Final Report Summary - ENDORSE (ENDORSE (ENergy DOwnstReam SErvices) - Providing energy components for GMES) Executive Summary:The ENDORSE project (ENergy DOwnstReam SErvices) has developed ten services that exploit Earth observation data and models to providing solutions at a regional scale to authorities and companies in renewable energies. The ENDORSE Consortium comprised Armines, Transvalor, and ENTPE in France, the aerospace agency DLR and Hochschule Ulm in Germany, iCons, Flyby and University of Genova in Italy, 3E in Belgium, and the Joint Research Center (JRC).The accurate assessment of the annual energy output of a planned onshore wind farm is an integral part of decision making, investment assessments, and sizing of the plant. By combining the various data required in the decision making process in an automated manner, the company 3E provides qualified estimates of wind resources with an important gain in productivity.Initial phases in the development of solar power plants request a quick assessment of the available resources, of location of potential sites and of electrical yield. The company Transvalor produces maps of solar resources at a spatial resolution of 200 m. They may be ingested into a Web geographical information system dedicated to the identification of possible sites made by DLR and JRC. Other web-based tools and data sets provided by Transvalor and the company Flyby help companies to designing solar power plants, simulating and monitoring their yield. An efficient system has been established by DLR to forecast everyday how much electricity taken from the sun can be sold in the market in southern Spain. Forecasts at regional level by DLR and Hochschule Ulm predict how much electricity will be gained from photovoltaic cells installed on houses in order to ensure stability and quality in the electricity distribution grid. By computing the annual energy consumption used for lighting a room defined by the user, the engineering school ENTPE provides architects with estimates of energy savings that can be gained from using sunlight to power the operation of blinds and artificial light, as well as with other information used in building design and retrofit, energy regulation policy planning and private investment.Biomass is currently used for energetic and material purposes with an expected increase in the use of short rotation forests as a biofuel resource. DLR has created a service for estimating biomass potential and providing maps of the net primary productivity and total stem of forested areas in order to support the sustainable, economic and ecological development of a region with respect to timber and energy production. Several countries outside Europe are also exploiting biomass to produce products for energy use. Before importing them, Europe must assess the sustainability of the crops. DLR has developed a cost effective tool that identifies direct land use/cover change based on publicly available satellite data, therefore helping certification institutes in taking the relevant decisions.All services developed within ENDORSE are operational. Many of them are already on the market or are ready to become so. Many services are very innovative, and the attention they received from possible users reveals their potential.By bringing advanced methods into operational services, ENDORSE has demonstrated that the exploitation of Copernicus services and Earth observation data and models is an efficient means of providing solutions at the regional scale with regard to renewable energies. The availability of these new services will further stimulate the development of similar services in other regions.Project Context and Objectives:The European Union has committed to providing 20% of Europe’s energy from renewable sources by 2020. Based on proven facts, the concept of the ENDORSE project was that the energy community is mature enough to exploit Earth Observation data in an efficient, replicable and sustainable manner, especially with the advent of stable and reliable data streams from the GMES space components and the GMES Services: MACC for irradiance and other atmosphere-related parameters, and Geoland 2 for data on land cover and land use. The project ENDORSE aimed at a user-driven development of downstream services promoting the energy use from sun, wind, and biomass, electricity grid management and building engineering through daylighting in buildings. It addressed regional services and has collaborated with relevant users to design, elaborate and test products and services, and to assess the conditions for self-sustainability of the developed services.The ENDORSE Consortium is composed of ten partners, experts in energy or market analysis, which are research institutes or companies (SMEs):• Armines (France, coordinator) and the French engineering school MINES ParisTech as a third party,• the German aerospace research institute DLR (Germany),• iCons (Italy),• Transvalor (France),• Flyby (Italy),• the Hochschule Ulm (Germany),• the University of Genova (Italy),• the French engineering school ENTPE (France),• 3E (Belgium),• the Joint Research Center (JRC in Italy).Starting from explicit requests made by local users, public or private, a portfolio of ten innovative services, possibly replicable, has been identified, each of them having a market potential a priori. Elements of choice were:• experience of the Consortium on existing market conditions and maturity level of on-going R&D,• requests made by users to the Consortium,• for each service, the prime user is known to the Consortium,• the representativeness of the prime user of a large number of users in Europe,• prime user's views on their near-term market chances or near-term use of the downstream service,• good experience with motivation and expertise of prime users;• pre-existing experience and/or willingness of prime users to explore GMES/GEOSS products in their daily work and decisions.These services were called pre-market services. The new products have been developed and tested in collaboration with prime users. Then, the pre-market services have been designed, developed and validated with an extended panel of users. This panel of users has assessed the benefits of the services in their daily work and decisions.The objectives of ENDORSE were to:• perform advances in environmental modelling to be used in the development of products and services. Emphasis was on the assessment of surface air temperature and solar irradiance, and methods for the fusion of various irradiance data sets; • develop and validate pre-market downstream services in collaboration with well-defined end-users or to enhance existing services by the new GMES Service capabilities and recent R&D results;• assess the conditions for self-sustainability of these services through surveys and workshops with end-users;• disseminate the achievements of the project to foster the use of GMES Services data and other EO data by the renewable energies community. This includes feedbacks to GMES Services and connection to other GMES projects relating to energy, such as EnerGEO;• stimulate the market of downstream services in renewable energies towards the end-users community, and the development of such services by SMEs and other service-oriented companies by demonstrating precursors with documented conditions of sustainability.The expected major outcomes of ENDORSE were:• scientific advances in environmental modelling, data fusion, and assessment of surface air temperature and solar radiation;• a set of validated and documented innovative methods exploiting GMES Services data and other EO data;• a portfolio of pre-market services, serving as precursors and examples of best practices for similar downstream services (other regions, other providers), with documented conditions of sustainability, and a portfolio of improved existing services showing the value of GMES Services;• a stimulation of the renewable energies community towards exploitation of GMES Services data and other EO data;• a stimulation of the service industry towards development of downstream services;• feedbacks to GMES Services on their data, and as a whole to GMES and GEOSS on the exploitation of EO data in renewable energies area.To achieve the objectives over the three years of the ENDORSE project (2011-2013), the Consortium has planned a total of 269 person-months and a total budget of 3 176 340 €, of which 2 409 500 € are provided by the European Commission.Project Results:The development of the services in ENDORSE requested advances in environmental modeling, and especially on a more accurate assessment of surface air temperature and solar irradiance, and on efficient methods for the fusion of various irradiance data sets in order to produce a more accurate data set. All scientific results obtained in ENDORSE have been described in public reports and can therefore be used by the community. They have been shown in scientific conferences and several of them are the subject of published or submitted peer-review articles. In addition, a few of them are available as Web services.Prior to these research activities, several basic tools were developed. Several methods for products and services in ENDORSE call upon a very accurate knowledge of the position of the Sun in the sky and on meteorological measurements that are used for development, validation or in operation.1. A fast and accurate computation of the position of the Sun in the sky.The published SPA algorithm provides very accurate positions of the Sun in the sky but is not fast enough when in operation in the ENDORSE services. The requests from services developers in ENDORSE were: i) accuracy of order of 9 10-5 rad (approx. 0.005°, or 19"), ii) period of validity: 1980-2030, and iii) approximately 20 times faster than SPA. Three available fast algorithms were analyzed; none of them complies with all requirements. Therefore, a new algorithm, named SG2 (Solar Geometry-2), has been developed. It meets all requirements in a very satisfactory way and is therefore suitable for use in ENDORSE and other projects, including operations and services. It has been disseminated outside ENDORSE: a peer-reviewed article has been published, C and Matlab codes are freely available in two Web sites, and a Web service offers on-line computation.2. Harmonization of meteorological data A series of procedures and advanced mathematical tools have been developed to harmonize the meteorological data sets to support validation activities as well as the integration of data into product generation and services. Three generic types of problems were identified:• the over-sampling of time series of meteorological data (e.g. from a 1-h time series of irradiance to a 1-min time series of irradiance);• the sub-sampling time shift of time series of meteorological data (e.g. application of the sub-hourly time translation to a 1-h TST time series of irradiance to get a 1-h UT time series);• the averaging, inducing a down-sampling, of time series of meteorological data with missing data (e.g. from a sub-daily time series of irradiance with data gaps to daily, monthly or yearly time series of irradiance).State-of-the-art techniques were discussed and it was shown that none of them obey the consistency property in case of over-sampling. Two standard techniques (linear and cubic spline interpolations) were modified to obey the consistency property and benefits of the correction were proven. It was shown that missing sub-daily data have significant impact on the resulting daily average time series. This impact was quantified as a function of the percentage of missing data. A limit of acceptable percentage was proposed as a function of the foreseen exploitation of the daily data.3. Qualification of meteorological data Meteorological data measured by ground stations are often a key element in the development and validation of products and methods. As always in measurements, meteorological measurements are prone to errors, and gaps in time-series. Their quality is seldom known and they should be qualified before entering a method or serving as reference for validation. A review of the state-of-the-art in quality check of meteorological data was performed. The published procedures were tested and their results were analyzed. A series of automatic procedures have been established to assessing the plausibility of the measurements of the meteorological variables under concern. A Web service offers on-line check of the plausibility of irradiance measurements.4. Assessing the quality of the satellite-derived retrievalsBased on the comprehensive work performed in the assessment of solar radiation retrieved from satellite images in previous international programs funded by the European Commission or under the umbrella of the Global Earth Observation, International Energy Agency and International Renewable Energy Agency, a protocol is proposed for assessing the uncertainty of the satellite-derived variable compared to a reference. A series of recommendations for its application and guidance is provided. The protocol can be applied to various meteorological data and has been notably applied to the air temperature at 2-m height produced by the partner University of Genova.5. Assessment of 2-m air temperature from satellite imagesAir temperature at surface or 2-m height is usually known through ground measurements therefore at a limited number of places, or by meteorological re-analyses at coarse spatial resolution. R&D activities have been performed by the University of Genova to produce maps of air temperature at fine scale. Current state-of-the-art methods for mapping air temperature at fine scale are based on spatial interpolation of ground measurements, possibly in conjunction with correlated variables known as fields, e.g. digital elevation models. The innovation brought by ENDORSE is the exploitation of EO data to obtain maps of air temperature as well as maps of the uncertainty.Air temperature at 2-m height, and more exactly close to the ground-atmosphere interface, is assessed from satellite thermal infrared images and spatially sparse in situ temperature measurements. The developed method is based on support vector machines (SVMs) and also allows maps of the uncertainty in the air-temperature regression to be obtained as an additional result. Comparisons against qualified ground measurements demonstrated that the new method provide estimates of similar accuracy than state-of-the-art methods. However, unlike such methods, the new method also provides a model for the pixelwise error statistics, thus offering, for each air-temperature map, an accompanying map characterizing the pointwise uncertainty (expressed in terms of standard deviation) in the estimation of the air temperature at each pixel. Compared to the empirical error statistics on the test samples, the experimental results have suggested a remarkable accuracy in the estimation of the pixelwise error variance. This last aspect is considered an especially significant result, because of the intrinsic difficulty of the problem of pixelwise statistical characterization of image data and the related need to develop advanced models for the involved non-stationary stochastic processes.The developed approach is fully automatic thanks to the integration of an automatic parameter optimization technique based on the numerical minimization of a generalization error bound. Compared to the use of manually defined grid searches, this automatic approach obtains very similar accuracies with a sharp reduction in computational burden. The integration of the MACC surface solar irradiance and of the MERIS-derived normalized difference vegetation index (NDVI) within the features employed for regression lead to improvements, albeit rather minor, in air-estimation accuracy. Furthermore, the optimal strategy for exploiting the in situ measurements and training the proposed approach in different time slots was experimentally investigated according to the seasonal and daily dynamics of satellite infrared observations and air temperature.6. Improving surface solar irradiance (SSI) estimates using recent MACC aerosol datasets in the clear sky modelRecent updates of the GMES preparatory project MACC have been studied for their application within products and services of ENDORSE. This included especially the new MACC-AER subproject results on aerosol modeling and the new MACC clear-sky model: McClear, producing estimates of the surface solar irradiance (SSI) that should be observed if the sky were clear whose validation was made in 2012 by the MACC-RAD subproject.The MACC-legacy Helioclim-3 database contains values of SSI that have been constructed using the ESRA clear-sky model and a database of the Linke turbidity factor as input to ESRA. The ESRA model was replaced by the McClear model with the new MACC aerosol model as input to McClear. Comparisons against qualified ground measurements have demonstrated that for all summarizations, i.e. integration periods, and almost all stations, the performances of corrected SSI based on McClear are better than the Helioclim-3 results. There are few stations not showing better performances, but in these cases, the difference is not large. Furthermore, large SSI values are better reproduced, which is of major importance in solar energy research and energy yield predictions. In addition, more accurate SSI results in a more accurate estimation of direct and diffuse components by the means of global-to-direct models. For these reasons, it has been recommended to the MACC project to systematically correct the HelioClim-3 data set using the MACC aerosol data set and McClear.Additionally, the MACC-legacy SOLEMI database from DLR has been tested against qualified ground measurements making use of three concurrent inputs: i) the original GACP aerosol climatology, ii) the MATCH chemical transport model run providing hourly resolution but without any data assimilation, and iii) the recent MACC chemical transport model run providing data assimilation of MODIS satellite retrievals. Overall, the new MACC capabilities provide the best results for all statistical parameters with respect to global radiation retrievals. For the direct component at normal incidence, again MACC provides the smallest root mean square errors and the largest correlation coefficients for all stations. However, this is at the expenses of increased biases and KSI, a parameter describing the accuracy in retrieving the statistical distribution of the SSI. Improvements are clear for the global radiation if using the MACC data set and results are mixed for the direct component.7. Improving the global-to-direct conversionQuite often, only the global radiation is available in measurements. This is the case of the MACC-legacy HelioClim-3 database. Several ENDORSE services need the two components: diffuse and direct, whose sum is equal to the global. Several published conversion global-to-direct or transposition models exist that have been empirically set up using a few measurements. New models have been studied whose innovation is the exploitation of the McClear clear-sky model and cloud properties data from MACC. Such additional data constitute local and concomitant information about the atmospheric optical transparency and the cloud coverage that are not used on standard approaches of transposition model. Significant improvements in terms of bias and root mean square error have been observed when compared against qualified ground measurements.8. Create a global BNI database of the direct SSI received on a plane normal to the sun raysThe assessment of the direct SSI received on a plane normal to the sun rays (DNI, direct normal irradiance) is of particular importance in the assessment of yield of concentrated solar power plants and their sizing. ISIS is the only available data set on DNI covering the world. It has a three-hourly temporal and 50 km spatial resolution. A comparison with the more-detailed SOLEMI database demonstrated that ISIS exhibits significant underestimations and does not reproduce adequately the spatial distribution of the DNI. Against this background, different physical and empirical methods have been tested in order to develop a representative data set of DNI based on hourly values for the period 1984 - 2004. A worldwide database containing direct irradiance on normal incidence has been created.9. Improvements of the MACC solar irradiance using the MACC snow cover indicatorThe recent updates of the GMES preparatory project MACC include an indicator on snow cover derived as an intermediate product in the Heliosat-4 radiation service chain within the MACC-RAD subproject. Snow may be mistaken as clouds by the Heliosat-2 method, thus leading to underestimation of the irradiance in the MACC-legacy HelioClim-3 database. It was proven that a method for warning of large underestimation in case of snow cover can be set up, based on the MACC snow indicator. A model for a posteriori correction of the HelioClim-3 data in case of snow-cover condition has been developed. It has been shown that this correction generally improve significantly the performance of HelioClim-3 in case of snow cover. The outcomes have been communicated to the MACC-RAD team as they should prove useful for establishing a detection tool and a correction model for the new Heliosat-4 method in MACC.10. Evaluation of the uncertainties of surface solar irradiance retrieved by meteorological re-analysesMeteorological re-analyses such as the ERA-Interim and the MERRA ones provide surface solar irradiance (SSI) for long periods of time. This capability is appealing in solar energy as it may help in determining the potential of a given site in any part of the world. A comparison was made between qualified ground measurements of daily means of the SSI with the same quantity extracted from the ERA-Interim and the MERRA respectively for the period 1985 to 2009. 40 stations with no marked orographic features were retained located in Europe and Africa. It was found that the SSI from re-analyses exhibits a strong bias, most often an over-estimation of the measured SSI. The correlation coefficient is low compared to what is usually observed when comparing satellite-derived assessments and ground measurements. Further analyses demonstrate that the cloud cover of the ERA-Interim and the MERRA re-analyses is not reliable in case of cloudy skies. The ERA-Interim and MERRA re-analyses often underestimate the cloud cover and therefore predict clear skies while the sky is actually overcast. It is concluded that the SSI derived from the ERA-Interim and MERRA re-analyses should not be recommended for use in solar energy applications.11. Methods for data fusionDifferent data sets representing the same parameter are available but may have different spatial resolutions and cover different periods but with an overlap. A time-series spanning over one year of measurements made by a pyranometer at ground level, and a time-series of several years of global radiation estimated by the MACC-SOLEMI data set is an example. The first data set provides a better accuracy in terms of radiation, the second data set offers a longer coverage in time. It is tempting to combine both data sets in order to benefit from the advantages of both: great accuracy and long temporal coverage. This is called data fusion, i.e. methods for combining two or more data sets in order to obtain a synthesized data set that exhibits the best of the properties of the original data sets. There is no single method in data fusion. It depends upon the input data sets and of the objectives.Three different activities have been undertaken, all supporting the development of one or more services in ENDORSE. They may share data sets and methods but their objectives are different. The first activity dealt with the production of cartographic products for the whole world knowing that the current MACC products: HelioClim-3 or SOLEMI have a limited geographical coverage. A data set covering the whole world and having the accuracy of the MACC products would be highly profitable for mapping and atlas. The second activity dealt with the fusion of the MACC product HelioClim-3 covering period 2004 up to now, and its precursor HelioClim-1, ranging from 1985 to 2005. Combining such data sets would yield harmonized long-term time-series of interest in solar energy studies and in other domains. The third activity dealt with the fusion of the MACC product SOLEMI and data sets of ground measurements of better quality. The objective is to calibrate SOLEMI with a limited data set, and then to extent this calibration in time to yield a long-term time-series of accurate radiation data. The three activities have been run concurrently with mutual exchanges in data, assessment tools and procedures.The second and third activities were tackling a similar problem: merging data sets having different information support: HelioClim-1 and HelioClim-3 for the second activity, and SOLEMI and instruments for the third one. To a lesser extent, this was also the case in the first activity. Mapping and therefore synthesizing data sets at a better spatial resolution was the main objective of this activity. The method was developed but its application reveals the limitations of the association model, i.e. the model that links the MERRA re-analyses (large size in information support, or more exactly large distances between grid nodes) and HelioClim-3 (smaller pixel size). A further comparison between ground-based measurements and MERRA (or ERA) shows that the MERRA or ERA does not exhibit the proper cumulative distribution function.This observation supports those of activities #2 and #3. Using an affine function in the association model is not sufficient. Whether it is for hourly or daily irradiation, the bias may be reduced by data fusion techniques but may be increased in some cases which may be unpredictable. An adjustment of the cumulative distribution functions may offer promises but the currently proposed methods are not satisfying.The work done here has explored several paths for associating estimates of solar irradiation, made on information supports of different sizes: pinpoint to 100 km or so. No major result was achieved. Intermediate results were attained that may pave the way for the development of more accurate, efficient and reliable methods. Associating such different data sets is of strong interest in different fields in solar energy and deserves intellectual investment.12. Typical Meteorological YearA typical meteorological year (TMY) is a "composite" year of relevant meteorological dataset made of blocks of data from actual historical meteorological dataset for several years, typically more than 10 years. An advanced method for creating TMY dedicated to solar energy conversion systems (e.g. photovoltaic, solar water heating or solar thermodynamic systems) has been developed in ENDORSE. This result has been exploited to create a TMY product and a service delivering this product.13. Solar Radiation Atlas for ProvenceThe Solar Radiation Atlas for Provence is a set of digital maps describing the solar radiation that can be exploited in Provence for the production of electricity. Provence is a province in the southeast of France where the sky is often clear. The Atlas comprise additional maps related to ground elevation (elevation, slope and aspect), administrative limits, land-use, and other relevant geographic information (electric grids, natural risks, etc.). The spatial resolution is 200 m. It aims at helping the development of solar energy in the region and a better efficiency in the sitting and sizing of solar plants. These maps are available on the Web: http://www.atlas-solaire.fr/. This Atlas is now part of the French Géoportail (http://www.geoportail.gouv.fr/accueil) which is the governmental web site for offering access to geodata to citizens and public administrations.The construction of the solar radiation part of the Atlas comprises the calibration of the MACC-legacy HelioClim-3 database with a set of irradiance measurements performed by MeteoFrance. The process takes into account the ground elevation and the orographic effects due to shadows and sub-pixel altitude variation. This is done for each instant of the day and every day. These corrected values are then summarized to produce maps of multi-annual means of monthly and yearly global, direct and diffuse irradiations for the different plane orientations. The method has been validated against qualified ground measurements not used in the process.14. Service for generating local atlases in solar energyThe purpose of this service provided by Transvalor is to generate maps at local scale related to solar energy for decision-support in solar energy policy planning and private investment. 15. Service providing calibrated HelioClim-3 data setsThis service provided by Transvalor performs the calibration of a long-term MACC-legacy HelioClim-3 time-series using measurements acquired on the site under concern during a short period. The purpose of this operation is to provide high quality long-term radiation values on a specific site.16. Service for solar concentrating systemsThe service provided by Flyby offers support to identify the best site where to install a Concentrated Solar Power (CSP) plant, to configure the plant optimally and to estimate the profitability of the investment. It then provides the capability to monitor in near real time the performances of an existing CSP plant. The service exploits EO satellite based data (irradiance and air temperature maps) and in-situ data. 17. Service for generating TMYThe service provided by Transvalor aims at generating highly spatially-resolved Typical Meteorological Year (TMY) data sets for the design and the performance assessment of complex or not, solar energy-based systems for electricity production. Transvalor now regularly provides TMYs calibrated with the data of the Solar Atlas PACA (Service S1). This calibration is now fully automatized which permits a gain in both productivity and accuracy in the provided TMYs.18. Service for solar potential calculation (for CSP plants)The web service set up by DLR calculates available solar potentials useable for CSP plants based on user assumptions. This service might be used by commercial or political bodies for identifying solar potentials and locating suitable areas for CSP plants. This service is currently limited to the area of Morocco.19. WebGIS for CSP serviceThe WebGIS was developed by DLR as web-based map viewer and analysis tool to enable users to load, search and add relevant data for CSP and to query information of this data. Users are able to carry out simple location analysis by overlaying data, etc. The WebGIS was later enhanced to also host the “Solar potential calculation web service”. The WebGIS is intended for providing access to non-expert users and targeted towards political and commercial bodies.20. Service for assessing wind AEOThe service provided by 3E is a fully automated web service for energy production assessment in Belgium. Users are able to insert details of a proposed future wind farm and quickly receive a result giving estimated yields. This calculation is possible for any location in Belgium. The service may be exploited by consultancy offices in wind development, wind developers and authorities involved in wind development (permitting procedure) in order to get a preliminary energy production assessment of a wind project in Belgium. 21. Service for mapping biomass potentialThe service provided by DLR is a set of digital maps describing the above ground biomass increase of forests in Brandenburg, Germany. It offers an unique knowledge of how the woody biomass increase develops over time even if no in situ data is available.22. Service for mapping direct Land Cover Change (dLCC)The purpose of the service provided by DLR is to support certification systems and certification bodies with maps identifying direct land cover change. Based on the European and national laws, all biofuels must comply with sustainability requirements. As an example, cultivation of energy plants is not allowed on areas which were covered by rain forest after January 1st, 2008. Therefore, an intelligent comparison of information layer is developed based on globally available medium resolution remote sensing data to identify direct land cover change. The benefit of the dLCC maps to certification bodies is the reduction of costs for their investigations which land cover existed before January 1st, 2007 and after this date.23. Solar roof potential analysis: advice for successful implementationIn the course of its work on the use of EO data for load balancing in low and medium voltage electricity grid in the area of Ulm, Germany, Hochschule Ulm has ordered a “solar roof potential analysis” from an external service provider. Such an analysis identifies roofs where photovoltaic panels are installed and their capacity. A precise solar roof potential helps for a better understanding of strategic grid planning with high share of photovoltaic power input. It also helps to better calculate and reduce the investment cost for grid enforcement. The “solar roof potential analysis” supplied in the beginning of ENDORSE did not meet the expectations. During the process to improve the results together with the external service provider, Hochschule Ulm has collected a solid understanding on the does and don’ts to for a successful implementation of such an analysis. It was found that such knowledge is rare and that it deserves to be shared given the growing interest in such analyses. It was further found that further research on the detailed assessment of the quality of solar roof potential is needed especially with the use of this important input in electricity network planning with high share of photovoltaic power input. 24. Strategic grid planningThe electricity network planning was a “do it once and forget about it job” before renewable energy had been spread out into the distribution grid. Hochschule Ulm has demonstrated to the key customer SWU the potential of a strategic grid planning - combing solar roof potential analysis with grid calculation in the test area Einsingen, Germany. The results help to choose the right measures and the right timing of these measures in the process of increasing share of photovoltaic power systems in the distribution grid. 25. Grid status for DSO at low voltage levelWith higher penetration levels of photovoltaic power systems detailed information is needed about the grid status in this area of the distribution grid. So far no information on the actual grid status is available. Therefore this service proposed by Hochschule Ulm helps to implement the cost effective monitoring of the low voltage grid. During ENDORSE such a complex service and several measures to reduce the investment cost have been demonstrated for the test area Einsingen at Ulm. A further exploitation of this service needs further development to reduce cost and complexity as it is needed for all low voltage grid areas with high penetration of photovoltaic power systems. This service helps to ensure the quality of the electricity network operation and enables the monitoring of this quality. Also different cost effective measures to react on disturbance of this quality (e.g. precise curtailment of dedicated systems on very few hours during a year) have been evaluated during ENDORSE. Further research topics have been defined (cost reduction with the help of smart meter technology, implementation of roof potential analysis into grid control centre technology, validation of the grid status information quality from solar inverters...).26. Service for assessing annual daylight availability and lighting energy consumption in buildingsThe purpose of this service proposed by the French engineering school ENTPE is to allow a wide access to a research tool, which using MACC/GMES solar irradiance data, provides monthly and yearly statistics on indoor daylight availability and lighting energy savings resulting from the control of blinds and artificial lights by daylight. This information is needed by architects and engineers to make the most of daylight and reduce as much as possible the energy spent for lighting. Since this should not be made at the expense of the occupant’s visual comfort, the use of shades and its impact has to be taken into account. This information is essential to compute the return of investment related to the systems controlling the shades and dimming the artificial lights according to daylight. It is also useful to companies designing shading and control systems to evaluate and demonstrate the performance of their systems under real climatic conditions.The service has been developed around the PHANIE-2010 light simulation code from CSTB-France, which has been validated using simultaneous indoor and outdoor illuminance measurements, in Nantes, as part of the ENDORSE project. However, this was one climate and one building configuration; additional datasets are needed to fully quantify the quality of the results provided by the service; in a way the availability of the service should motivate building designers and owners to monitor daylight availability in buildings. The information provided by this service was not easily available prior to ENDORSE. Thanks to the service, the impact of architecture and climate on indoor daylight availability, use of shades, need for lights, will be better understood and this should result in optimizing the design of bioclimatic buildings. The energy savings related to dimming artificial lights according to daylight can now be quantified. This should increase the market penetration of control systems for lighting fixtures.27. Enhanced cloud statisticsEnhanced cloud statistics are derived by DLR from cloud physical parameter time series as provided by the MSG satellites within the GMES/Copernicus MACC-II project. In ENDORSE, they have been used to characterize the solar forecasting service validation locations further more. Nevertheless, the development has been done having in mind their potential use for the assessment of suitable solar power plant sites in an early planning phase of project development. After developing a set of metrics, an extended test phase with about 15 international users is on-going at the moment within ESA’s Value Adding Element program’s project ResGrow. Many of these users mention the use of such data as soon as solar systems are combined with local thermal or electric storage systems. This project supports a further market uptake of existing services/products in the renewable energy sector. A further enhancement of the metrics used and an improved visualization with respect to different user groups/application needs are expected. This is currently under development within ResGrow in an interactive and user-driven process.28. Service for irradiance forecasts for electricity productionENDORSE’s prime user Flagsol has developed jointly with DLR a solar production forecasting software tool, that makes use of numerical weather prediction forecasts combined with satellite imagery and nowcasting, ground measurements and cameras. Within ENDORSE DLR has further evaluated and enhanced methods used in this product. Up to now the prototype has been run in a split and distributed manner – partly at DLR premises and partly at Flagsol premises. A prototype system is installed in the Andasol-3 power plant in Spain. Potential Impact:The development of the services in ENDORSE has resulted in advances in environmental modeling, and especially on a more accurate assessment of surface air temperature and solar irradiance, and on efficient methods for the fusion of various irradiance data sets in order to produce a more accurate data set. All scientific results obtained in ENDORSE have been described in public reports and can therefore be used by the community. They have been shown in 26 international scientific conferences. Five articles have been published in peer-reviewed journals. Six others are in preparation and will be submitted in the forthcoming months.Several methods have been made available as Web services, i.e. computation can be performed on-line. These Web services are available in the SoDa Service, a Web portal dedicated to professionals in solar energy. In this way, the results are exposed to the 40000 users of the SoDa Service and one may expect a large uptake of these results by professionals, whether public or private. For example, after one year of existence, there have been approximately 300 launches of the Web service for an accurate computation of the position of the sun. In addition, 130 downloads of the corresponding software have been made in one year.ENDORSE has set up a series of innovative services for providing solutions at a regional scale. Five domains were targeted: wind, sun and biomass energies, electricity load balancing, and daylighting in building. Each product was specified in agreement with users. A protocol for the assessment of products by users was established with them. Methods for producing products were developed whose main innovation is the exploitation of Earth observation data and models. They call upon the results of the research activities in environmental modeling. In addition, ENDORSE has set up protocols for the scientific validation of methods and products. After tests of products and refinements, services were specified in collaboration with users as well as the protocol for service assessment and prototypes of these services, so-called pre-market services, were developed and tested by users. These methods and pre-market services are documented in public reports as part of the dissemination.The pre-market services were demonstrated in dedicated workshops. The workshops have addressed the target customer groups, the service industry as well as other relevant stakeholders such as authorities, policy makers and public administrations. Each workshop was dedicated to one or two services only to better interact with users, except InterSolar 2013, the largest fair in Europe in solar energy, where five services were demonstrated. Within workshops, ENDORSE has provided interactive demonstration of the services, where possible, and has consulted the audience on the compliance of the services to the market needs, their potentials and conditions of sustainability. Ten such workshops were held. Further intensive consultation and brainstorming with a much larger panel of users and other relevant stakeholders were held in the final phase of ENDORSE in order to understand and express the strategic value and the market potential of each pre-market service. Findings of the tests of services by users and comments collected during workshops were taken into account. The main challenge in this activity was to keep a consistent approach across all the 10 ENDORSE services and to apply the same methodological framework throughout, bearing in mind the characteristics of each service. A framework was proposed and refined after discussions with the Advisory Board and tests on a few cases. It proves successful and was used for all services. As a result, each pre-market service was given a report containing a market analysis, and guidance and recommendations to maximize its exploitation and integrate it in existing sectoral business practices. These conditions of sustainability are documented in a public report as part of the dissemination.By demonstrating the pre-market services and participating to international conferences in renewable energy and energy, ENDORSE has contributed to promote the use of GMES Services data and other EO data by the renewable energies community. This dissemination activity aimed at showing that EO data, GMES Services and services similar to those developed in ENDORSE may be used by consultants, companies and other stakeholders in their daily work. Dissemination has also targeted other GMES projects relating to energy, such as EnerGEO, GMES4REGIONS, GMES-PURE. One pre-market service is currently used in the EnergizAIR project funded by Intelligent Energy Europe which communicates the energy production from renewable energies to the wide public.An objective of ENDORSE was to stimulate the market by the virtue of examples, i.e. by showing to the service industry what has been done in ENDORSE and to suggest similar services might be created with possibly adaption to other regions and other local conditions, including cultural differences. ENDORSE used its portfolio of pre-market services as precursors and examples of best practices to this goal. The documented conditions of sustainability, the documented methods and pre-market services, and the availability of GMES Services should help companies not members of the ENDORSE Consortium to build new services in renewable energy. In five occasions during international events, a total of eight companies (Austria, Germany, India), two associations of companies in Poland, two scientific associations (Hungary, Romania), and four Weather Bureaus in Africa were approached with a demonstration of one or more pre-market services. In addition, one German company contacted ENDORSE for a possible implementation of a pre-market service. Despite reminders from ENDORSE, there was no result but it may be too early to judge.By participating to GMES events and demonstrating the pre-market services, ENDORSE has contributed to create awareness on the needs of the renewable energy community in GMES. Being a user of GMES Services, ENDORSE has documented the value of the GMES Services, namely MACC and GeoLand 2, and of GMES space components in renewable energy. Feedback on products was provided to MACC and GeoLand Services. ENDORSE has also contributed to further direct and indirect validation of MACC products (radiation, snow cover, cloud properties, aerosols).Looking backward to the development of the successful pre-market services, ENDORSE has drawn lessons that may help the European Commission in promoting successful downstream services.On a more technical ground, ENDORSE has provided two innovative algorithms to MACC-II where they are routinely used for the computation of the sun position and the quality control of meteorological data used in validation.ENDORSE has created several data, maps and Web services that deserve an efficient dissemination. The INSPIRE metadata and standards recommended by the GEOSS (Global Earth Observation System of Systems) were followed to ensure the interoperability of data and services. Being interoperable, these data, maps and Web services are visible from any GEOSS-compliant portal and catalogue. As an example, several ENDORSE maps are part of the Global Atlas for Renewable Energies of the International Renewable Energy Agency (IRENA). Because these standards were new, the activity was made in collaboration with the INSPIRE working group and the GEOSS Architecture and Data Committee and ENDORSE has served as a test case. As a consequence, the concerns of the renewable energy are now better taken into account in GEOSS. The knowledge gained and the best practices established during the ENDORSE project have contributed to the interoperability of data and applications within the project itself and within the GEOSS and GMES worlds. ENDORSE has started with a total of ten pre-market services. In the course of the project, 4 new results emerged having a priori market potential. The pre-market service proposed by Hochschule Ulm and the 2 related new services have revealed themselves complex to operate and not enough mature to be set in operation in the forthcoming months. They hold promises but need more development. The pre-market service proposed by the SME Flyby needs additional validation and hence refinements before entering the market.A total of 10 pre-market services and related services are being put in operation at the end of ENDORSE. The German aerospace center DLR will internally exploit four services by i) integrating the International Renewable Energy Agency (IRENA) as a new global user, ii) intensifying the collaboration with the SME Flagsol, iii) expanding the certification service to high resolution maps, and iv) increasing the biomass application to global demands. The French ENTPE will offer access to its service during 2014 before transferring marketing and management to the SME ESTIA, the prime-user. The market strategy is clear for the five other services that will be proposed by the SMEs Transvalor and 3E, part of the Consortium, and Flagsol that was associated as a prime-user. Products will be on sale in the coming months thus demonstrating that products from GMES Services may be exploited into sustainable services.List of Websites:http://www.endorse-fp7.eu/Prof. Dr. Lucien Wald, MINES ParisTech / Armines, CS10207, 06904 Sophia Antipolis cedex, France
