Customers' Requirements of Solar Energy Resource Information
Results of the IEA SHC Task 36 on-line survey
Thomas Huld1, Marcel Súri1, Richard Meyer2, Richard Perez3, Lucien Wald4, David Renne5, Paul W. Stackhouse6, Dominique Dumortier7, Marion Schroedter-Homscheidt8, Jesus Polo9, Christoph Schillings10, Jan Remund11, Elke Lorenz12, Pierre Ineichen13, Hans Georg Beyer14 European Commission, Joint Research Centre, 21020 Ispra, Italy SunTechnics GmbH, Anckelmannsplatz 1, 20537 Hamburg, Germany 3 University at Albany, ASRC, 251 Fuller Road, Albany, NY 12203, USA 4 Ecole des Mines de Paris/Armines, BP 207, 06904 Sophia Antipolis cedex, France 5 National Renewable Energy Laboratory, 1617 Cole Boulevard, 80401-3393 Golden, CO, USA 6 NASA Langley Research Center, 21 Langley Blvd. ,M.S. 420, Hampton, VA 23681, USA 7 LASH ENTPE, rue Maurice Audin, 69518 Vaulx-en-Velin, France 8 DLR, German Remote Sensing Data Center, Postfach 1116, 82234 Wessling, Germany 9 CIEMAT, Avd. Complutense, 22 28040 Madrid, Spain 10 DLR, Institut für Technische Thermodynamik, Pfaffenwaldring 38, 70569 Stuttgart, Germany 11 Meteotest, Fabrikstrasse, 3012 Bern, Switzerland 12 Carl von Ossietzky University Oldenburg, Institute of Physics, 26111 Oldenburg, Germany 13 University of Geneva, Battelle Bat A, 7 rte de Drize, 1227 Carouge, Switzerland 14 Hochschule Magdeburg-Stendal (FH), Breitscheidstrasse 2, 39114 Magdeburg, Germany
2 1
Background
Task 36 of the International Energy Agency (IEA) Solar Heating and Cooling Programme (SHC) concerns the quality and availability of solar radiation data for solar energy applications. A part of the work planned for Task 36 was to conduct a survey of the needs in solar radiation data by users. These needs are defined in terms of types of data as well as quality and availability. Purpose of this study is to identify where currently the greatest demands for improvement in the field of solar resources are.
Methods
It was decided from the start that the survey should be conducted by an on-line questionnaire, rather than by e-mail or mailed questionnaires. In this way it would be easy to perform the analysis of the survey results since the data would already be in a database. Furthermore, the survey would then be accessible to potential users not previously known to the Task participants. After a series of drafts the final version of the questionnaire was then implemented as a web application by the JRC and hosted by the same organization. It was made available in four languages: English, German, French and Italian. Data entered into the questionnaire were automatically inserted into a PostgreSQL database. The survey was announced on a number of web sites run by the Task members, and was on-line from June 2006 to February 2007. The entire web page as seen by users is reproduced here as Annex 1. 1
The questions in the questionnaire were divided into a number of topics: 1. Personal data (optional, the survey could be taken anonymously) 2. Profession 3. Technology: for which type of technology does the user need solar radiation data 4. Purpose: what does the user need radiation data for? 5. Type of data (which parameters?) 6. Temporal resolution 7. Timeliness, how recent should data be? 8. Synthetic data of interest 9. Spatial extent of data (single sites, maps) 10. Spatial resolution of gridded data 11. Use of data (how to use them, not for what) 12. Satisfaction with currently available solar radiation data 13. Radiation forecasts 14. Climate change 15. Additional comments For most of these categories the possible answers to questions were in the form of a number between 0 and 5 indicating the interest/importance the respondent would assign to the given topic. For a number of questions it was also possible to enter text and comments. For the answers where a number was requested, the default answer was always 0. This means that it was not possible to distinguish whether the respondent answered that a given option was not important, or whether (s)he did not answer at all. For this reason the number of valid replies in each of the categories with this type of answer has been calculated as the number of respondents who gave at least one non-zero answer in that category. The survey web site was announced in the following web sites: · · · · · PVGIS European Commission, Joint Research Centre (http://re.jrc.ec.europa.eu/pvgis/) SoDa Ecole des Mines de Paris (http://www.soda-is.com/) Satel-light ENTPE (http://www.satellight.com/core.htm) IEA SolarPACES (http://www.solarpaces.org/Tasks/Task5/task_v.htm) Meteotest (http://www.meteotest.ch)
Results
Removing empty replies, the total number of respondents was 111.
1.
Nationality
As part of the optional personal data the respondent could select his/her country of residence. A total of 79 respondents gave this information. Respondents from European countries account for 84% of the 79 replies and this preponderance of replies will be discussed below in "Discussion and Conclusions".
2
Country Spain Germany Italy France Switzerland Greece Belgium Portugal United Kingdom United States Czech Republic Pakistan Morocco Mexico Gambia Albania Canada Netherlands Croatia Japan Djibouti Panama Iran Luxembourg Thailand Tunisia
Number of answers 16 8 7 7 6 6 5 3 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Table 1
2.
Profession of respondents
A total of 111 respondents chose at least one of these options, out of which 36 chose two and 12 chose three options. Profession engineering company consultant university other (please specify): manufacturer utility company public research laboratory governmental, public agency / service private research laboratory non-governmental organization or other promoter of solar energy Table 2 Among those that chose more than one profession the most common combinations were consultant + engineering company or manufacturer + engineering company, or a combination of university and public research or governmental service, probably reflecting the fact that most European universities are state-run or closely involved with state-run organizations. 3 Number 31 20 15 15 14 7 7 6 6 6 Percent 28 18 14 14 13 6 6 6 6 6
15 respondents chose the "other" option. Five identified themselves as private persons/individuals, three as students, and two as systems distributors/wholesalers. In addition, there was one architect, one telecom-industry person, one system integrator, one project developer, and one person setting up a company. Overall, the answers are dominated by engineering companies and consultants, as well as universities. There is quite a bit of interest also from manufacturers, though the question did not ask for details of what is manufactured. Utility companies and governmental organizations outside research are not so well represented.
3.
Technology
The question: "Which technology are you concerned with?" For each category the respondent could give a number from 0 (not important) to 5 (very important). A total of 96 valid replies were received. A summary in the Tab. 3 shows the average number of points given to each of the categories, and the number of high values (4 or 5) showing a high level of interest. Technology photovoltaics (PV): solar heating: concentrating (thermal) solar power: building engineering, architecture: solar cooling: concentrator PV: water desalination: chemical systems: Table 3 It is clearly seen that the highest level of interest is in the fields of PV and solar heating systems, whereas in particular chemical systems and water desalination are not of concern to the respondents. Concentrating solar energy systems (both PV and thermal) and building engineering have an intermediate level of interest. The choice "other" had 8 replies. Two gave the field of interest as PV in buildings (obviously not considering this to be part of the "building/architecture" option). Two were interested in agriculture (one specifically in greenhouses). Two were in the field of electrical energy production. Finally there was one reply each for UV radiation protection and for "environmental research" (not detailed). Average points 2.98 2.23 1.72 1.72 1.38 1.18 0.49 0.20 Answer >3 60 40 31 34 18 21 6 2
4.
Purpose
The question: "For which purpose do you use solar radiation?" For each category the respondent could give a number from 0 (not important) to 5 (very important). 97 valid replies were received. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5).
4
Purpose system design: feasibility study: investment decision: cost assessment: site selection: research/education/promotion: monitoring: plant operation: guarantee/certification/insurance: deployment: energy trading: grid operation: fault detection: plant maintenance: policy-making: plant decommissioning: Table 4
Average points 3.10 2.97 2.47 2.42 2.31 2.06 1.40 1.27 0.96 0.94 0.79 0.78 0.73 0.71 0.56 0.26
Answer >3 65 60 52 46 45 39 24 24 17 19 13 13 12 11 8 1
Here the most popular answers are system design, followed by other aspects of planning of systems (presumably PV though it could also be other solar technologies). Research and education also receives a fairly high number of replies. On the other hand, decommissioning seems not to need solar radiation data (this actually makes sense). Policy making, energy trading and plant maintenance are also low on the list. Somewhat surprising is that there is also not much interest in solar radiation data for plant monitoring, even though an obvious way to check for performance problems is to compare actual power production with expected power production calculated from the irradiation. This may be partly explained by the novelty of this approach which is not yet well known. On the other hand, it may be explained by the fact that only a few companies are mastering this technology and that the respondents, which are companies, do not consider this market segment as they do not want to enter it. Five respondents chose the "other" category, giving purposes such as "research", radiative models, business plan creation and "photosynthesis" (biomass??)
5.
Type of data
The question: "Which type of data are you interested in?" The answers were given as a number from 0 to 5, indicating level of interest. A total of 95 respondents had at least one non-zero answer. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5).
5
Type of data global horizontal radiation: direct radiation: diffuse radiation: radiation on tilted and tracking surfaces: daylight (illuminance, luminance): spectral distribution of irradiance: ambient temperature: snow cover: wind speed / wind direction: relative humidity / dew point: atmospheric pressure: Table 5
Average points 3.32 3.66 2.83 2.83 1.91 1.60 3.03 1.02 2.33 1.49 1.17
Answer >3 69 75 52 59 35 27 59 16 46 27 20
The most popular choices here are direct radiation, closely followed by global horizontal radiation, ambient temperature and inclined plane radiation. The choices of least interest are snow cover and pressure. Also the solar radiation spectrum is not interesting to so many of the respondents (though the number of respondents who rated this as very interesting (>3) is still more than ¼ of the total number of respondents). Four chose "other", with the fields of interest given as tilted plane radiation (though this was in the list of choices), "isolation" (insolation maybe?), aerosol content and visibility, and water temperature (possibly for solar water heating estimates).
6.
Temporal resolution
The question: "For the type of data most important to you, which type of temporal resolution do you need?". 94 valid replies were received. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). Temporal resolution annual averages: monthly averages: weekly averages: daily averages: hourly averages: 15 minute averages: 5 minute averages: instantaneous values: Table 6 Here we see a general tendency that the long-term averages are of greater interest than the very shortterm averages (less than 1 hour), with the most popular being monthly averages. One "other" reply was received, asking for monthly averages on an hourly basis. Average points 2.69 3.19 1.61 2.06 2.39 1.22 0.84 0.77 Answer >3 53 67 31 38 48 21 14 12
6
7.
Timeliness
The question: "For your applications, how recent should the data be?" There were 92 valid replies. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). Timeliness very recent (e.g. last hours/days): recent (e.g. last month/last year): from older archives: Table 7 Results show a moderate amount of interest from the users, with the least interest shown in the very recent data. Interestingly, the recent data is more interesting than the older data, even though the replies to the use of data indicated that people were interested mostly for planning/siting of solar energy systems, for which long-term data might be thought the most relevant. Average points 1.79 3.06 2.83 Answer >3 29 64 60
8.
Synthetic data
The question: "If you are using synthetic data, what type?" The number of valid answers was a little lower than for the previous sections, at 82. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). Synthetic data type Typical meteorological years: Design reference years: Test reference years: Stochastically generated typical years: Table 8 The clear winner here is the typical meteorological year, which seems to be used much more than the alternatives. Average points 3.06 1.83 1.44 1.05 Answer >3 59 37 27 18
9.
Type of spatial data
The question: "For the data you are using the most, are you interested in sites (locations), maps" 92 valid answers were recorded. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). Type of spatial data A single site: Several sites, but less than 10: More than 10 sites: A grid (map) of point values covering a country/region: A grid (map) of average values covering a country/region: Table 9 Average points Answer >3 1.65 33 1.95 39 1.60 28 2.16 43 2.62 54
7
In this case the spread of values is not very large. There is a general preference for gridded data rather then single sites, but both types of data seem to be useful to the respondents.
10. Spatial resolution
The question: "What spatial resolution of maps (what pixel size) do you think is needed for your main purpose?" 88 valid replies. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). Spatial resolution Values or time-series averaged over regions/countries: (approx. 300 km x 300 km, global coverage): (approx. 100 km x 100 km, global/regional coverage): (approx. 10 km x 10 km): 5 km x 5 km: 1 km x 1 km: Table 10 There is not very much variation between the options, though the very coarse resolutions are less popular. Generally it seems that for maps, 5x5 or 10x10km are sufficient for most people, although about half the respondents indicate a high level of interest in the detailed maps with resolution 1x1km. Average points Answer >3 1.21 19 1.03 12 1.95 35 2.41 48 1.97 44 2.12 42
11. Use of data
The question "Use of data. When accessing either location-specific or mapped observations do you:..." This part produced 89 valid replies. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5). When accessing either location-specific or mapped observations do you: Average points Answer >3 Use observations themselves for documentation or customized purposes: 2.17 42 Use observations as inputs to a simulator/software: 2.61 54 Use time-averaged observations for documentation / custom applications: 1.80 32 Use statistical summaries other than averages (e.g., percentiles, probability distribution, extremes), for documentation / custom applications: 1.40 23 Use time-averages or statistical summaries as inputs to a simulator/software: 1 .86 39 Table 11 The most widely reported use of the data is as input to simulations and other software. Documentation had a lower priority. However, there seems to be a rather broad range of data applications among the respondents.
12. Data quality
The question: "Is the solar resource and meteorological information currently available satisfactory when considering..." 8
83 respondents gave non-zero answers to at least one of the above questions which needed a numerical value (all except the last question). The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5) and low score (0 or 1) Is the solar resource and meteorological information currently available satisfactory when considering... Access: Up-to-date: Geographical coverage: Accuracy : Clarity in description of products: As a whole: Table 12 Since these questions not only asked about the needs of users but also about their view of the quality of existing data we have also included the counts of the low scores. From the average number of points and how they are distributed among the different questions, there is not a large difference in how the quality of data is seen by the users. However, one thing is clear from this part of the survey: the number of people who are satisfied with the data is significantly lower than the number of people who are strongly dissatisfied. Obviously, there is room for improvement. The last question in this part asked about where users currently go to get solar and meteorological data. Here there were a large number of answers. These are reproduced in their entirety in Appendix 2. Average points Answer >3 2.05 2.29 2.35 2.13 2.06 2.02 29 36 38 32 35 28 Answer <2 51 45 43 47 50 47
13. Are radiation forecasts important to you?
For this section there were only 63 valid replies. The results are given as average points awarded to each category, as well as the number of answers with a high score (4 or 5) and low score (0 or 1). Are radiation forecasts important to you? Nowcasting (up to 6 hours): Forecast up to 24 hours: Forecast up to 3 days: Forecast up to 1 week: Forecast up to 1 month: Seasonal forecast (2-6 months): Table 13 The general level of replies is low, with an increase towards longer forecasts. It seems that at present there is still not much interest in forecasting of solar radiation among the types of users represented in this survey. Of course, this could just as well reflect the lack of representation of the groups to whom forecasts would be interesting. Average points 0.79 1.1 0.8 0.87 1.12 1.51 Answer >3 14 23 12 8 21 25
14. Climate change
The question: "Is climate change an issue for your activities?" The following three questions could be answered only by yes or no, or not at all.
9
Is climate change an issue for your activities? Do you expect that regional climate change could influence the feasibility of solar applications of concern to you? Do you expect that regional climate change could influence their operational efficacy during the expected operation period? Would you like to get better information on how solar installations help to reduce greenhouse gases and what this means for our environment? If yes, what kind of information do you request? Table 14
Yes 65 58 44
No 28 34 42
For the next two questions, the respondent was required to enter a number giving the period in years: · How many years is the typical amortization period of systems you are dealing with? · How many years do you expect that your solar systems are running? Here we have divided the replies into groups depending on the number of years in the answer. Number of replies Interval (years) Amortization time Lifetime >0, <=5 15 7 >5, <=10 32 5 >10, <=20 18 25 Table 15 >20, <=30 9 27 >30, <=40 0 9 >40 1 2
Discussion
Problems with the method
A number of issues have become apparent after the survey ended and the analysis started. The total number of people responding is 111 which is quite poor for a global survey, in a field of rising economic interest. Because of the low sample size the results have a relatively high uncertainty. Nevertheless it is high enough to derive some clear tendencies for questions, where there are clear tendencies in the answers. The survey was announced on a few web sites among the members of the Task. However, this was not done immediately for all the sites. Furthermore, some of the sites have more visitors than others. The end result is that the respondents tend to be selected according to the following non-random criteria: Users of the Web services offered on the web pages that advertised the survey, e.g. PVGIS, SoDa, etc. are overrepresented. For instance, PVGIS provides services mainly for the PV community, so this community may be overrepresented relative to other potential users which do not use these services. The most popular sites would be expected to generate more replies than the less frequented ones. 10
Most of the sites which advertised the survey were based in Europe and offered data or services mainly for Europe. The geographical distortion is clearly seen in the results. 83% of respondents are from Europe. But this can also reflect that Europe is the continent, where solar energy currently receives its strongest development. The observed regional distortion should only matter with questions, where regional aspects dominate. For the many questions not covering regional aspects (interest in synthetic data, timeliness etc.) it can perhaps be assumed that the needs of users elsewhere are not significantly different from those of European users of solar radiation data. The relative interest in the different technologies could be significantly different elsewhere (such as cooling, desalination, and concentrating solar thermal or PV). It would have been possible to include in the database information about the referring page from which the survey web page was accessed. Unfortunately, this was not done, so there are no hard data on where the respondents found the survey (this was also not one of the questions). Another weakness of the way the survey web application was programmed was that each numerical answer (rating questions from 0 to 5) had a default value equal to 0. This means that it is not possible to be sure if a user deliberately gave a vote of 0 to a question or if (s)he just did not answer at all. We have tried to exclude the "no answer" by looking at all the answers of a given respondent to the questions in each section of the survey. If the respondent gave a 0 value to ALL questions, it is assumed that the questions in that section were not answered by that respondent. The many sections with 0 votes in the later part of the questionnaire suggest that the total length might have been too long. Due to lack of time and motivation many solar resource data users did not start filling in or quit before completing the questionnaire.
Main results of the survey
Despite the problems mentioned above, there are still some clear results left. The distribution of respondents among the different professions corresponds to the initial target of the study. The main classes of respondents are engineering and consultancy, and the main fields of interest are in PV and solar heating systems, both of which are (up to now) dominated by rather small installations. The main purpose for solar radiation data was given as "system design" and "feasibility studies". Utilities are not well represented in the survey, and this is reflected in the lack of interest in certain types of data which would be of particular interest to utilities, as will be discussed below. The choices of type of technology also reflect the current status of commercial development of the given technology. PV and solar thermal heating are prominent, with concentrating systems less pronounced, and technologies such as solar cooling and water desalination not of great interest. The choice of purpose for the data follows a logical pattern: solar radiation data are needed especially before a solar energy system is built, to show the feasibility and to optimise design and site selection. We see a strong response for choices like site selection and investment decisions, but not enough for maintenance and grid operation. Somewhat surprising, the interest in using radiation data for monitoring is not very strong. This may be because there are still only rather few large installations which often run their own monitoring. There are companies who specialize in providing monitoring for small installations, but they probably have their own sources of radiation data and may not be answering surveys like this one. The answers for the type of radiation data needed contain surprise: direct radiation is the type most interesting to the respondents. This may reflect the fact that it is still difficult to find such data, but also that tracking PV systems and concentrating solar thermal systems are rising technologies, which have high demand in such information. Here there is a clear indication of where future efforts of the community should go. It is also clear that the most needed ancillary data are the ambient temperatures, which are much more 11
sought after than pressure, humidity or spectral distribution of the radiation. Daylight does not score very high, again reflecting the distribution of respondents, who are primarily from the solar energy field. The responses to the query about temporal resolution show clearly that users are mostly interested in long-term averages, with the most popular choices being monthly averages and yearly averages. Very short term averages or instantaneous values are not of interest to most people. The small attendance of people involved in solar thermal power plants might be a reason, because proper simulation of such systems would require much higher time resolution. Here we also see the lack of participation of the utilities who might be interested in high-resolution instantaneous data for statistical analysis of power output. This is also shown in the question about timeliness where the most recent data are less requested than archived data. When it comes to spatial resolution of data, the results clearly show that respondents are more interested in gridded data than in single sites, and that the grid resolution should be reasonably high, with the strongest request for a resolution of 10 km x 10 km. This again fits in well with the profile of respondents who mainly need the data for siting and system design. The replies to the question about quality/adequacy of the data do not show differences among the various aspects. All aspects reach the same average score, ranging from 2.0 to 2.4; which is fairly low. It is surprising that none of the aspects reach a higher note on average. For example, another survey, made for the SoDa Service, indicates that access is more than satisfactory and partly conflicts with this survey. We believe that this should be investigated in more details, especially since it may impact on the design of the graphical user interface of the prototype of the information system. Beside this point, one thing is clear from this part of the survey: the number of persons satisfied with the data is significantly lower than the number of persons dissatisfied. Obviously, there is room for improvement. Forecasts did not show a strong response. The main users for such data would be utilities and energy traders. Utilities are not strongly represented in the survey and energy trading seems to be almost absent. Of course, this is more a reflection of the current low penetration of solar electricity in the overall electricity market than of the intrinsic usefulness of forecasts. The most requested type of forecast currently is the long-term forecast aiming to predict on seasonal time-scales. This sort of forecast is the one which will have the highest uncertainty. It might have raised the respondents' interest, because the feasibility of getting such information is not widely known. Therefore seasonal forecasting might attract more attention. Again this may be connected with the lack of respondents from utilities that would be seriously interested in short-term forecasts. The answers concerning climate change issues indicate that many respondents are of the opinion that climate change could significantly affect conditions controlling the feasibility of solar energy systems. This shows the importance of further studies on the influence of climate change on solar irradiation and the importance of using relatively recent data or at least to validate that older data still reflect the current situation.
Variation in user needs depending on users' technological interests.
The above analysis is limited to looking only at the complete survey to determine what users require most. However, as the users have varying uses for the solar radiation data it would be reasonable to expect that there would be differences in requirements depending on what field the users work in. For this analysis we chose the categories Question 3: "Which technology are you concerned with?", that had received the highest average scores: · · · photovoltaics (PV) solar heating (SH) Concentrating (thermal) solar power (CSP) 12
For each of these three categories we chose the respondents that had indicated a high level of interest in these topics (a score >=4) and looked at the distribution of answers to questions 5 to 10.For each of the categories the number of respondents concerned are given in Table 3. For question 5: "Which type of data are you interested in?" the most popular answers are shown in Table 16, along with the percentage of high answers. PV global horizontal radiation (82%) direct radiation (73%) radiation on tilted and tracking surfaces (65%) SH direct radiation (75%) global horizontal radiation (68%) radiation on tilted and tracking surfaces (65%) Table 16: CSP direct radiation (87%) global horizontal radiation (55%) radiation on tilted and tracking surfaces (55%)
Here there is a clear difference between the PV community and the SHC respondents, with the latter significantly more interested in the direct radiation component. Solar heating users are intermediate in that there is a weak preponderance of interest in direct radiation. For question 6: "For the type of data most important to you, which type of temporal resolution do you need?", the equivalent results are shown in Table 17. PV Monthly averages (72%) Annual averages (62%) Hourly averages (35%) SH Monthly averages (78%) Hourly averages (55%) Annual averages (48%) Table 17 CSP Monthly averages (68%) Annual averages (61%) Hourly averages (52%)
As for the overall results shown in Table 6, the monthly averages are the most popular with all three user groups. However, for Solar Heating the hourly averages are more important than annual averages, though the effect may not be significant. One possible explanation is that the detailed calculation of solar heating systems depends on knowing the short-term irradiation values, whereas the overall output of grid-connected PV systems is rather independent of exactly when the radiation energy arrives. For question 7: "For your applications, how recent should the data be?", there was no clear deviation of the three user groups from the overall results shown in Table 7, with recent data being the most popular followed closely by older archived data and with relative little interest in very recent data. Only the solar heating users were somewhat more interested in very recent data, with 16 out of 40 (40%) respondents choosing this option, compared to 29 out of 92 (32%). Question 8 was: "If you are using synthetic data, what type?", and as shown in Table 8 it did not generate a high level of enthusiasm. For the three user groups there are not really significant changes compared to the overall result, with Typical Meteorological Years being the most popular. Again here, the solar heating people deviate a little, with a higher proportion of users choosing Test Reference Years (22 out of 40, compared to an overall result of 37 out of 82). Again for question 9, "For the data you are using the most, are you interested in sites (locations), maps", the replies of the three user groups follow the overall trend rather closely, with the two grid options being the most popular and no obvious deviations Results from question 10: "What spatial resolution of maps (what pixel size) do you think is needed for your main purpose?" are shown in Table 18 below.
13
PV Approx. 10 km × 10 km (55%) (approx. 100 km × 100 km, global/regional coverage) (47%) 5 km × 5 km (47%)
SH approx. 10 km × 10 km (48%) (approx. 100 km × 100 km, global/regional coverage) (43%) 1 km × 1 km (40%) Table 18
CSP 5 km × 5 km (58%) 1 km × 1 km (55%) approx. 10 km × 10 km (52%)
In this case we see a significant difference between the 3 user groups. While the PV and the solar heating users are content with 10 km ×10 km resolution, a large part of the SHC users require a better spatial resolution, 5 km × 5 km or even 1 km ×1 km. In addition, the coarser resolution answers got very few replies from the users interested in SHC (less than 10 answers out of 31 users). This seems to indicate that estimates of solar thermal power requires a higher spatial resolution, possibly because these systems depend on the level of direct normal irradiation, which is sensitive to terrain shadows and local climatic conditions. The overall picture of this analysis is that the requirements of users vary significantly between user communities, and that it is important for data providers to know their customers.
14