Scope of the report
The objective of this report is to estimate the current public and industrial R&D investments directed towards SET-Plan priority technologies in the EU. These comprise wind energy; concentrating solar power (CSP) and solar photovoltaic (PV); carbon dioxide capture and storage (CCS); smart grids; transport biofuels; hydrogen and fuel cells; nuclear fission (with a focus on generation IV reactors); and nuclear fusion.
The assessment is focused on a single indicator representing research and development inputs: the R&D investments. R&D outputs, such as innovation surveys or analyses of patents that are considered valuable indicators for innovation activity (Griliches, 1990; Jaumotte and Pain, 2005), are not included. The efficiency of the innovation process is thus not considered.
Besides, the indicator 'R&D investments' may be too narrow for capturing the scope of industrial R&D activities, parts of which may be conducted within departments or groups that are not formally designated as 'R&D' departments (see Freeman and Soete, 2009 with further references Freeman and Soete (2009) also apply Goodhart's law (Goodhart,1972) to Science, Technology and Innovation Indicators: once these indicatorsare made policy targets, they loose much of their information content thatqualified them to play such a role.). Furthermore, a considerable part of innovation in the energy sector takes place on the side of the component supplier, i.e. new technologies are rather bought in by the energy companies than being developed in-house (Jacquier-Roux and Bourgeois, 2002; Kaloudis and Pedersen, 2008). Even though the approach of the present report aims at going beyond the boundaries of the classical energy sector, it can capture this phenomenon only to a limited extent, thus leading to a systematic under-estimation of innovation efforts related to energy. Future work that considers major parts of the supply chains in various energy sectors might be able to overcome this.
Above and beyond, innovation depends on a wide variety of factors throughout all phases of the innovation chain, i.e. the scientific research, development and market introduction of new technologies. Hence, a more comprehensive approach would need to consider the broader context of 'push' and 'pull' policies and measures that address the research, development, demonstration and deployment of innovative technologiesEmpirically, Johnstone et al. (2008) show on the basis of anassessment of exhaustive panel data on patent applications that both, dedicatedR&D spending and (quantityand price-based) 'pull-policies' are significant determinants of patenting inrenewable energy. (see Figure 2), as well as institutional capacities, the role of stakeholders, related policies and measures and their use and interplay (see e.g. Foxon, 2003; Grubb, 2004; Kaloudis and Pedersen, 2008, chapter 5.2).
Notwithstanding the above-listed limitations of the focus on one single indicator, the present analysis of energy-related R&D investments can help in better understanding the status quo of one central step in the innovation cycle, without claiming to provide an indication of the total European innovation capacities with regard to SET-Plan priority technologies.
Figure 2: Steps of technological innovation and scope of the assessment
Source: Grubb, 2004 (original figure)
But even the present assessment which focuses only on R&D investments requires an upfront definition of the research, development and demonstration activities that it covers:
R&D covers three activities according to the Frascati Manual (OECD, 2002): basic research, applied research, and experimental development. Basic research is experimental or theoretical work undertaken primarily to acquire new knowledge of the underlying foundation of phenomena and observable facts, without any particular application or use in view. Applied research is also original investigation undertaken in order to acquire new knowledge. It is, however, directed primarily towards a specific practical aim or objective. Experimental development is systematic work, drawing on existing knowledge gained from research and/or practical experience, which is directed to producing new materials, products or devices, to installing new processes, systems and services, or to improving substantially those already produced or installed. RD&D includes demonstration in addition.
The degree to which the financing of different RD&D activities are included in the figures of the present report differs between industrial and public national and EU funds as well as across individual Member States The aim of this section is toclarify the scope of RD&D covered in the present report :
- Data on public national R&D investments of EU Member States are generally taken from the IEA RD&D statistics (see section 2.3.2) or the GBOARD (socio-economic objective 5). While the latter focuses on R&D, the IEA also includes demonstration activities. In practice, however, most Member States do either not provide data on funds directed towards demonstration or do not display them separately. Hence, data on aggregated public national funds of EU Member States dedicated to demonstration amount to some 9% of the total energy R&D budget, only. Differences in the share of funds directed towards demonstration occur between technologies, as illustrated in Figure 3.
Given the high amounts needed for large-scale demonstration projects and considering the data gaps, the figure on public support to demonstration activities included in the IEA database seem to be an under-estimation at the aggregated EU level. In the following, we thus assume that the IEA database largely focuses on R&D (a hypothesis that is supported by the large similarity of the aggregated EU figures with data from the GBOARD, the latter of which focusing on R&D only).
Also that basic research shall be excluded in the IEA database unless it is clearly oriented towards energy-related technologies. However, in some Member States, the institutional budgets included in the submission data may partially cover research of more basic nature.
Figure 3: Share of demonstration activities in the IEA RD&D statistics, 2007
- Regarding the EU public R&D spending, only funds within the 6th Research Framework Programme have been assessed. While these indeed include some support to demonstration activities, their main focus lies on R&D.
- For estimating industrial R&D investments, companies' annual report (largely via the 'EU Industrial R&D Investment Scoreboard') have been used as a starting point. They thus follow the accounting definitions of R&D, such as within the International Accounting Standard 38 ('Intangible Assets'), which uses the definition of R&D of the Frascati Manual (OECD, 2002). In general, technology demonstration mostly incurs engineering costs and is thus recommended to not be included under R&D investment. However, this can be expected to strongly depend on the type of sector/activity, influenced e.g. by the maturity of the technology and/or the policy support to its deployment.
- Following these considerations, the term R&D will be used in the following despite the fact that demonstration activities are included to a certain extent that varies across different funders, countries and companies. Moreover, even within the category 'R&D' systematic differences may occur, for example between public and industrial research with the former often focusing on research of a more basic nature, while industry tends to finance more applied research.
Industrial R&D investments
Allocating companies' annual R&D investments by technology
Data on corporate R&D expenditures are difficult to obtain (see e.g. de Nigris et al., 2008; SRS Project, 2008; van Beeck et al., 2009; Wiesenthal et al., 2008). The data become even sketchier when looking at the R&D expenditure by technology. The difficulties can be explained by the lack of a regulatory framework that obliges private companies to report their R&D investment, the fact that some companies consider this information as confidential, and that others use them for strategic purposes (Gioria, 2007).
For this reason, there is currently no single database that can provide a comprehensive overview of industrial R&D investments dedicated to individual technologies. In order to nevertheless gather a maximum of information, various data sources have been assessed in parallel in this report, all of which have their own specific strengths and shortcomings. The most promising method turned out to be a novel and assumption-based approach. It combines data on R&D investments of individual companies, primarily taken from the EU Industrial R&D Investment Scoreboard (Hernández Guevara et al., 2008), with additional information about the company, thus allowing a rough estimation of a companies' R&D investment by SET-Plan priority technology.
This bottom-up approach has thus been chosen as a first option for deriving an estimation of the corporate R&D investments in the relevant technological fields. In brief, this approach includes the following steps:
- The identification of key industrial players for a certain sector.
- The gathering of information on R&D investments of these companies preferably through the EU Industrial R&D Investment Scoreboard, but also web-published annual reports or direct contacts.
- For companies active in several technological fields (such as electrical utilities, oil companies, component suppliers), an allocation of the R&D investments to individual technologies to the extent possible. For companies that are specialised in one technological field only, their entire R&D expenditure was allocated to that technology.
- The summing up of the individual companies' R&D investments by technology.
This approach has been complemented by data extracted from official databases (BERD) and EU-financed projects (e.g. SRS NET: Scientific Reference System on new energy technologies, energy end-use efficiency and energy RTD; and ERMINE: Electricity Research Road Map in Europe) as well as other literature, as sketched out in Figure 4. This is described in more detail in the following sections.
Figure 4: Schematic overview of the methodology applied for estimating corporate R&D investments by technology
Source: Present report
Breaking down corporate R&D investments
As introduced above, an allocation of individual companies' annual R&D investment to the various technologies of interest was the central approach followed in the present report.
The EU Industrial R&D Investment Scoreboard proved to be the single most important data source for obtaining basic information on annual corporate R&D investments (Hernández Guevara et al., 2008), which was used as a starting point for the subsequent assessment. The data analysed relate to the companies R&D investments in 2007.
The EU Industrial R&D Investment Scoreboard provides data on investment in R&D from 2000 companies (1000 EU-based and 1000 non-EU based). It is prepared from companies' annual audited reports and accounts Ageneral concern raised with regard to assessing R&D expenditures based onthe companies' annual reports is that the figures may include the partsfinanced through public budgets, thus creating problems with double-counting(SRS project, 2007). This problem does, however, not occur for the Scoreboarddatabase, which only includes the R&D financed by the company as a generalprinciple. If disclosed, the externally funded R&D parts (i.e. by publicsector as well as companies outside the group) are deducted. In case that acompany does not disclose the part of the R&D that has been externally funded,it cannot be deducted and the public part is thus included in the company'sinvestment, thus accepting a slight systematic error. The companies are grouped by sectors of activity following the ICB classification (Industry Classification Benchmark). Companies are allocated to the country of their registered office, which may differ from the operational or R&D headquarters in some cases Notethat the way in which corporate R&D investments are allocated to countriescan significantly influence the outcome of the analysis. The EU IndustrialR&D Investment Scoreboard allocates companies to the country of theirregistered office, while BERD refers to all R&D activities performed bybusinesses within a particular sector and territory, regardless of the locationof the business's headquarters. This important difference needs to be kept inmind when comparing results from different studies one with another..
The Scoreboard's breakdown by field of activity does not allow for the technology-oriented grouping required for this work. Furthermore, Jacquier-Roux and Bourgeois (2002) showed that much of the R&D efforts relevant for the energy sector are being carried out by the supplier of energy equipment, making the energy sector a supplier-dominated sector in the taxonomy developed by Pavitt (1984). In order to capture the R&D investments of the different types of companies, the present report therefore had to identify a number of companies that are considered relevant for research in a certain SET-Plan priority technology instead of relying on existing classifications of companies (such ICB or NACE, the statistical classification of economic activities). The assessment of the present report is thus based on the estimation of R&D budgets from both traditional energy companies - such as 'oil and gas producers'; 'electricity'; 'gas, water and multiutilities' following the ICB classification - and companies that are active in sectors like 'alternative energy'; 'automobiles and parts'; 'chemicals'; 'construction and materials'; 'electrical components and equipment'; 'industrial machinery'; 'industrial metals'; and 'general industrials'.
For each of the technological fields of interest, key industrial EU-based companies have been identified. This has been based on analyses of the various markets (e.g. through the barometers from EUROBSERV'ER 2008a-c), expert knowledge and other sources (such as the members of Technology Platforms or associations; companies' internet websites). Of course, a number of companies are active in various fields simultaneously, which meant that they figure in various 'technology groups' at the same time.
Such an approach bears the risk of missing a central player for R&D as the selection is strongly based on the market share of companies. However, a large market share does not necessarily imply a high R&D intensity. This is supported by the fact that innovation may happen on the component supplier side rather than within the known energy companies. Since the lists of companies that are considered within a certain field are not exhaustive, neglecting minor players that might, in sum, provide a far greater R&D commitment, they tend to underestimate the total R&D efforts dedicated to SET-Plan technologies.
Overall, a total of 136 companies were identified as central R&D investors on the SET-Plan priority technologies, a large number of which being simultaneously active in several SET-Plan technologies. 72 of the 136 companies identified are listed among the TOP1000 R&D investors of the Scoreboard, allowing the direct extraction of R&D investment data. However, for four of them, no further breakdown of the R&D investment by technology could be performed. Some of the companies of interest are smaller overall R&D investors and therefore do not rank among the TOP1000 R&D investors; as they are nevertheless important for a certain (smaller) technological field, an in-depth research has been carried out by looking at a more detailed database containing some 7000 companies, which is the basis underlying the EU Scoreboard. Data for 12 companies was extracted from this extended database. A web-based search provided some information for 26 additional enterprises through e.g. annual reports or other information for those companies that are not listed on the stock exchange and thus are not obliged to publish their financial report. Haug et al. (2009) provided information for nine additional companies active in research on CSP. Combining data from the various sources, data were available for 115 out of the 136 companies identified.
The R&D Investment Scoreboard does not allocate the R&D investments to individual energy technologies. For companies being active mostly in one technological field (e.g. Vestas for wind power), one may assume that their R&D investment remains within their main field of business activity. This has been the case for around 25 companies, for which the R&D investments by technological field could thus be identified with a high accuracy. For another eight companies, a very high confidence level could be obtained in the assumptions made in the present analysis building on information from official sources (see also chapter "Analysis of uncertainties").
For large companies operating in manifold areas, however, the amount of research investments that is dedicated to a certain technology cannot be directly derived from the R&D Investment Scoreboard. This is the case in particular for car manufacturers, oil companies, electric utilities, and large component manufacturers. For this reason, the Scoreboard data proved to be of more direct help in the areas of PV and wind than for CCS, biofuels or hydrogen/fuel cells, with large multinational utilities and oil businesses being more active in the latter fields.
For the purpose of this report, the R&D investments of multi-business companies was divided up and allocated to the different technologies on the basis of assumptions. Companies' annual reports and corporate sustainability reports were systematically analysed for additional information on the breakdown of R&D investments. Moreover, the websites of individual companies and associations were screened for further information, enhanced by free searches that delivered additional information in the form of e.g. presentations and speeches from company key actors or press releases. In very few cases, newspaper articles provided helpful information. For some companies, work that had previously been carried out at DG RTD and TREN has been used as input (Naneva and Paschos, 2008). In addition, information from literature (New Energy Finance, 2008; PWC et al., 2007) was used as a starting point. Furthermore, some 30 companies or industry associations were contacted. This direct contact allowed obtaining exact figures for four large companies, and helped to refine assumptions for another handful of companies.
In the easiest cases, this additional information revealed the allocation of the R&D investment to the different technologies. For most companies, however, the R&D expenditures could be narrowed down to a particular field (e.g. 'renewables') with a certain accuracy but then needed to be further split between the various renewable energy sources based on qualitative information. In that cases, some substantiated "guess-timates" had to be performed in order to allocate their R&D investment to single technologies, based on information available for the individual companies obtained through the sources described above.
This included, for example, the number of researchers by field that allowed a rough estimation of the R&D investments by applying an average R&D investment per research employee. Figure 5 shows information on R&D investment per research employee gathered for 31 companies or research centres. An average investment of €120000 to €150000 per research employee was found for 55% of them. This range was then used for further estimates, unless more precise figures could be obtained for the specific company.
Figure 5: Distribution of R&D investments per research employee
Source: Own analysis based on a variety of information sources
Other companies announced some future R&D investment plans, which were subsequently "extrapolated" back to the 2007 data. In few cases, the substantiated assumptions were applied to other, similar companies as well. Often, the R&D expenditures could be narrowed down to a certain field (e.g. 'renewables') with a certain accuracy but then needed to be further split between the various renewable energy sources based on qualitative information.
For some selected important R&D investors, patent applications have been used as an indicator of the R&D breakdown. Based on the assumption that patents may reflect a company's research effort, which is supported by assessment that show a significant correlation between patents and R&D spending (e.g. Griliches, 1990; Jaumotte and Pain, 2005), the distribution of patents across the relevant technologies was used as a proxy for the distribution of its R&D expenditures. Linking input indicators such as energy R&D spending to output indicators (such as patents Theoutputs of research are manifold, ranging from the better understanding ofpresently used techniques to finding new technologies and attracting publicresearch funding (see overview table in Ernst, H., 1998, p.3). The use ofpatents for measuring R&D outputs may thus fall short of the complexity ofthe motivations for research, but can nevertheless be used as a suitableindicator (Griliches et al., 1986; Griliches, 1990). However, Ernst (1998)shows that research does not necessarily lead to more patents, but to patentsof higher quality. ) entails a number of problems. These occur in particular as the 'energy sector' includes a broad variety of technologies and industries with distinct characteristics regarding the research intensity needed for a patent and the propensity to patent. In addition, the tendency to patent may also differ across countries. As a consequence, the average R&D intensity per patent may differ considerably across technologies. Companies may also decide to classify or label patents in a way that makes it difficult to detect them with the simplified patent assessment applied in this report.
Despite these general constraints regarding the use of patents, they may nevertheless be used as a rough indicator within the scope of this report, given that studies show a strong correlation between the number of energy-related patents granted and the energy R&D investments (Margolis and Kammen, 1999; Johnstone et al., 2008). Popp(2005) shows that patents are a suitable mean for obtaining R&D activity inhighly disaggregated forms. However, it needs to be mentioned that within the present report the patent-based approach could not be exploited in full. For the above-mentioned reasons, this would have required an in-depth assessment of the contents of each patent instead of using a search by keyword and the inclusion of patent applications that were handed in by subsidiaries.
Additional information has been derived from a detailed database of EU-FP6 projects and was used for cross-checking the R&D investment of selected companies in e.g. CCS and H2/FC. From this, the contribution of a company within a certain technological field could be deducted by summing up the company's contribution in all projects within a certain area. These figures were only used for verifying that the results of the other approaches lie above it - if that had not been the case, the estimations would have had to be revisited.
Illustration of the approach
In the following, the methodology applied in this report shall be illustrated. The 'wind energy' sector is a representative example for an area in which the approach allows a relatively accurate analysis.
Table 1 shows a selection of EU-based industries that are considered of outstanding importance in the area of wind energy. For these industries, the R&D investment has been extracted from the EU Industrial R&D Investment Scoreboard and other sources. A first limitation of the analysis is due to the fact that some important companies are missing, such as some large electric utilities that carry out wind energy R&D, even though one may safely assume that much of the innovation is carried out by the component suppliers that are included. A crucial step lies in the breakdown of the R&D budgets and the allocation of contributions to the different technologies, as described in detail for the cases of Acciona, Alstom, Dong Energy, EDF, Iberdrola and Siemens.
Total R&D investment 2007 (€m)
Assumed share of R&D to wind
Assumed R&D investment in wind (€m)
Acciona states that €16.3m of its R&D budget (i.e. 42%) is dedicated to energy R&D. The company is largely active in wind energy and solar (PV and CSP), but also in H2 and bioenergy (solid biomass and biofuels). Given the importance of wind energy, we assume that slightly more than half of the energy-related R&D budget is allocated to wind energy, i.e. €9-10m.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on Acciona's annual and corporate social responsibility reports. Furthermore, through direct contact the assumed R&D break-down could be improved.
Alstom is a major component supplier for power generation and transport. R&D activities in the power sector are carried out through the two sub-sectors "Power Systems" (mainly CCS and gas/steam turbines R&D) and "Power Services". Wind energy would classify under the first. An analysis based on the number of researcher, the split of turnover between the various financial sectors and literature (ZEP, 2008a), leads to an estimated R&D investments within "Power Systems" in the order of €200-220m. Out of this, some share would be dedicated to wind energy. As a rough guess, we assume that some 10% of the 'Power Systems R&D' would go to wind energy research (ca. €20m). Moreover, the Alstom Group acquired 100% of Ecotècnia in 2007, a Spanish wind turbine company. From a newspaper article (El Pais) we know that Ecotecnia will invest around 2% of its turnover in R&D (ca. €5m). In total, we thus assume that some €25m are dedicated to wind energy R&D.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on Alstom's annual and corporate social responsibility reports, their websites, patent analysis, newspaper and expert guesses. Furthermore, through direct contact the assumed R&D break-down could be improved. However, figures are not official figures from the company but remain own estimates.
Mostly active in wind energy, thus 100% of R&D is assumed to go to wind.
EU Industrial R&D Investment Scoreboard
Wind energy is central for Dong Energy. The company states that it currently accounts for the largest proportion of wind energy in Europe and plans massive new investments. However, coal and other renewables also play a key role for Dong Energy: the company is very active in CCS (see e.g. Castor FP6 project) and second generation biofuels (e.g. from straw). Furthermore, we know that Dong Energy intends to invest around DKK350m (€46.8m) in R&D of sustainable energy, including renewables and CCS. On this basis, we roughly assume that 1/5 of Dong Energy's R&D budget is allocated to wind.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on Dong's annual and corporate social responsibility reports and expert estimates.
Out of the total R&D budget, EDF invested around €100m on 'environment', including energy eco-efficiency, research into Renewable Energies, local impact of climate change, and other studies furthering knowledge of environmental issues. Out of this, renewables (excl. hydro) account for 9% according to official data. Based on information of EDF's R&D efforts on PV, we estimate that around €5.5m are dedicated to R&D on wind energy, biomass, geothermal and ocean power. Given the perspectives of wind energy within EDF, we assume that half of that is dedicated to wind energy research.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on EdF's annual and corporate social responsibility reports, their websites and expert guesses.
Furthermore, through direct contact the assumed R&D break-down could be improved. However, figures are not official figures from the company but remain own estimates.
Enercon's turnover was in the order of €2.4bn in 2007. According to their website, Enercon employs around 10000 people and over 130 engineers in R&D. If we assume R&D expenses per R&D employees to be in the typical range of €120000-150000 per R&D staff, Enercon's R&D expenses were in the order of €15.6m to €19.5m spent in R&D, with the central value taken for this report. However, this would mean that Enercon's R&D intensity remains low at less than 1% of the turnover. For this reason, the above estimate may well be an under-estimation.
Mostly active in wind energy, thus 100% of R&D is assumed to go to wind.
EU Industrial R&D Investment Scoreboard
Iberdrola spent €65m in R&D in 2007. According to their website, Iberdrola plans to invest €225m in R&D over the period 2008-2010 (i.e. €75m per year) on the three following activities: 1) Renewables (€70m) and deregulated power business (€70m); 2) Regulated business (€50m); 3) Information technology and other areas (€35m). Considering that on the one hand, wind power plays an important role in Iberdrola's renewables portfolio but on the other hand, Iberdrola is also active in other renewable energy sources (such as solar thermal and biomass) and on grid integration, we assume that 20% of the €70m announced for Renewable Energy research would go to wind energy over the 3 years (i.e. €4.7m per year). We also know that the R&D figures announced for 2008 (and beyond) imply a 15% increase to the 2007 figures. If we thus assume that the 2007 wind energy R&D investment is 15% below the above-estimated figure for 2008, the R&D investment on wind energy in 2007 can be estimated to be some €4m.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on Iberdrola Press Release 24/03/2008 and official reports. Furthermore, through direct contact the assumed R&D break-down could be improved. However, figures are not official figures from the company but remain own estimates.
Mostly active in wind energy, thus 100% of R&D is assumed to go to wind.
EU Industrial R&D Investment Scoreboard
Mostly active in wind energy, thus 100% of R&D is assumed to go to wind.
EU Industrial R&D Investment Scoreboard
For SIEMENS, several approaches have been combined:
Firstly, SIEMENS annual report states that 15% (i.e. €505m) of its total R&D investments are dedicated to energy research. The energy sector is made up of 6 divisions, one of which is the renewable energy division, including wind energy. Knowing the total staff by division and for wind, the relation of total staff to total R&D personnel (ca. 5.5%-9% across the various sectors) and the R&D investments per R&D employee (€132000 per R&D employee in the energy sector), we estimate that R&D investments dedicated to wind research may be some in the range of €16-27m (i.e. 0.5-0.8% of total R&D).
Secondly, a research in patent databases showed that 0.5% (depatisnet) and 1.2% (USPTO) of all SIEMENS patent applications in 2004-2006 were associated with wind energy.
Thirdly, Siemens maintains wind turbine R&D centres in Denmark, Germany, the Netherlands, UK and the USA. For the newly opened USA centre we know that they aim at a staff number of some 50 researchers, but currently only have 12. Assuming a similar number of researchers for all centres would give a total number of employees of 212 people. With an average investment of €132000 per researcher, this would mean an R&D investment of €28m.
From the approaches above, we deduce that the share of R&D dedicated to wind may be in the order of 0.5-0.8%. We use 0.6% as best guess, yet with high uncertainties.
EU Industrial R&D Investment Scoreboard for data on total R&D investment; assumptions based on various inputs, such as annual and corporate social responsibility reports and work done by Naneva and Paschos (2008); press releases; patent application assessment.
In 2007, Vergnet's turnover was €37m out of which 10% were dedicated to R&D activities, meaning that around €3.7m were spent in R&D for the same year. Moreover, Vergnet had around 20 engineers in R&D in 2007 . Assuming that €150000 is spent per R&D employee, on average, this would mean that their R&D expenses account for €3m which is in line with the €3.7m. We assume that most of this investment was allocated to wind energy.
Oseo website (08/2008)
Mostly active in wind energy, thus 100% of R&D is assumed to go to wind.
EU Industrial R&D Investment Scoreboard
Table 1: Example of the approach, illustrated for the wind energy sector
Source: Own analysis. Data on total R&D investments are taken from the EU Industrial R&D Investment Scoreboard (except for Enercon and Vergnet). Assumptions on the share of total R&D investments dedicated to research on wind energy are based on a variety of sources, including companies' annual reports and corporate social responsibility report as well as websites and direct contact where possible.
Note: Data on the R&D investment in wind energy are own estimates and are related to uncertainties. They do not present any official company figure.
BERD (Business enterprise sector's R&D expenditure)
In order to complement the company-based information obtained through the Industrial R&D Investment Scoreboard, the Eurostat/OECD BERD (Business enterprise sector's R&D expenditure) database has been searched for data on corporate R&D investment. BERD contains figures on the business enterprise sector's expenditure in R&D, broken down by different socio-economic objectives following the NACE (Rev 1.1) classification. Furthermore, the expenditures are broken down by sources of funds, disaggregated into business enterprise sector (BES), government sector (GOV), higher education sector (HES), private non-profit sector (PNO) and abroad (ABR).
Within the present report, the energy-related BERD data have been assessed for funds from all sources and those funds that stem from the business enterprise sector BES. The latter would be more comparable to the central approach of this report, which assesses the corporate R&D investments that stem from the companies' funds (to the extent that the publicly funded parts can be identified and subtracted, see footnote 10).
Unfortunately, the NACE classification does not provide the technological breakdown required for this report. For this reason, the BERD results on the corporate energy-related R&D expenditures could only be used for comparison with the aggregate figure of the analysis based on the method described in this chapter. Table 2 shows those sectors that are considered as relevant in the context of energy-related R&D as assessed in this report Note that the NACE classes chosen in the present assessmentcomprise more categories than only classical energy sectors (see the selectionin Kaloudis and Pedersen, 2008). A broader approach was chosen in order to alsocapture companies that are active in the manufacturing of energy components,and thus enhance the comparability with the central approach of this study..
The BERD database often misses data for several EU Member States at the high level of detail that is required for this work, which makes it difficult to obtain a comprehensive picture. In a number of cases, entries by category are not available for a certain country in the last year, but for some years before. These gaps for 2007 were filled with the latest data available back to the year 2003. Even though this approach leads to an error, the mistake made seems to be smaller than assuming no R&D expenditure for those countries where entries are missing.
Funds from all sectors (€ million)
BES funds only
Mining and quarrying
Manufacture of coke, refined petroleum products and nuclear fuel
Manufacture of engines & turbines, except aircraft,
vehicle & cycle engines
Manufacture of electric motors, generators and transformers
Manufacture of electricity distribution and control apparatus
Manufacture of insulated wire and cable
Manufacture of accumulators, primary cells and primary batteries
Manufacture of lighting equipment and electric lamps
Manufacture of electrical equipment n.e.c.
Electricity, gas and water supply
Total Energy-Related BERD
Total EU27 Business and Enterprise R&D expenditure
Share of energy-related over total BERD
Table 2: Business and enterprise R&D expenditures in energy-related fields in 2007 aggregated for EU Member States
Source: BERD (data retrieved in January 2009)
Note: Data gaps in 2007 have been filled with entries for 2003-2006 where necessary. Less data is available for 'funds by BES' than for 'all funds', thus distorting the aggregates.
Research Projects: SRS NET & EEE and ERMINE
Two research projects were used as information sources on industrial energy R&D spending, namely the ERMINE project http://www.ermine.cesiricerca.it/ and the SRS NET & EEE project http://srs.epu.ntua.gr/. IPTS accessed the reports and databases and contacted the project coordinators for clarifications.
The SRS NET & EEE (Scientific Reference System on new energy technologies, energy end-use efficiency and energy RTD) project aimed at enhancing the availability, quality and completeness of data on new energy technologies and energy efficiency and at producing unbiased, validated, organised and scientifically agreed technical and economic information on renewable and efficient technologies. As part of this work, an extensive database on R&D expenditure was constructed.
The database contains public and private energy R&D expenditure up to the year 2005. Private R&D data are taken from several sources such as specific studies, R&D programmes, activity reports, etc. and are available for wind energy, CCS, H2/FC and nuclear (fission). Note that PV is included within the "solar energy" technology.
The approach used is presented in the "R&TD Countries' Reference Report" (SRS project, 2007), which includes the name of the sources and their quality level for every Member State. Unfortunately, data on private R&D spending are only available for a limited number of EU countries. In some cases, this lack of data unfortunately means that countries that are central for a given technological field are not covered.
For public R&D investment, the SRS project primarily used the IEA database in conjunction with other sources such as Eurostat or various studies. See van Beeck et al. (2009) for a more detailed overview of the project results and data gathering procedures.
Given the difficulties encountered in collecting information on private R&D efforts, one of the project's conclusions is that "The data collection on ERTD expenditures in Europe would be facilitated with the appointment of a European institution that ensures a systematic collection of validated and disaggregated data on public and private ERTD expenditures. For a comprehensive database, it is vital that such an institution has enough power and prestige, in particular to urge companies to provide data on private expenditures."
As part of the ERMINE (Electricity Research Road Map in Europe) project, a database on R&D expenditure in the EU electricity sector has been created based on various prime resources and direct contact to companies (de Nigris et al., 2008). A central part of the data collection consisted of questionnaires sent to the main actors of the EU electricity sector (government institutes, universities, research centres, electricity companies, etc.). Public energy R&D budgets are primarily taken from the IEA database (but calculated in euro 2004 which may explain some differences to the results of the SRS database). The ERMINE data have been used for verifying the order of magnitude of the results of the present work to the extent possible.
New Energy Finance
New Energy Finance www.newenergyfinance.com offers an on-line desktop with information on investors and transactions in clean energy. A core part of this service is based on an extensive database containing information on financial transactions, such as private equity and venture capital. Unfortunately, the database does not contain explicit data on R&D expenditure.
However, New Energy Finance recently published an analysis on R&D investments in clean energies as part of the publication 'Global Trends in Sustainable Energy Investment 2008' (Boyle et al., 2008) and as a Research Note (New Energy Finance, 2008). The data are based on the information of 47 companies, mainly extracted from their annual and corporate social responsibility reports. Similarly to the main approach used in this report, the R&D expenditures are allocated to the country that hosts the company's headquarters. To the extent possible, the publicly financed part of the corporate R&D investments is subtracted (similar to the EU Scoreboard).
According to these studies, global R&D spending into renewable energy technologies and energy efficiency has increased slightly from 2006 to 2007 to reach a total of US$16.9 billion. Of this, corporate R&D accounted for US$9.8 billion and public R&D for US$7.1 billion (Boyle et al., 2008). In 2006, the EMEA (Europe, Middle East and Africa) region accounted for US$4.8 billion of corporate research in clean energy technologies in 2006, with the largest part coming from energy and utility companies followed by technology and automobile sectors (New Energy Finance, 2008).
The assessment done by New Energy Finance defines the category 'clean energy' in a much broader way than how this report defines 'SET-Plan priority technologies'. The category 'clean energy' also includes energy efficiency R&D efforts as well as co-generation, but does not include nuclear. Unfortunately, no further breakdown of clean energy research into individual technologies has been available. However, the clean energy investments that were provided for a number of large companies have been used as one input for the allocation of those companies' R&D budgets to individual technologies in the present report and for comparing the order of magnitude of the results of the present report.
The uncertainties related to the various approaches pursued in this report are described above. In the following, suggestions are made on how to further enhance the accuracy of the outcome:
- Firstly, and despite the methodological problems related to linking R&D output to -input indicators (see above with further references), an in-depth analysis of companies' patent applications may provide useful additional - yet not sufficient - information from which to derive an indication of their R&D directions. However, such an assessment would also need to pay attention to systematic problems such as the fact that some sectors are more "patent-intensive" than others and that the specific R&D costs per patent differ between technologies.
- Secondly, a company's current (and announced future) positioning across various business-fields may provide some insights into the areas of research that could be regarded as strategic. The Compustat database, a database of 88000 companies worldwide maintained by Standard and Poor's, may be of use here.
- Finally and probably most importantly, broadening the number of companies directly surveyed on a systematic basis would refine the results and help to validate the assumptions made.
Member States' public R&D investments
Unlike the estimation of corporate R&D investments in energy technologies, the assessment of public energy R&D investments in Member States is primarily based on available supranational datasets, which has been enhanced by the use of some national data which result from the direct contact with EU Member States. The following two datasets have been used as the principal starting points: the Eurostat GBAORD and the IEA RD&D statistics. They will be described in more detail in the following.
The Eurostat GERD (Gross Domestic Expenditure on R&D) database GERD is maintained by Eurostat/OECD on the basis of datacollected from all R&D performers. It has a sectoral breakdown (BES:business and enterprise, GOV: government, HES: higher education; PNP: privatenon-profit). , which contains R&D expenditure by R&D performers, could not be used as its breakdown does not provide the level of detail on different energy technologies required for this report.
GBAORD (Government Budget Appropriations or Outlays on R&D)
Government Budget Appropriations or Outlays on R&D are all appropriations allocated to R&D in central government or federal budgets. It is also recommended that provincial or state government should be included when its contribution is significant, while local government funds should be excluded (OECD, 2002). Data are collected from government R&D funders and maintained by Eurostat and the OECD and follows the NABS (Nomenclature for the Analysis and Comparison of Scientific Programmes and Budgets) classification.
Socio-economic objective 5 'Production, distribution and rational utilisation of energy' is the most relevant main category for the present report. It covers research into the production, storage, transportation, distribution and rational use of all forms of energy. It also includes research on processes designed to increase the efficiency of energy production and distribution, and the study of energy conservation.
For this report, data for 2007 have been used mostly. Due to major data gaps that would have hampered a comprehensive analysis at the EU-27 level, a simple gap filling procedure has been applied using data from the latest available years back to the year 2003. This is noted in detail in the notes below the figures presenting the results of this analysis.
The 4-digit level of detail provided by the GBAORD database can be considered as accurate enough for the objectives of this report (see Table 3). Unfortunately, disaggregated data on GBAORD energy subgroups were only available for the Czech Republic, Germany, Ireland, Greece, Spain, Malta, the Netherlands, Romania, Slovenia and the UK In the case of IE and MT the reported values are zero.. This means that for some of the major energy R&D funding Member States such as France or Italy no data is available at this disaggregated level, inhibiting an aggregated figure on the EU-27 Member States' public R&D investment by technology (see Figure 6).
Subsectors of Production, distribution and rational utilisation of energy" (category 5)
Term used in this analysis
General research on production, distribution, and rational utilization of energy
General research on energy
Fossil fuels and their derivatives
Radioactive waste management including decommissioning with regard to fuel/energy
Renewable energy sources
Solar thermal and photovoltaic energy
Water, wind and wave energy
Water, wind & wave
Research into biomass conversion (particularly into the areas of pyrolysis, gasification, extraction and enzyme processing); research on the processing of waste from industry, agriculture and the domestic sector with a view to energy
Rational utilization of energy
Other research on production, distribution and rational utilization of energy
Table 3: Classification of energy-relevant sectors in GBAORD
- Figure 6: Distribution of energy-related R&D budgets of EU Member States, 2007
Source: GBAORD (data retrieved from Eurostat in January 2009)
Note: No data for BG. Gap-filling was applied by using 2006 data for CZ, DE, ES (all sectors) and UK (500, 501, 503, 505, 506); 2005 data for NL (501) and HU (category 5); 2004 data for SI (all sectors), UK (502, 509) and NL (509).
IEA RD&D statistics
The International Energy Agency (IEA) hosts a publicly accessible database on energy RD&D budgets from the IEA member countries. Data is collected from government RD&D funders.
Unlike the GBAORD, the IEA database covers demonstration activities on top of pure research and development activities. 'Demonstration projects' are of large scale, but are not expected to operate on a commercial basis (IEA, 2008). In practice, however, most IEA member countries do either not provide data on funds directed towards demonstration, or do not display them separately. As a consequence, the aggregated demonstration activities of EU Member States that are explicitly specified us such in the IEA database amount to some 9% of total R&D activities of these countries in 2007 (gap-filled), yet with important differences across technologies as shown in Figure 3 (see also section "Scope of the report" for a further clarification of the scope covered by the present assessment). For the purpose of this report, we thus consider the IEA data as mainly related to R&D investments.
The breakdown of the IEA R&D data follows a scientific/technical nomenclatureSee also European Commission (2005) for a comparison ofenergy R&D statistics in the European Union.. The level of detail more or less matches the requirements of this report. Furthermore, most countries provided data at this high level of detail, which renders the IEA database central for this work.
Only 19 of the 27 EU Member States are IEA members. Consequently, the database systematically contains no data for the other countries, i.e. for: Bulgaria, Cyprus, Estonia, Latvia, Lithuania, Malta, Romania, and Slovenia. Nevertheless, the aggregated R&D budgets of the Member States covered by the IEA database account for almost 99% of the overall EU-27 energy budget according to GBAORD data, thus limiting the errors incurred by a lack of data in the missing EU Member States, notwithstanding that the contributions of the excluded Member States may be higher for individual technologies.
The latest available data are for the year 2007. Similar to the procedure applied for GBAORD data, some straight forward 'gap filling' process was applied for the IEA data. For entries missing for 2007, the value from the latest available year was applied down to the year 2003; data older than 2003 were not considered. This means for instance that IEA data for Greece (latest available year at detailed level: 2002), and Luxembourg (2000) were neglected. In the case of Finland and the Netherlands, mainly 2006 data have been used. For the Czech Republic and the Slovak Republic 2003 data were used. This approach slightly distorts the overall picture (see chapter "Analysis of uncertainties"), but is nevertheless justified given that the main interest of this report lies on the aggregated EU figures.
As a result of the consultation process with Member States, it was decided to use national figures for the year 2007 in the case of France, the UK, Germany, Austria and Belgium Note that in Belgium,nuclear-related R&D (fusion and fission) is under federal responsibilitynamely through the DG Energy by the FPS Economy, while administration ofnon-nuclear-related R&D activities is the main responsibility of theregional governments. The non-nuclear energy R&D figures thus are theaggregate of the regional funds. , either replacing the IEA figures or complementing them. In the case of Ireland, it was decided to use official national figures for the year 2006 as proxy for the R&D investments in the year 2007 instead of using the data from the IEA RD&D statistics.
In addition to the 2007 figures, an average of the national R&D investments over the time span 2002-2007 is displayed in most of the charts shown in the report. This supplementary information prevents the risk of giving too much weight to data mavericks and one-off events while at the same time providing a longer term perspective.
EU FP6 public energy R&D investments
European funds complement the Member States' public R&D support. The Research Framework Programme and EURATOM Framework Programme are the key source of R&D financing on energy technologies. Other EU funding schemes such as the Competitiveness and Innovation Programme with its pillar Intelligent Energy Europe, the Cohesion funds, Trans-European Networks, etc. could either not be assessed quantitatively on the level of detail needed for this report, or were considered less relevant for research as they mainly focus on deployment. They are mentioned in box 1.
For the purpose of this report, the expenditures The assessment is based on commitments, not payments. of the 6th Research Framework Programme and the EURATOM Framework Programme 2002-2006 have been analysed. As FP6 ended in 2006, detailed data on R&D expenditures are available on a project level, which allows a decent allocation of R&D spending to the various technology types. Such detailed data is not yet available for FP7, thus preventing the use of FP7-figures for an in-depth assessment. Nevertheless, some detail is provided on the evolution of energy-related R&D investments from FP6 to FP7 to allow for an assessment of future trends. The reflections on FP7 are based on the budget decisions taken so far.
The breakdown of the FP6 investments has not been based on the budget lines (such as 'Sustainable Energy Systems'), as this would not provide the level of detail required by the present work. Instead, R&D expenditures have been allocated on the basis of individual projects to the various SET-Plan priority technologies, which allowed estimating the total EU-funded expenditures for various technologies under FP6. The assessment systematically includes all projects funded within of the core budget line used for energy-R&D projects (Sustainable Energy Systems); to the extent possible it has been complemented by other energy-relevant projects that are funded through other budget lines (e.g. 'sustainable surface transport' or 'horizontal research activities involving SMEs') based on various publications (European Commission, 2004; 2007b-e).
This allocation approach is associated with (limited) uncertainty as some projects simultaneously address various technologies (e.g. projects on alternative motor fuels comprise work on biofuels, H2 and natural gas; projects on integrating fluctuating renewable electricity sources may be allocated to renewables or to grids). In those cases, a decision was taken on to which technology the expenditures should be allocated. Even though this process creates a potential source of error (see section "Uncertainties associated with estimates for public energy R&D investments" in chapter "Analysis of uncertainties") it is justified as it avoids a double-counting of the budget of one single project. In future work, one may go a step further and allocate fractions of the project budget to individual technologies; this would nevertheless require an in-depth knowledge of the contents of each project. Despite the uncertainties related to the present approach, a comparison with other sources (Rossetti di Valdalbero et al., 2007; European Commission, 2007b-e; Langlois D'Estainot, 2009; EPIA, 2007; Orion Innovations, 2008; Filiou et al., 2009) does reveal only limited differences.
The analysis of EURATOM funds is based on budgets, not on actual expenditures. Furthermore, not only project-related funds have been included but also the JRC funds dedicated to nuclear energy.
As the EU Research Framework Programmes are of multiannual nature, while the present report aims at presenting the EU R&D investments for the most recent year available, they had to be broken down further in order to determine the specific budgets available for one single year. In order to level out annual fluctuations in the budget that are due to the project cycles, it was decided to assume an even allocation of the total expenses to every year of the FP6 duration (the financially effective duration of FP6 was four years). The figures shown for the EU public R&D investment thus relate to an average annual investment over the years of the duration of FP6.