Paper presented at Second Photovoltaic World Conference, Vienna July 1998



Bent Sørensen and Peter Meibom

Roskilde University, Institute of Mathematics and Physics, Energy & Environment Group.

email:, www:, fax +45 4674 3020

P.O.Box 260, DK-4000 Roskilde, Denmark, Phone +45 4674 2028


ABSTRACT: The purpose of this work is to establish a framework for assessing solar resources and performing demand matching in the context of photovoltaic systems modelling, assessment and planning purposes, all on a global level.

Keywords: GIS: geographical information system - 1: solar resource evaluation - 2: energy options -3


The approach is to employ a geographical information system (GIS) to map solar resources on the basis of satellite data (ra-diation at top of atmosphere, albedo, downward radiation at surface) and to match it with demand modelling on a habitat basis (population density, energy demand intensity). PV use is based on estimates of practical areas for collection use (building roof areas, suitably inclined and oriented surfaces) combined with land use data (important for central receiver fields). Local measurement data have been converted to the GIS grid employed. This is a novel approach compared to previous assessment, planning and scenario work in the pho-tovoltaic field, which has traditionally been country based.

By comparing the PV resource maps generated with present and projected future population and energy demand projections, assessments of supply-demand matching quality, transmission requirements and import-export options can be assessed and used to formulate planning strategies and build scenarios for future utilisation of photovoltaic power in the overall energy system. We use a similar approach for estimating the energy produced by other sources (wind, biomass).


Figures 1 shows the October solar radiation incident on a horizontal plane, calculated from satellite data (NASA, 1997) by a method devised by Pinker and Laszlo (1992). We use such data are used to estimate the solar radiation available for capture by photovoltaic devices. A very simplified approach is the following: Looking at measured data on tilted surfaces at selected locations (Sørensen, 1979), it appears that the Octo-ber and April horizontal data constitute a fair approximation to the average January and July radiation on a plane tilted towards North or South by an angle equal to the latitude. January and July span the range between extremities. For locations where ground-based measurements of direct and scattered radiation on tilted surfaces are available, one can do better (e.g. for Europe, a "reference year" has been constructed as an engineering and architectural design basis, for a total of 29 cities, according to Mørck, 1993). These specific data can also be used to construct refined global models translating horizontal to tilted data, but it should be kept in mind, that there is no universally valid relationship, as aerosol content of the atmosphere, details of cloud type and cloud cover, as well as the possible presence of reflecting surfaces, all influence the actual radiation on a tilted surface.

The electricity produced by solar panels is subsequently estimated on the basis of two assumptions: One is that the area of suitably oriented surfaces used for energy collection is at most 0.1% of the horizontal land area (the collectors being actually mounted on building roofs and facades, or in dedicated park-like collector fields) and the other is that of a fixed average efficiency for the collectors. This average efficiency, which pertains to year 2050 technology, is con-servatively taken as 15% for flat-plate photovoltaic panels. Figures 2 shows, for July, the resulting power production based on these assumptions together with the simplified translation of horizontal radiation data to tilted surfaces described above. The Figure thus depicts the expected average production for the month of July, for latitude-degree tilted N- or S-facing planes considered to be optimal for solar collection (Sørensen, 1979). There is a trade-off involved between maximum annual energy production and having some spread of production over summer and winter periods. For high-latitude locations, the winter power production on a horizontal surface would be almost nil, and the latitude tilting almost as good as a vertical orientation.

For deserts and other non-productive land, it is certainly possible to use more than 0.1% of the land area for PV. In these regions, the potential production may be over 100 times larger than the one indicated in the Figures.


As an example of the use of the GIS-based PV production data we use an energy demand model developed by one of the authors (Sørensen, 1996). It assumes the UN forecast of a 9800M world population for year 2050 (World Bank 1988), corresponding to an optimistic economic development and a relatively high level of goal satisfaction in currently deprived regions. The rate of total energy conversion to useful services at the end-user is 644 W/cap. , ranging from 1100 W/cap. in presently developed regions, over 700-800 W/cap. in the rest of the World, except for Africa that only reaches 200 W/cap. (Sørensen, 1998). For comparison the range today is from 1020 W/cap. (North America) to 130 W/cap. (Africa). Food energy is included in these numbers, which in the current, little efficient energy supply system requires a primary energy input of 2000 W/cap. (global average; OECD, 1996). It seems reasonable to assume, that if societies move towards high utilisation of renewable energy, it is because they are concerned with environmental issues and worried about sustainability, and therefore they will set energy efficiency as a high priority. Measures to improve the efficiency of transforming energy to a final product or service are typically cheaper than any supply increase, per unit of energy saved or supplied (Nielsen and Sørensen, 1998).

Figure 1. October radiation on horizontal plane used to represent average total January radiation on a plane tilted South or North by the latitude angle (based on satellite data on 2.5° × 2.5° grid, NASA, 1997; analysis by Pinker and Laszlo, 1992; Sørensen, 1979).

Figure 2. Total average PV power production in July, for panels tilted South or North by the latitude angle (based on radiation data for April).

The population distribution in GIS form is derived by data from CIESIN (1997), extrapolated to year 2050 in such a way, that they agree with UN forecasts on a regional level, and urbanisation trends are reflected except for a ceiling on maximum population density corresponding to the largest city kernel densities at present. The population GIS map is then folded with the per capita 2050 energy demand as derived above for each region. These data are given distributed on specific categories of end-use in Sørensen (1998), except that the space heating and cooling demands have been further elaborated to reflect the detailed geographical distribution of temperatures. Figures 3 and 4 give the annual average density of potential PV production (such as the content of Figures 2) expressed as percentages of the density of energy demand, separately for areas of deficit (<100%) and surplus (>100%). We have constructed similar GIS maps for wind power production (Sørensen and Meibom, 1998) and are currently working on maps of biofuel production.

Figure 3. Ratio of potential PV power production and total end-use energy demand in a 2050 energy scenario, expressed as percentage for all populated areas. Only areas of surplus (over 100%) are shown.

Figure 4. Ratio of potential PV power production and total end-use energy demand in a 2050 energy scenario, expressed as percentage for all populated areas. Only areas of deficit (under 100%) are shown.


Figures 3 and 4 show that the potential coverage of total energy requirements by photovoltaic power is quite high, despite the modest collector area assumptions made. As expected, there is a deficit in the most populated areas: all major cities, and extended areas notably in Europe, India and China. Addition of other renewable energy sources, such as wind and bioenergy is not going to change this pattern, as these sources are equally small in densely populated areas.

In order to establish a viable energy system, import and export of energy between regions, through transmission or other transport, is thus required. The calculations depicted show that this is indeed possible, as there is a substantial surplus in some regions. For several continents, the areas of deficit and surplus are sufficiently near to each other to allow regional transmission to cover all demand, e.g. by transferring power from rural to urban areas. Only in the three mentioned areas, a solution requires long-term transport of power. In Europe, this could be wind power from off-shore facilities or PV from Sahara, while in Asia, wind power from Manchuria and the Gobi desert could fulfil the need. The precise solution is part of the scenario work and it will depend on the energy sources chosen beyond solar power.


GIS maps such as the ones derived here for photovoltaics can be used to determine, if there is enough renewable energy to cover demands. From regions of surplus and regions of deficit, the need for exchange of power by electricity transmission, biogas or hydrogen pipelines, liquid biofuel transport or district heating lines can be determined. An example of this was shown in Sørensen (1998). Whether a high-PV scenario will be economically attractive may in the early phase depend on whether environmental and social costs are included in marked energy prices (cf. Kuemmel et al., 1997; Sørensen, 1997), but in the long range, no fossil or nuclear fuel based energy system can be sustainable, although the period could be extended to several hundred years with technologies based on clean coal or safe nuclear breeder reactors, technologies that yet have to be developed. With a renewable energy system proven technically and economically feasible, the task at hand is to devise policy measures, that have to be taken in order to make the transition happen with sufficient speed, and at the same time without forcing undesirable behaviour on the part of the investors and consumers (such as retiring equipment before its economical pay-back is achieved, unless retirement should turn out less expensive under some new external conditions imposed).

This work is part of a project on global energy alternatives performed for the Danish Energy Agency, looking at two renewable energy scenarios (centralised or decentralised), a carbon dioxide controlling fossil scenario and an advanced nuclear technology "safe nuclear" scenario.


CIESIN, 1997. Gridded population of the world. David Simonett Center for Space Studies at University of Santa Barbara, California, NCGIA Technical Report TR-95-6; Consortium for Int. Earth Science Information Network:

Kuemmel, B., Nielsen, S. and Sørensen, B., Life-cycle analy-sis of energy systems, Roskilde University Press.

Mørck, O., 1993. personal communication from CENERGIA on European reference year projects of IEA and EC.

NASA, 1997. Surface solar energy data set v1.0; NASA Langley Research Center EOSDIS Distributed Active Archive Center:

Nielsen, S. and Sørensen, B., 1998. A fair market scenario for the European energy system. pp. 127-186 in "Long-term integration of renewable energy sources into the European energy system" (LTI Research Group, ed.), Physica-Verlag, Heidelberg.

OECD, 1996. Energy balances of OECD countries 1993-4. Energy statistics of OECD countries 1993-4, Energy statistics of non-OECD countries 1993-4. Paris

Pinker, R. and Laszlo, I., 1992. Modelling surface solar irradiance for satellite applications on a global scale. J. Applied Meteorology, vol. 31, pp. 194-211

Sørensen, B., 1979. Renewable Energy. 687 pp. Academic Press, London and New York

Sørensen, B., 1996. Scenarios for greenhouse warming mitigation. Energy Conv. and Management, vol. 37, pp. 693-698.

Sørensen, B., 1997. Impacts of energy use, pp. 243-266 in "Human ecology, human ecology" (Diesendorf and Hamilton, eds.), Allen & Unwin, Sydney

Sørensen, B., 1998. Addressing market needs: a global per-spective. Paper for 4th Renewable Energy Technology Conf., Hobart 1998;

Sørensen, B. and Meibom, P., 1998. GIS tools for renewable energy modelling. Paper for World Renewable Energy Conf. V, Firenze;

World Bank, 1988. World population projections (Zacharia and Wu, eds.). Baltimore