [8]. Discourse,then, is not only about talking but also about understanding: itis, broadly, a “particularway of talking about and understandingthe world (or aspects of the world)” [8 pp. 1].

The concept of discourse has come to be used by variousfields, including linguistics, psychology, sociology and polit-ical sciences. As such, different approaches to the analysis ofdiscourse have been articulated. These approaches differ,among other things, on their levels of analysis, and theirempirical focus (see [9–13]). For our goal of understandingthe role of the grid parity debate in solar PV discourse, the

grid parity in 2008, 2010, 2015, 2020 and 2030 respectively [5].



1 Throughout the first stage of the research project (i.e. the grid paritycalculation methodology), it was found that grid parity calculations weremade before 2005 as well. However, these started as more simple break-even analyses and did not make use of the “grid parity” label as will beshown in the next section. Therefore, these were not included in thediscourse analysis. Nonetheless, three calculation studies [22–24] wereincluded that both used the term grid parity and showed a specific stancetowards the subject.

Fig. 2. Number of worldwide web searches on Google of the term “grid parity” (normalized at 100; absolute data not available). Image generated by: GoogleTrends.

3L.A.H. Munoz et al. / Technological Forecasting & Social Change xxx (2014) xxx–xxx

work of Maarten Hajer is particularly interesting, because ituses a similar level of analysis and empirical case: he useddiscourse analysis to understand the role of the ‘acid rain’debate in the 1980s environmental discourse [7]. Hajer [14]defined discourse as an “ensemble of ideas, concepts andcategories through which meaning is given to social andphysical phenomena, and which is produced and reproducedthrough an identifiable set of practices” [14 pp.67]. In Hajer'sdiscourse analysis, a researcher examines (written or spoken)statements, which typically take the form of ‘storylines’. Astoryline is a “condensed statement summarizing complexnarratives, used by people as ‘short hand’ in discussions” [14 pp.69]. In any field, Hajer argued, there are several such storylinesthat fulfill a particularly important role [14 pp. 69]. Storylinesare the medium through which actors try to convince others oftheir positions, suggest certain practices, and criticize alterna-tives [14 pp. 71]. Around such storylines, actors can formdiscourse coalitions. In Hajer'swords, a discourse coalition is thus“the ensemble of a set of storylines, the actors that utter thesestorylines, and the practices through which these storylines getexpressed” (14 pp. 71). Analyzing concrete utterances by PVactors can then reveal not only their constituent storylines, butalso (the coalitions of) actors that rally around them [14].

So, in order to understand the role of the grid paritydebate in the PV field, we make a two-stage analysis:

• In the first stage, we address the first research question.Eleven grid parity calculation studies (see [15–25]) werefound based on a literature study of scientific journals (e.g.Solar Energy), energy conferences (e.g. PV Solar EnergyConference and Exhibition) and technical reports. They wereanalyzed according to themethodology used, the assumptionsmade, and the factors included in the calculationmethodologyof the grid parity point.We describe the emergence of the gridparity concept from experience or learning curves; review thevarious options for calculating it; and critically evaluate thecurrently dominant method.

• In the second stage, we make a discourse analysis based onHajer's methodology [14 pp. 73]. This stage consists of sixsteps:

1. Using the search strings “grid parity”, “PV grid parity” and“break-even point” on scientific, energy and technical tradejournals (e.g. Energy Policy, Solar Energy and TechnologicalManagement & Strategic Management), technical reports,web sites (e.g. Assolare, EPIA, IEA, EuPd and EIA) andscientific related magazines (e.g. Renewable Energy World,Future Photovoltaics, Spectrum IEEE and Climate spectator)we identified (59) documents (at least one article per actor)


written by (45) actors1 from a variety of public and privateorganizations actively concernedwith the grid parity debate

2. We did a general survey of these documents in order to“come up with a first reading” [14 pp. 73], examining thedocuments for their constituent storylines [14 pp. 73] andcoded these.

3. We then constructed a matrix of actors and storylines.Key storylines were identified (storylines that exhibit aclear stance and define the stance of the actor), andaround them coalitions were constructed. This meansthat actors that made use of that key storyline weregrouped together into the same coalition. Table 1 givesan example to illustrate this process. Storylines 2 and 3are defined as the “key storylines”. Thus, authors 1 and2 are grouped into coalition 1 and authors 3 and 4 intocoalition 2. This process is independent of storylines 1and 4. However, it does not mean they are not part ofthe analysis.

4. Based on this, we identified 6 groups (6 stances or keystorylines) consisting of actors and the storylines theyshared. We take these as proxies for discourse coalitions inthe PV field.

5. In order to further verify our group selection, wetriangulated [26] our identification of the various discoursecoalitions by sending out a survey to 45 stakeholders(including actors from each of the six groups identified, aswell as a PV industry associations and PV companies acrossEurope). The survey contained a summary of our initialfindings as well as a two-part questionnaire aimed at (1)validating our interpretations, and (2) adding any keystorylines our initial analysis may have missed (responserate was 22.2%).

6. Fifth, we fed back interview and survey results into ourcategorization and interpretation and drew conclusions.

The results of the first stage of our analysis are detailed inSection 3. This section then provides a context for the resultsof the second stage in our analysis, which is presented inSection 4.



Table 1Mapping of authors and storylines. The authors and the storylines they shareare referred to as the various ‘discourse coalitions’. Storylines 2 and 3 areidentified to be key storylines. As a result, authors 1 and 2 (gray areas) aredefined as coalition 1 and authors 3 and 4 (dark grey areas) as coalition 2.

Storyline 1 Storyline 2 Storyline 3 Storyline 4

Author 1 X X

Author 2 X X

Author 3 X X

Author 4 X X


3. Emergence and calculation of grid parity: acritical assessment

The grid parity concept can be traced back to experiencecurve theory. This idea (i.e. experience curve) emerged first inpsychology (1885) as a tool for memory studies and has beendescribed by Hermann Ebbinghaus [27]. However, it wasquickly translated to the study of economic and technicalperformance in industrial processes in the 1930s [28]. Here,experience curves illustrate the learning development processof an industry as experience is gained through repetition. Thesecurves are constructed using empirical data usually simplifiedto a learning ratio (LR) or progress ratio (PR) (LR = 1 − PR).The two main uses for experience curves are (1) as amonitoring tool and away to quantify the learning investments(i.e. investments required to bring down costs to a desiredlevel), and (2) as a forecast tool (e.g. in energy models fortechnology development through its cost evolution [29]). Anextensive literature on experience curves exists (e.g. [28–31]),and as such, many different mathematical models of experi-ence curves have been articulated. Table 2 summarizes severalof these.

Despite of the great variety in possiblemodels (see Table 2),we found that only the “log-linear model”, which also happensto be the simplestmodel in experience curve literature, is being

Table 2Overview of experience curve models [32].

Learning curve model

Log-linear models Log-linear modelDe Joung modelStanford-B modelS-curve modelPlateau modelKnecht log-linear modelYelle modelTwo-factor learning curveMulti-factor learning curve

Exponential models Knecht exponential modelTwo-parameter exponentialThree-parameter exponentialConstant time

Hyperbolic models Two-parameter hyperbolicThree-parameter hyperbolic

Differential model Differential modelSecond and higherorder models

Cherrington and Towill constant time modelHackett model with n = 1

Please cite this article as: L.A.H. Munoz, et al., The power o(2014), http://dx.doi.org/10.1016/j.techfore.2013.12.012

f grid p

used for grid parity calculations. Eq. (1) gives themathematicalrepresentation of the model.

y ¼ C1xX0

� �b

ð1Þ

with, C1 as the initial price or cost; x the cumulative production,sales or installed volume at a time t; X0 the initial cumulativeproduction, sales or installed volume; and b (−1 b b b 0) therate of innovation or learning parameter. The latter representsthe slope of the curve and is related to the progress ratio withthe following equation:

PR ¼ 2b ð2Þ

Nonetheless, other models make use of different param-eters, which could give a better estimation. For instance, thetwo-factor learning curve includes cumulative R&D expendi-tures as a second factor besides the cumulative capacity.Eq. (3) gives the mathematical representation of this model.

y ¼ C1xbKSn ð3Þ

Here, KS represents the R&D knowledge stock and n thelearning by searching index. The latter relates to the learningby searching progress ratio PRLBS in the same way as the rateof innovation relates to the progress ratio. Furthermore,exponential and hyperbolic models make use of exponentialand 1

�x factors, respectively, with the aim to extract more

information on the learning process and a more accurateestimation of the production rate. Another type of model isthe differential model, which tries to describe better how thelearning process occurs by the use of a differential multipli-cative relationship. Finally, the second and higher modelsinclude a certain form of periodicity after fitting the timeconstant model to the data [28–31].

In general, the PV industry's experience curve (log-linearmodel), when plotted on a logarithmic scale, see Fig. 3,appears as a straight line with a constant learning rate ofabout 20% for every doubling of the cumulative capacity [29].The extrapolation in time of the experience curve allows thecalculation of the point (i.e. cumulative production) whenthe technology will reach a certain price. Applied to PV,experience curves are used in combination with marketgrowth rates to estimate the year in which PV will cost thesame as its fossil fuel alternative (e.g.[16–18,21,30]): a

Fig. 3. Solar PV break-even study, using log-linear model [15].



Experience Curve[LR]

Selection of a technology for

comparison, Break even price [CT]

Break even price and EC

with the same units

LCOE

Calculate the break even

market size [XT]

Calculate the time to reach the

desired market size [GP]

PV Grid Parity

Learning investments

[LI]

Yes

No

Fig. 4. Grid parity calculation (schematic) [32].


break-even point. For example, an estimation made by theInternational Energy Agency (IEA) in 2000 stated that PVmodules would reach the break-even point around 2025,assuming an annual growth rate of 15% and a learning rate of20% [15].

By the mid-2000s, this break-even point for PV came to bereferred to as ‘grid parity’ (e.g. [22–25,33,34]). The earliestuse of the term grid parity we found is in a 2005 article in theoil and gas multinational BP's industry magazine Frontiers [6].Broadly speaking, the term typically refers to intersection ofthe price of the electricity generated by a PV system and theprice of conventional electricity production. Although exactdefinitions vary between the various grid parity studies(e.g. [35,36]), what these different definitions of grid paritycalculations have in common is that they are all comparisonsbetween the levelized cost of electricity2 (LCOE) of the PVtechnology and the LCOE of a competing technology or a gridelectricity price. The methodology for calculating the gridparity year is shown in Fig. 4. It starts with the extrapolationof the PV experience curve for a specific market. Next, thetarget price CT [$] (fossil fuel alternative price in Fig. 2) has tobe selected.3 Extrapolation of the experience curve to thetarget price determines the break-even market size XT [GW](‘break-even point’ in Fig. 3) and the learning investmentsLI [$GW] (shaded area under the EC in Fig. 3). With thisrequired market size now determined, the final step is to usethe technology's market growth rate to forecast when thismarket size will be reached (i.e. the point in time when gridparity will be reached).

Grid parity calculations have been made many times, andwith widely different outcomes. We have analyzed a selectionof 11 grid parity studies, and made the following observations:

• Despite the multiple different models available, only thelog-linear model is used. Therefore, unlike in other modelssuch as two-factor LC (TFLC) and multi-factor LC (MFLC),RD expenditures, input price changes, and scale effects arenot considered.

• The learning rate is assumed to be constant and typicallyaround 20%. However, it is not clear for how long this rate canbe sustained in the future. Market penetration is believed tobe oneof the reasons for the experience curve to flatten out. Itis argued that cost reductions differ in rate according to thedevelopment stage of the technology [8].

• A global value for market growth rate is typically assumed,despite such growth rates being country- and technology-dependent.

• The calculations typically either refer to crystalline silicon(c-Si) technology, or specify no technology, while c-Si is justone of many possible PV technologies, whichwould arguablyhave different experience curves.

• Although the experience curves should be constructed usingcost data, price data are typically used instead due to betteravailability.

2 The levelized cost of electricity (LCOE) is the price at which electricityshould be generated by a source in order to break even over its lifetime. Ittherefore includes initial investment, operation and maintenance, cost offuel, cost of capital, the energy produced and the system lifetime.

3 This target price corresponds to the LCOE of the competing technologyor the grid electricity price.


• Experience curves are typically constructed using price datafor PV modules, while the outcomes of the grid paritycalculations claim to be for PV systems (of whose cost themodule represents only around 50% — and moreover, thisshare likely decreases as module cost is reduced).

• While electricity prices (i.e. target cost) are time-, technology-and market-dependent, this is typically not included in thegrid parity calculations — nor are specific policies (e.g. taxes)and support mechanisms (e.g. subsidies and feed-in tariffs).

• While the PV system electricity price clearly depends on solarirradiation, and is therefore time- and location-dependent,this is typically not taken into account — nor are systemlifetime, system efficiency, degradation rate, rate of return oninvestment, or operation and maintenance (O&M) costs.

• Grid connection costs are not taken into account in grid paritystudies. However, as the PVmarket is growing investments inan updated, heavier electricity grid are needed. Also, at somepoints in time, feeding back electricity to the local grid mightnot be desirable andwill therefore be charged by the networkoperator.

These observations highlight that while grid paritycalculation seems straightforward and objective, it is decid-edly not. To arrive at a meaningful result, assumptions madein each step of the calculations have to be evaluated carefully.Sensitivity analysis is required, as well as optimistic andpessimistic scenarios, to come to a meaningful grid parityestimation (which, for this reason, is more likely to be a range




than a year). Nevertheless, we found that this is almost neverdone: the grid parity studies performed invariably use thesimplest model available, and typically neglect the issuesmentioned above. The grid parity estimations seem to servethe purpose of informing future PV competitiveness only,leaving out any unnecessary complexity. Moreover, whenpresented in reports (figures or tables) or during conferencepresentations its complexity is often even further dimin-ished. Fig. 1 is an example of this. It is being used to describehow PV systems will reach grid parity in Europe, throughthe forecasting of lower system costs, which will make PVgeneration progressively cheaper at lower solar irradiationlevels and competitive in northern Europe as well. However,it fails to include factors like the electricity price increase,the type of PV systems, subsidies, etc. This situation hasconsequences for the debate around grid parity, as we willshow in Section 4.

4. Role of the grid parity debate in the PV field: adiscourse analysis

We refer to the different way that actors in the PV fieldtalk about — and understand — solar PV technology broadlyas ‘PV discourse’. As we are interested in the role of gridparity in the PV discourse, we analyzed a selection of writtenutterances by PV field actors in which the term grid parity isused (step 1; 59 documents produced by 45 actors). Thissection contains the results of this analysis.

4.1. Storylines

Analysis of our selection of documents yielded tendifferent storylines which mobilize the concept of grid parity.Their main tenets are described here. For each of thestorylines, we have added a quote from one of the relatedliterature sources to indicate its essence.

[1] Grid parity as the key milestone of the PV industry.“Solar photovoltaic (PV) power is set to achieve theenvironmentalists' holy grail of grid parity — the samecost price as fossil fuels — across the European Unionby 2017” [37]. The actors that make use of thisstoryline (see [35,37–44]) in general argue that gridparity should be the main priority of the PV industry.By reducing production costs the industry gets closerto independence of incentives and subsidies: the pointof self-sufficiency. Additionally, after grid parity, therewill be an increase of the demand (a ‘boom’) since themain barrier for mass adoption is long payback times(high system costs, resulting in high electricity cost).Furthermore, producing electricity at the same level asthe grid conventional price means that PV technologyis able to compete with the traditional energy sources,making it a viable option for centralized electricitygeneration (‘PV power plants’) as well. “Achieve gridparity with the mainstream power generation tech-niques and the world changes” [40].

[2] Cost reductions are not just needed in PV modules.“Cost parity for solar power cannot be attained solelythrough improvements in solar cells. The cells them-selves account for less than 50% of system cost”


[36].The main argument of this storyline is that inorder to reach grid parity, cost reductions are neededin the system as a whole and not just in the modules.Module costs represent only half of the total systemprice: ‘Balance of system’ (BOS) components andinstallations account for the rest. Therefore, costreduction through further development of BOS com-ponents is crucial. This requires R&D support andinvestments in the development of new materials;significant cost reductions are only achievable throughthis route. Furthermore, it is also important that thelifetime of the entire elements matches, as this willreduce the operation and maintenance costs of thesystem, which in turn reduces electricity generationcosts (see [6,36,38–40,45–47]).

[3] The drivers of grid parity. “To encourage the use of PV,governmental agencies across the world are intro-ducing various incentive mechanisms. Rising electric-ity prices coupled with increasing environmentalconcerns amongst consumers have also contributed toan increase in PV installations” [42]. This storylineargues that the main drivers of grid parity are higherelectricity prices (i.e. political dimension), high solarirradiation (i.e. geographical dimension) and technicalimprovements. Grid parity is defined as the intersectionbetween grid electricity prices and PV electricity cost.Higher electricity prices result in this intersectionoccurring at a lower cumulative capacity, which willtranslate into an earlier grid parity point in time.Additionally, higher solar irradiation and technical im-provementswill increase the power output of the system,reducing the generation costs thereby also bringing thegrid parity point closer. This storyline connects PV toenvironmental concerns which are also part of thepolitical dimension of each country: if the externalitiesrelated to CO2 emissions and environmental problems areincluded in the electricity price scheme, the PV industrywould not need support. Thus, including externalitieswillbring closer grid parity (see [38,41,42,47–59]).

[4] Grid parity is almost/already here. “The 2010s arecharacterized by ongoing grid parity events throughoutmost regions in the world, reaching an addressablemarket of about 75 to 90 percent of total globalelectricity market” [51]. This is a storyline aboutcountries which very nearly will reach — or in someversions already have reached. For instance, it is arguedby EPIA that grid parity will be reached in Spain, Italyand Germany by 2015 and in most of Europe by 2020.This storyline distinguishes between retail grid parity(between 2012 and 2015), and wholesale grid parity(around 2030). Thus, grid parity is not a single point intime: it should be seen as something that slowly evolvesover a number of years across different markets andlocations. Nonetheless, among actors using this storyline,there is no general consensus on where grid parity isgoing to be reached first, although the majority point toItaly (see [23,24,35,37,43,45,46,48–53,57,60–70]).

[5] Subsidies, incentives and R&D support determine gridparity. “Unless governments continue and expand theirfinancial incentives and policy mandates for solarenergy, the recent high growth rate will not be



4 The definition at which the author makes reference is the following:“The common definition of grid parity compares the Levelized cost of PVgenerated electricity with the prevailing retail tariff simply because this ishow the customer would be charged if the electricity was purchased fromthe grid” [72 pp. 3].


sustainable” [71]. In this storyline, the PV industry iscurrently being driven by policy measurements thatresult in cost reductions in the PV system. Each countryhas different renewable and solar support mechanisms.Economic incentives, financing, subsidies, regulationsand other mechanisms have a direct effect on thedevelopment of the industry, and on its market growth.They are driving the cost reductions of the industry andcancel out market externalities, thereby determininghow fast grid parity is reached. Without them, most PVmarkets would not exist. In addition, the storyline warnsabout the effects of cuts in these mechanisms: uncer-tainties related to their future can endanger reaching gridparity. Crucially, the storyline argues that supporting thePV industry makes sense since oil and natural gas havebeen the main beneficiaries of energy incentives [49]. Inthis storyline, FiTs (feed-in tariffs) are commonly cited aseffective mechanisms, removing price uncertainties ofthe market. Even after grid parity has been reached, thismechanism should continue to help ensure that manu-facturers keep reducing prices, to avoid industry stagna-tion [56]. However, recently, Germany's FiTs cuts havedrawnmuch criticism, and the Spanish case is commonlycited as an example to show the potential problems thatthese actions can bring on the industry (i.e. slowergrowth rates). Some actors using this storyline thereforeclaim that zero interest financing of investment costscould be a better incentive mechanism than FiTs [47].This storyline also calls for further developments interms of support mechanisms, e.g. R&D support for newmaterials to cope with the increased demand [30], andpolicy incentives for addressing barriers like upfrontcosts [71] (see also [6,39,42,48,49,52,53,57,60,61,64,65]).

[6] Grid parity as one of the PV industry's milestones.“Policymakers, however, must recognize that droppingcosts in solar technology will not automatically resolveour energy problems. If policymakers wish to helpdistributed solar technologies across the chasm intocommercialization, political mandates to further encour-age their adoption would be necessary” [71]. In thisstoryline, grid parity is a useful benchmark, but for arobust indicator of competitiveness, a more complexformulation is needed [46]. It also argues that gridparity's impact on demand is overestimated — acommon example is the Hawaii solar thermal market:despite grid parity, solar heat systems are still notdominant, because of a chasm between ‘visionary’ and‘mainstream’ consumers in Hawaii, which grid parity didnot manage to bridge [71]. This storyline, too, arguesthat cost competitiveness is not the only barrier for themass adoption of a technology like PV, and highlightsthat barriers like high upfront costs, long rate ofreturn on investment, and unfamiliarity with thetechnology need to be addressed as well. Customersneed a compelling reason to buy and consumer behaviorchange is needed. A common metaphor used in thisstoryline is that of the green banana: despite the obviouslong-term benefits, people want the yellow banananow [50]. Additionally, it is argued that current marketgrowth is artificially high, and not global: the PVindustry is growing in a limited number of countries


due to their specific support mechanisms. For theindustry to become mainstream, the technologyshould be widely spread around the world (see also[50,52–55,57,58,60,61,70–72]).

[7] Shortcomings of the grid parity concept and theindustry. “The first shortcoming in this definition4 isthat it uses a flat tariff for grid electricity as the basis forcomparison. Wholesale electricity prices vary consider-ably throughout the day. A flat tariff, regulated in manyjurisdictions, obscures the time-dependent cost ofelectricity” [72]. This storyline highlights the shortcom-ings and missing points of the concept of grid parity.Unlike, storyline 6, this storyline focus on the conceptitself and what is missing to be a better benchmark,rather than on the other milestones needed to start theup-scaling process. First of all, the industry's market islimited by the demand's growth rate and by competitionwith other energy technologies. PV will still need tocompete with the other technologies after grid parity:this means that not all the new required capacity will besupplied by PV. Additionally, the grid parity concept hasseveral fundamental problems. First, grid parity calcula-tions fail to recognize grid connection costs (transmis-sion and distributions networks which interconnect thesystems), which represent 45% of the grid electricityprice [72]. These costs are important especially for PVtechnologies, due to the fact thatwithout energy storagetechnologies PV systems are not able to meet the totalenergy demand of any user. Second, the grid paritycalculation methodology uses a flat tariff (during a day)as a basis for comparison, which obscures the electricitycost's time-dependency. And finally, analyses are usual-lymade based on themodule production cost, but this isnot the price at which the modules nor the systems aresold to the end consumers (see also [59,71,77]).

[8] Problems after grid parity. “The question, though, iswhether First Solar or any other solar manufacturerwould be able to handle the flood of orders that wouldensue if they reached competitive cost. At that point, itcomes down to a matter of having enough of rawmaterials” [39]. This storyline argues that, while afterreaching grid parity demand for PV systems will indeedincrease, this might not be a good thing if the industry isnot prepared for it. First, an increase in demand willcreate a higher pressure on raw material supply, whichwill lead to a fluctuation in the system prices dependingon the suppliers' capacity (or willingness) to meetincreased demand. Second, increased demand willresult in problems with grid capacity (the network'sability to cope with the extra and varying energy flow):the intermittent nature of distributed PV will result incongestion and overloading of transmission and distri-bution lines. Additionally, the uncontrollable nature ofthe power generation of PV leads to uncertainties in thebase load generation, which in turn leads to economic



5 This classification is the result of an ex post grouping rather than an exante one. This means, that first the document research was made, and thentraced the authors' institutional affiliations.


losses (this issue highlights the necessity of short-termenergy storage [60]). Third, there will be economicrepercussions: high PV penetration means the displace-ment of conventional generation technologies, whichneed to be on standby mode in case solar irradiationis not high enough. These technologies need to befinancially compensated. Furthermore, extra networkinfrastructure costs are brought as a consequence ofmeeting the above grid capacity issues mentioned.These costs then may need to be recovered through ahigher tariff applied to the electricity bought from thegrid. Finally, problems such as the industry's depen-dence on subsidies, support and incentives haveresulted in specific consumer behavior (a profit mental-ity), which means that people are looking at PV as amoney-making opportunity. While this may be trueunder the current pricing schemes (FiTs), it may not beso after grid parity (see also [39,43,56,72,73]).

[9] Grid parity as an outdated and harmful concept. “Lackof support for incremental improvements, too muchfocus on champion cell results, and using announce-ments and roadmaps as data have placed the PVindustry — and all solar — firmly on the slippery slopeof unmet expectations. Profit, or simply breaking even,is sacrificed on the altar of grid parity” [74]. The core ofthis storyline is that a new target is needed. While thestoryline argues that the development, support and useof PV technology are desirable, it questions the rele-vance and use of the grid parity concept. First, it isargued that it is impossible to get a complete andcoherent picture of the industry's cost–price develop-ment process, because the price dynamics are now toofast and complex when taking into account thecomplete value chain (e.g. large number of manufactur-ing process, multiple components, choice of distributionchannels and market differences). It argues that gridparity has been useful as an abstract metric for R&Dprograms and as a tool to legitimize support for the PVindustry. However, it has become outdated as a tool forpolicy decisions because it is based on constantlyoutdated data [75]. Grid parity is pressing companiesto reduce their production and operation costs, whichresults in continuous shakeouts and decreased oppor-tunities for start-ups. The promise of being competitivewithout subsidies reduces the opportunity for companiesto enjoy revenues necessary in any industry to guaranteestability. In this situation, only stable well-establishedcompanies can survive. The storyline argues that thepromise of grid parity is ‘blackmailing’ the PV industry:continuously lower prices, or the incentives necessary tostimulate demand will be removed [76]. As a conse-quence, the storyline argues that the industry requires anew focal point, away from grid parity. Instead, theindustry should focus on developing markets withoutincentives, reducing the high upfront costs, and address-ing the various other market barriers. Additionally,before even talking about cost competitiveness, theindustry should address market externalities like CO2

abatement costs. By not including such negative exter-nalities in the current energy market, customers tend topay too little for what they are getting (electricity from


fossil fuel sources). If the benefits of renewable technol-ogies would be included, it would make sense to paymore for PV electricity (see also [74,77]).

[10] Grid parity excludes off-grid markets. “Prerequisite forgrid-parity analysis is access to an electric grid….Howevergrid-parity concept makes no sense for people withoutaccess to electricity, but most of them live in areas ofexcellent solar conditions, hence highly economic off-gridPV systems are a best adapted and least cost solutionfor their energy needs.” [sic] [33]. This storyline, too,challenges the usefulness and relevance of grid parityin the actual PV context. One highlighted issue is theexclusion of the off-grid markets: while there is a hugepotential market for PV systems in off-grid applications,especially in developing countries [75], grid paritydefinitions explicitly require the connection to theelectrical grid. This obscures the potential socialbenefits of PV systems in the local developmentprocess of developing societies (see also [78]).

4.2. Discourse coalitions

In the next step, we researched the documents by 45actors, checking their names and institutional affiliations. Wefound three broad categories of actors: (1) private persons(e.g. scientists, journalists, PV enthusiasts), (2) industryactors (e.g. manufacturers, installers, R&D companies) and(3) PV trade association actors (e.g. EPIA, Assolare).5 Makinga matrix of these actors and the various storylines in thedocuments they authored (as in the example of Table 1)yielded six discourse coalitions: combinations of actors andshared storylines. Results are shown in Table 3 and visuallysummarized in Fig. 5.

Several storylines are shared among most actors (e.g.“Cost reductions are not just needed in PV modules”; “thedrivers of grid parity”; “grid parity is almost/already here”;and “subsidies, incentives and R&D support determine gridparity”), indicating a general agreement over some elementsin the grid parity debate. Nevertheless, some storylines areshared by only a few actors (e.g. “grid parity as the keymilestone of the PV industry”; “grid parity as one of themilestones” and “grid parity as outdated and harmfulconcept”). The majority of actors are most interested inwhen the grid parity point is going to be reached (coalition Cand storyline 4). When looking at the institutional affiliationsof actors that form that coalition, it has the highestparticipation of industry members. This is not surprising,since under any definitions of grid parity, ensuring that PVtechnology keeps lowering its production cost ensuresmarket growth. Thus, the commitment from industry mem-bers towards reaching grid parity is to be expected.

Table 3 and Fig. 5 clearly show that several storylines areshared — not only by actors within a discourse coalition, butalso between coalitions. Storylines 2, 3, 4 and 5 are the mostshared, while storylines 1, 6, 9 and 10 are shared the least.

Each of the six discourse coalitions identified uses asimilar set of storylines, only being separated by one or two



Table 3Discourse coalitions and storylines (the number in square brackets represent the storyline number, as listed in Section 4.1).

Discourse coalition No. actors Storylines shared No. actors

A Grid parity as the key milestone of the PV industry 9 [1] Grid parity as the key milestone of the PV industry 9[2] Cost reductions are not just needed in PV modules 3[3] The drivers of grid parity 3[4] Grid parity is almost/already here 3[5] Subsidies, incentives and R&D support determine grid parity 2[8] Problems after grid parity 3

B Grid parity as one of the PV industry's milestones 11 [6] Grid parity as one of the PV industry's milestones 11[2] Cost reductions are not just needed in PV modules 1[3] The drivers of grid parity 5[4] Grid parity is almost/already here 6[5] Subsidies, incentives and R&D support determine grid parity 4[7] Shortcomings of grid parity concept and the industry 2

C Grid parity is at hand 16 [4] GP is almost there 16[2] Cost reductions are needed not only in the PV modules 2[3] The drivers of GP 3[5] Subsidies and incentives are important for the industry 4[10] Grid parity excludes off-grid markets 1

D Grid parity, useless and even harmful concept 3 [9] Grid parity as outdated and harmful concept 3[8] Problems after grid parity 1[7] Shortcomings of grid parity concept and the industry 1[10] Grid parity excludes off-grid markets 2

E Cost reduction not only in the modules 2 [2] Cost reductions are not just needed in PV modules 2[5] Subsidies, incentives and R&D support determine grid parity 1

F The drivers for grid parity 4 [3] The drivers of grid parity 4[5] Subsidies, incentives and R&D support determine grid parity 2[7] Shortcomings of grid parity concept and the industry 1[8] Problems after grid parity 2


divisive storylines that keep them apart. For example,coalition A (“grid parity as the key milestone of the PVindustry”) and B (“grid parity as one of the PV industry'smilestones”) only fundamentally disagrees on the relevanceof the grid parity point for the industry:

• Coalition A sustains that grid parity is the milestone of theindustry and therefore all efforts should be concentrated onreaching it. However, they recognize that cost reductionsare required not just in the modules but in the wholesystem, and that there might be problems after reachinggrid parity (e.g. grid capacity and raw material supply).When looking at the type of actors that form this groupthey are mostly private persons, writing for technicalmagazines.

• Coalition B argues that, despite the fact that reducingproduction costs is a good thing for the industry, otherissues should be addressed before subsidies and incentivesare removed in order to ensure mass adoption. Therefore,they also engage in discussion about what the grid parityconcept is ‘missing’. We interpret this to mean that as thepromised grid parity point (in some of the more optimisticassessments) approaches, other barriers come into focus.

Nonetheless, the majority of the actors seem to beconcerned about when the grid parity point is going to bereached. Coalition C (i.e. grid parity is at hand) is formed by16 actors. When looking at the type of actors that form thisgroup, it has the highest participation of industry members[32]. This is not surprising, since under any of the definitionsof grid parity, ensuring that PV technology keeps loweringits production costs ensures market growth. Thus, the


commitment from industry member towards reaching gridparity is to be expected. The remaining three coalitions D(i.e. grid parity, useless and even harmful concept), E(i.e. cost reduction not only in the modules) and F (i.e. thedrivers for grid parity), are relatively small in sample,composed of about 11% of the authors:

• From our data, we were not able to deduce a clear positiontowards the question of the relevance of grid parity forcoalitions E and F. An argument could be made for omittingthem as distinct coalitions, as all shared storylines in thesecoalitions are already ‘covered’ by other coalitions (see: Fig. 5).However, we opted to conceptualize them as distinctcoalitions because, importantly, they do not share the twostorylines that distinguish between coalitions A and B(storylines 1 and 6, respectively).

• Coalition D, however, on the other hand, adopts a clearposition in the grid parity debate: it questions the usefulness ofgrid parity and criticizes the ubiquitous focus of the concept.Actors in this coalition reject its use, however, for differentreasons. For instance, it is argued that the shift of focus fromindustry failures toward a cost target is harmful for companiesin the sector and that there is the need for a new target.Starting companies are not having the required conditions tosucceed, or maintain themselves in the market, due toconstant pressure (i.e. cost reductions) from well-establishedcompanies. Therefore grid parity is harmful for the industry,and support should focus on new companies instead ofmaking the strong ones more cost effective.

This classification was supported by the answers providedby several actors to the survey sent to them. As mentionedbefore, in order to triangulate [26] our identification of the



Fig. 5. Coalition diagram: circles (with letters) correspond todiscourse coalitions;black dots (with numbers) correspond to storylines.The size and shape of the circles are not meaningful. The diagram wasconstructed with the sole purpose of showing how a few storylines separatedthe different coalitions, and thatmutual understanding and discourse affinity isfound in the debate. General beliefs are shared across the analyzed actors, butsome core beliefs bring them apart and create the debate.


various discourse coalitions, a questionnaire was sent todifferent stakeholders (including actors from each of the sixgroups identified, as well as PV industry associations and PVcompanies across Europe). This questionnaire was formed bya set of general questions, with the objective of betterunderstanding the actor's position towards grid parity. Fromtheir answers, it was clear that there were three mainopinions regarding the concept. Some of the stakeholdersconsulted agreed that grid parity indicates the starting pointfor the up-scaling process (i.e. mass adoption). Some othersindicated that grid parity has served its purpose, but thatnow the time has come to move to the other barriers thatneed to be addressed before mass adoption can take place.Finally, there was a group of stakeholders that rejectedthe use of the concept, either because of the shortcomings ofthe concept or because it is still far away from being achieved.Thereby, the questionnaire confirmed the existence ofcoalitions A, B and D. Regarding the remaining threecoalitions (i.e. C, E and F), while the questionnaires did notallow to confirm their existence, the document analysisconducted for the discourse analysis showed enough evidenceto keep them as separate coalitions.

5. Conclusions and discussion

The concept of grid parity appears to be used exclusivelyby proponents of solar PV technology. Adopting a role in away quite similar to ‘acid rain’ in 1980s environmentaldiscourse [7], ‘grid parity’ is used by actors in the PV field as ashort-hand ‘cue’ to evoke a specific, more complex, narrativein other actors' minds. And again as in Hajer's ‘acid rain’ study[14], this assumption of mutual understanding is demonstra-bly false. We have shown the term, which is commonlyunderstood as referring to the point in time when solar PVelectricity costs the same as competing conventional sources


or the grid electricity price, to be only deceptively simple: toarrive at a meaningful estimation of the grid parity point,assumptions made in each step of the calculations have to bearticulated and carefully evaluated. Nevertheless, we foundthat this is almost never done: scenarios are made withoutclear information or justification of the assumptions made,and the grid parity studies performed invariably use thesimplest model available of grid parity as a ‘black box’ withthree inputs (solar PV market growth, learning rate andelectricity price) and one output (the grid parity year). Usingthe term thisway is a political act, which obfuscates issues suchas the following: Which PV technology? Which conventionalsource it is compared to? Which grid electricity price? Areexternalities being internalized?

We have used a discourse analysis to open up this blackbox. Our analysis has yielded ten storylines and six actorcoalitions (which are ensembles of actors and the storylinesthey share). Yet mapping these storylines and coalitionsreveals that, while several storylines are shared by multiplecoalitions, others are not. These are what we call divisivestorylines, which indicate (and advocate) different strategiesfor the future of solar PV. We have shown that PV discourse isnot homogeneous, and that analyzing the grid parity debatereveals tensions and conflicts in the PV community aboutfuture directions.

We have found that the majority of actors in what we havecalled coalitions A (“grid parity as the key milestone of the PVindustry”), C (“grid parity is at hand”), E (“cost reduction notonly in the modules”) and F (“the drivers for grid parity”)indeed use the term as a black box, providing either noconcrete definition of grid parity or the very simplest one. This‘version’ of grid parity promises competitiveness, mass adop-tion and self-sustainability (i.e. no need of incentives andsupport mechanisms) and gives a positive perspective of theindustry's future, showing that it will become an economicallyviable option. It helps to articulate positive expectations in theindustry and legitimizes mechanisms, reduces uncertaintiesabout future performance and attracting new actors. It isbelieved that after reaching grid parity, the industry willexperience a boom in the demand.

Yet different storylines have emerged as the predictedgrid parity point approaches and these promises do not seemtomaterialize. Actors who argue that reaching grid parity willnot be enough for PV to become a success reframe grid parity,(re-)introducing some of the complexity to ‘fix’, or evenarguing that it is harmful as it might blind actors to the otherbarriers and issues necessary for the up-process of the PVindustry. The majority of actors in what we have calledcoalitions B (“grid parity as one of the PV industry'smilestones”) and D (“grid parity, useless and even harmfulconcept”) do highlight these issues and expose the underly-ing complexity of the concept.

The degree of ‘vagueness’ that actors display in using theconcept of grid parity is thus indicative of their preferredstrategy for the PV industry: a more complex definition of theconcept emphasizes that there are issues and blank spaces tobe filled, while a more simple (or no) definition avoids anydiscussions and highlights the apparent inevitability of gridparity. The ‘vagueness’ with which the grid parity concept ispredominantly used, has, in a sense, been valuable. In Hajer'sterms, the fact that the assumption of mutual understanding




(when using the concept) is false, “can be very functional forcreating a political coalition.…people that can be proven tonot fully understand one another can nevertheless togethercan produce meaningful political interventions” [14 pp. 69].

The ubiquitous use of ‘the simple version’ of grid parityhas focused attention squarely on the economic and financialbarriers that PV technology faces. This has gone at the cost,however, of attention for other barriers. Actors who seek toemphasize these other barriers, thus need to marginalize ordelegitimize (the hegemonic version of) the grid parityconcept by attempting to fill it with a more precise, andmore complex, meaning. This is how we understand the roleof the grid parity debate: as a key ‘battleground’ in a broaderdiscursive conflict over the optimal PV up-scaling strategy.The concept of grid parity has served the purpose ofobtaining resources (e.g. via support mechanisms) andsteering/guiding activities (as a heuristic). However, whiledoing so, much of its complexity is left out as this does notserve the message about PV competitiveness that actors wantto make. This conclusion is supported by the fact that gridparity calculations only use the relatively simple log-linearmodel and exclude many variables. However, this does notmean that actors doing so are unaware of the complexity ofthe concept: their simplification of it is strategic Interestingly,our discourse analysis of the grid parity debate revealed aplethora of ideas on PV up-scaling strategies. By opening upthe black box of grid parity we revealed tensions aboutpreferred strategies in an otherwise seemingly homogenousPV discourse.

However, we also make the following qualifications. First,the fact that all actors who use the grid parity conceptseem to be generally in favor of PV may stem in part fromthe databases used for acquiring the written utterancesfor our discourse analysis. Second, some discourse theorists(e.g. [14]) argue that because discourse is reproducedthrough an identifiable set of practices, one should examinethese practices in conjunction with written utterances. Whileour analysis of the written utterances by 45 actors wasmethodical, our exploration of their practices (e.g. dissemina-tion of the discourse through presentations etc.) was ratherad-hoc, and so, we cannot claim to havemapped their practicessystematically. Third, we would briefly address our own rolein determining the boundaries of the grid parity discourse.As Jorgensen and Phillips argued “(…) a discourse is notsomething that the researcher finds in reality, rather, it isconstructed analytically with a point of departure in theresearch questions” [8 pp. 147]. In this perspective, dis-courses do not independently ‘exist’ (which would amountto reification of discourse). Instead, they are analyticalcategories (co-)constructed by their researchers. Thismeans, however, that these researchers “…have to estab-lish…that the delimitation they have made is reasonable.Delimitation can begin with the aid of secondary literaturethat identifies particular discourses, but obviously the workcontinues in the analysis of the material”. We believe wehave met this criterion by first engaging in a criticalassessment of the evolution of the grid parity concept inSection 3, which indicated that conceptualizing grid parity asa discoursewould bemeaningful. Subsequently, in our analysisin Section 4, we then proceeded to explore its constituentstorylines.


References

[1] L. El Chaar, L. Lamont, N. El Zein, Review of photovoltaic technologies,Renew. Sust. Energ. Rev. (2011) 2165–2175.

[2] B. Verhees, R. Raven, F. Veraart, A. Smith, F. Kern, The development ofsolar PV in The Netherlands: a case of survival in unfriendly contexts,Renew. Sustain. Energy Rev. 19 (2013) 275–289.

[3] E. Beck, F. Martinot, Renewable energy policies and barriers, Encycl.Energy (2004) 365–383.

[4] J. Painuly, Barriers to renewable energy penetration: a framework foranalysis, Renew. Energy (2001) 73–89.

[5] J. Poortmans, W. Sinke, The strategic research agenda of the EuropeanPV technology platform, IEA: Energy Technology Roadmap Workshop,PV ERA NET 3rd Joint Workshop, Paris-France, 2008.

[6] M. Brown, Going for grid parity: making solar power economicallycompetitive, Frontiers, BP, 2005, pp. 6–10.

[7] M. Hajer, Discourse coalition and the institutionalization of practice:the case of acid rain in Britain, The Argumentative Turn in PolicyAnalysis and Planning, Duke University Press, London, England, 1993.

[8] M. Jorgensen, L. Phillips, Discourse Analysis: As Theory and Method,SAGE Publications Ltd., London, England, 2002.

[9] E. Laclau, C. Mouffe, Hegemony and Socialist Strategy: Towards aRadical Democratic Politics, Verso, London, U.K., 1985

[10] N. Fairclough, Critical Discourse Analysis, Longman, London, U.K., 1995[11] N. Fairclough, R. Wodak, Critical discourse analysis, in: T. van Dijk (Ed.),

Discourse as Social Interaction: Discourse Studies: A MultidisciplinaryIntroduction, Sage, London U.K., 1997

[12] D. Edwards, Discourse and Cognition, Sage, London, U.K., 1996[13] D. Edwards, J. Potter, Discursive Psychology, Sage, London, U.K., 1992[14] M. Hajer, Doing discourse analysis: coalitions, practices, meaning, in: M.

van den Brink, T. Metze (Eds.), Words Matter in Policy and Planning:Discourse Theory and Method in the Social Sciences, NetherlandsGeographical Studies, Utrecht, Netherlands, 2006.

[15] IEA, Experience Curves for Energy Technology Policy, IEA Publications,Paris, France, 2000.

[16] D. Poponi, Analysis of diffusion paths for photovoltaic technology basedon experience curves, Sol. Energy (2003) 331–340.

[17] B. van der Zwaan, A. Rabl, Prospects for PV: a learning curve analysis,Sol. Energy (2003) 19–31.

[18] G. Schaeffer, H. de Moor, Learning in PV trends and future prospects,19th European PV Solar Energy Conference and Exhibition, ECN, Paris,France, 2004, pp. 0–7.

[19] K. Neuhoff, Large Scale Deployment of Renewables for ElectricityGeneration, Oxf. Rev. Econ. Policy 21 (2005) 88–110.

[20] B.A. Sanden, The economic and institutional rationale of PV subsidies,Sol. Energy (2005) 137–146.

[21] M. Staffhorst, The way to competiveness of PV—an experience curveand break-even analysis, Kassel University, 2006.

[22] W. van Sar, E. Alsema, H. de Moor, G. Schaeffer, Accuracy of progressratios determined from experience curves: the case of crystallinesilicon photovoltaic module technology development, Prog. Photovolt.Res. Appl. (2008) 441–453.

[23] R. Bhandari, I. Stadler, Grid parity analysis of solar photovoltaic systemsin Germany using experience curves, Sol. Energy (2009) 1634–1644.

[24] P. Lund, Boosting new renewable technologies towards grid parity:economic and policy aspects, Renew. Energy (2011) 2776–2784.

[25] EPIA, Solar photovoltaics competing in the energy sector: on the roadto competitiveness, EPIA, Brussels, Belgium, 2011.

[26] R. Ying, Case Study Research: Design andMethods, Fourth edition SAGEPublications, California, USA, 2009.

[27] R.H. Wozniak, Classic in the History of Psychology: Introduction toMemory, Hermann Ebbinghaus (1985/1913), 1999.

[28] M. Anzanello, F. Fogliatto, Learning curve models and applications: liter-ature review and research directions, Int. J. Ind. Ergon. (2011) 573–583.

[29] M. Junginger,W. van Sark, A. Faaij, Technological Learning in the EnergySector: Lessons for Policy, Industry and Science, Edward Elgar Pub,Northamptom, MA, USA, 2010.

[30] IEA, Experiences Curves for Technology Policy, IEA Publications, Paris,France, 2000.

[31] W. Zangwill, P. Kantor, The learning curve: a new perspective, Int.J. Ind. Ergon. (2000) 595–607.

[32] L.A. Hurtado, The power of grid parity: unraveling its role and relevance inthe up-scaling of the PV industry, Industrial Engineering and InnovationSciences, Technische Universiteit Eindhoven, Eindhoven, Netherlands,2012, p. 161.

[33] C. Breyer, A. Gerlach, Global overview on grid-parity event dynamics,25th European PV solar Energy Conference and Exhibition, Valencia,Spain, 2010.

[34] C. Breyer, A. Gerlach, J. Mueller, H. Behacker, A. Milner, Grid-parity analysisfor EU and US regions and market segments—dynamics of grid-parity and




dependence on solar irradiance, local electricity prices and PV progressratio, 24th European PV Solar Energy Conference and Exhibition, Hamburg,Germany, 2009, pp. 4492–4500.

[35] L. Dittmar, J. Henkel, N. Ehlers, What is grid parity? A contribution to thediscussion about a frequently used but rarely understood term, 34th IAEEInternational Conference: Institutions, Efficiency and Evolving EnergyTechnologies, Stockholm, Sweden, 2011.

[36] T. Esram, P. Krein, B. Kuhn, R. Balog, P. Chapman, Power electronics needsfor achieving grid-parity solar energy costs, Energy 2030 IEEE, 2008.

[37] A. Neslen, F. Simon, Special report: EU on track for solar grid parity by2017, EurActiv, 2011.

[38] First-Solar, Annual report, First Solar, 2009.[39] A. Hutchinson, Solar panel drops to $1 per watt: is this a milestone or

the bottom for the silicon-based panels? Popular Mechanics, 2009.[40] A. Levine, Accelerating the movement to grid parity in photovoltaic

manufacturing, Future Photovoltaics, 2010.[41] D. Lewis, Solar grid parity, Engineering & Technology, 2009.[42] T. Olbrecht, Grid parity—the holy grail? PV News, EuPd Research, 2011,

(pp. http://www.eupd-research.com/de/news/index.php?id=49&tx_ttnews%45Btt_news%45D=769&txttnews%765BbackPid%765D=750&cHash=760f769ecf797b767).

[43] G. Parkinson, Solar PV at grid parity! Now what? Climate Spectator,2011.

[44] Sharp-Corporation, Environmental and social report 2011, Sharp, 2008.[45] J. Farrel, Grid parity for solar PV with balance of system cost reductions,

Renewable Energy World, 2011.[46] W. Sinke, Grid parity: holy grail or hype? International Sustainable

Energy Review, 2009, pp. 34–37.[47] H. Trabish, New study: solar grid parity is here today, Green-techmedia,

2011.[48] APVA, Australian PV survey report 2010, APVA, 2011.[49] APVA, PV market developments in Australia, at the crossroads? EcoGen

2011, APVA, Brisbane, Australia, 2011.[50] R. Baile, Solar's search: seeking the elusive milestone, Renewable

Energy World, 2011.[51] C. Breyer, A. Gerlach, C. Werner, Grid parity: coming sooner than you

think, Future Photovoltaics, 2011.[52] F. Didier, PV competing in the energy sector: grid parity and beyond,

PV SEC—8th European Industry Submit—CEO Panel Discussion, EPIA,Hamburg, Germany, 2011.

[53] EPIA, B2b work on the 4 September 2008 —grid parity, EPIA, 2008,(pp. http://www.epia.org/events/past-events/epia-events-within-the-last-year/b2b-workshop-on-the-4-september-2008-grid-parity.html).

[54] M. Feinstein, Does grid parity matter? PV Magazine, 2011.[55] P. Gipe, New German solar PV tariff, the price to beat worldwide,

Renewable Energy World, 2011.[56] C. Morris, P. Plante, Feed-in tariffs needed after grid parity, Renewable

Energy World, 2011.[57] Shareholders-Unite, The real issue with solar energy isn't its cost,

Seeking Alpha, 2011, (pp. http://seekingalpha.com/article/306468-the-real-issue-with-solar-energy-isn-t-its-cost).

[58] W. Schirtzinger, Does grid parity matter? Renewable Energy World,2010.

[59] R. Swanson, PV: from the present to 2020, Future Photovoltaics, 2011.[60] P. Bhamre, PV at the edge of grid parity—the dawning era of a new era,

2011. (pp. http://www.eupd-research.com/de/news/index.php?id=49&tx_ttnews%45Btt_news%45D=831&txttnews%835BbackPid%835D=850&cHash=ab5711795f).

[61] P. Bhamre, Europe focus: dawning of a new era beyond grid parity—thethird demand cycle: PV power generators, 2011. (pp. http://www.pvgroup.org/node/1856).

[62] Young Ernst, UK solar PV industry outlook: the UK 50 KW to 5 MWsolar power market: industry outlook, 2011.

[63] FirstGreentech,Glossary of terms—solar PV installation, FirstGreentech2011.(pp. http://www.firstgreentechsolar.co.uk/glossary.htm).

[64] GetSolar, Grid parity: are we there yet? GetSolar, 2011, (pp. http://www.getsolar.com/News/Solar-Energy-Facts/General/Grid-Parity-Are-We-There-Yet-800531257).

[65] GetSolar, Question answered: solar has already met the grid, GetSolar,2011, (pp. http://www.getsolar.com/News/Solar-Energy-Facts/General/Question-Answered:-Solar-Has-Already-Met-the-Grid-800659857).


[66] HanwhaSolarOne, Company news, HanwhaSolarOne, 2010, (pp. http://www.hanwha-solarone.com/en/news/company-news/company-news-006).

[67] OerlikonSolar, Press releases: closing the gap to grid parity, OerlikonSolar,2010, (pp. http://www.oerlikon.com/ecomaXL/index.php?site=SOLAR_EN_press_releases_detail&udtx_id=7299).

[68] OerlikonSolar, The way to grid parity, OerlikonSolar, 2010, (pp. http://www.oerlikon.com/ecomaXL/index.php?site=SOLAR_EN_grid_parity-1).

[69] M. Pinto, Has the sun finally risen on photovoltaics? Symposium onVLSI Technology, IEEE, Honolulu, Hawaii, 2008, pp. 2–5.

[70] PowerEngineer-RenewableEnergy, Photovoltaic cells: will grid paritychange everything? PowerEngineer: Renewable Energy, 2009.

[71] C. Yang, Reconsidering solar grid parity, Energy Policy (2010) 3270–3273.[72] B. Elliston, I. MacGil, M. Diesendorf, Grid parity: a potentially misleading

concept? Solar 2010, the 48th AuSES Annual Conference, University ofSouth Wales, Canberra, Australia, 2010, pp. 1–10.

[73] P. Mints, Solar PV perspective 2011: part one—Paula Mints looks at thebackground of the great solar squeeze, pre Solyndra, Renewable EnergyFocus, 2011.

[74] P. Mints, Stop the solar industry's self-destruction over grid parity, 2012.(pp. http://www.electroiq.com/articles/pvw/2012/february/stop-solar-self-destruction-over-grid-parity.html).

[75] M. Bazilian, M. Liebreich IjeomaOnyeji, I. MacGil, J. Chase, J. Shah, D.Gielen, D. Arent, D. Landfear, Z. Zhengrong, Reconsidering the economicsof photovoltaic power, Bloomberg New Energy Finance, 2012.

[76] P. Mints, The pied piper of grid parity, Renewable Energy World,2012.

[77] S. Letendre, The grid parity fallacy, Renewable Energy World, 2010.[78] Ch. Breyer, A. Gerlach, M. Hlusiak, C. Peters, P. Adelmann, J. Winiecki, H.

Schützeichel, S. Tsegaye, W. Gashie, Electrifying the poor: highlyeconomic off-grid PV systems in Ethiopia—a basis for sustainable ruraldevelopment, Proceedings 24th European Photovoltaic Solar EnergyConference, Hamburg, 2009, September 21–25, 2009, http://dx.doi.org/10.4229/24thEUPVSEC2009-5EP.2.3.

Luis Hurtado Munoz (1987) MSc is a PhD student at the Department ofElectrical Engineering of the Eindhoven University of Technology researchingthe optimization of energy utilization in the built environment (officebuildings) and the interaction with the electricity grid. Before starting hisPhD he was a master's student in Sustainable Energy Technology (SET) at thesame university. This article is the result of his master's thesis on the role of thegrid parity concept in PV up-scaling processes which he performed atthe Department of Innovation Sciences under the supervision of BoukjeHuijben and Geert Verbong.

Boukje Huijben (1986) MSc is a PhD student at the Department ofInnovation Sciences of the Eindhoven University of Technology. For herPhD she performs research on the implementation and up-scaling of solarenergy in The Netherlands using both business model and transition studies(Strategic Niche Management) literature. Previously she studied Chemistryand Environmental Sciences at the Radboud University of Nijmegen (i.e.both obtained in 2010).

Bram Verhees (1977) PhD is a postdoctoral fellow at the Department ofInnovation Sciences of the Eindhoven University of Technology researching theconcept of protection, which is central to Strategic Niche Management (SNM)and transition studies but not yet systematically interrogated. Bram holds aPhD in Innovation Sciences from the same university. For his PhDhe researchedthe role of cultural legitimacy in innovation journeys using discourse analysis.

Geert Verbong (1955) PhD is a Full Professor in the School of InnovationSciences at TU/e. He has been a core member of the Dutch KnowledgeNetwork on System Innovations or Transitions, working on the socialdimensions of smart grids and the implementation of renewable energysources. Currently he is also manager and coordinator at the TU/e EindhovenEnergy Institute. Recent publications include books on the history ofrenewable energy in The Netherlands (2001), the Dutch Energy ResearchCentre (2005) and Governing the Energy Transition (2012). He has managedseveral research projects and provides policy advices.



the power of grid parity: a discursive approach

Documents