Figure: Possible electricity supply area divided
into 19 regions with schematic representation of potential
electricity transmission paths using HVDC to the geographic
population centres of the regions
Diminishing natural resources and global climate change are threatening the peaceful course of human development. A fundamental prerequisite for alleviating these dangers is to convert our energy system to renewable generation technologies that neither consume exhaustible resources nor degrade environmental quality despite continuous operation. Therefore also the questions have to be answered how the future electricity system should be structured, which techniques should be used and, of course, how costly the shift to a renewable electricity might be. This also raises the question how far we can get with the existing technologies and what costs are to be expected if one applies technologies at their today's costs, which is meant as a worst case assumption. It is apparent, however, only taking into account currently available technologies at their actual costs would constitute a worst-case cost assumption, since future developments will unquestionably improve economic performance. But on the other hand it is a conservative approach not overstressing the fantasy with optimistic cost assumptions and therefore it is resulting in a sound basis for further considerations.
The above-named questions have constituted the focus of a study to determine the minimum cost of an electricity supply realized for Europe and near-proximity Asian and African regions. The approach includes the option of supplying electrical energy to national economies not only or mainly from domestic resources but likewise in cooperation with neighbouring countries and distant regions using transmission systems to interconnect all participants within a wide-supply area containing huge renewable energy resources. The freedom for international cooperation between the nations opens up for synergetic benefits. Many of the nations within the area of consideration are emerging nations bounding on Europe with renewable potentials far in excess of national demand. Furthermore low population densities in deserts, steppes, tundra, and other regions largely devoid of human settlement neighbouring the European Union, however, may more easily allow for the expansion of wind power capacities and other renewable capacities than in highly populated areas. Very interesting resources are e.g. the huge potentials of wind energy of the coastal regions of Morocco and Mauritania which are particularly advantageous due to their summer peaks in production, which are the reverse of seasonal conditions in Europe. Solar electricity from concentrating parabolic arrays could likewise complement the output of wind farms in Germany, both inland and offshore. Due to these and further circumstances, the possibilities of wide-area interconnection – via the relatively cost-effective HVDC (High Voltage Direct Current) transmission - promise unprecedented economic and technical benefits for all participating nations.
Figure: Possible electricity supply area divided
into 19 regions with schematic representation of potential
electricity transmission paths using HVDC to the geographic
population centres of the regions
Taking this as starting point in Kassel scenarios for a future electricity supply entirely with renewable energies have been developed at the Institut für Solare Energieversorgungstechnik (ISET), Kassel, Germany. Various concepts have been studied for providing renewable electricity to Europe and neighbouring regions. Therefore an extensive region (s. figure) with approx. 1.1 billion inhabitants and an electricity consumption of roughly 4,000 TWh/a has been analysed to determine the available potentials for a future energy system. This process has taken into account data from ECMWF (European Centre for Medium-Range Weather Forecasts) as the meteorological basis and the population density as a restrictive factor for the wind energy potentials or estimated roof areas in all countries within the shown regions for determining the roof top photovoltaic potentials, combined with data on solar irradiation from ECMWF and National Center for Environmental Prediction (NCEP) and the National Center for Atmospheric Research (NCAR), wind speeds, and also temperatures used e.g. for photovoltaic electricity production and for solar thermal electricity production [ERA-15] [NCEP 99]. Also other renewable resources such as biomass and hydropower have been investigated or included at the level of current knowledge. All this has been fed into mathematical optimisation routines which have been applied to the question of which renewable resources with their individual temporal behaviour at different sites and with different yields should be used, and how selection should be made to achieve optimum cost performance. (A linear optimisation with roughly 2.45 million restrictions and about 2.2 million free variables was employed to find the best combination in each scenario.). The optimisation takes into account the temporal behaviour of the combined consumption of all countries within every individual region shown in the figure as well as all requirements imposed by resource-constrained production. Both sets of data, electricity demand and temporal behaviour of the possible production, have been compiled for optimisation (using time series with three-hour intervals) for all of the 19 regions to be supplied with electricity. The optimisation process ensures that supply will meet demand at any time, while determining if and to which extent any potential source is to be used, and how every part of the supply system will operate, including the dimensioning and operation of a HVDC grid that is superimposed on the current grid infrastructure. The criterion of optimisation is the minimization of overall annual costs of electricity when fed into the regional high-voltage grids, enabling these costs to be compared directly with those from regular power stations feeding into the conventional AC-high-voltage grid. However, the economic optimisation of all power plant operations for a time frame greater than, or equal to, three hours has simultaneously been included using sets of time series extending over one year.
The promising results for the base-case scenario – which assumes an electricity supply system implemented entirely with current technology using only renewable energies at today’s costs for all components (see [Czi 01] and [Czi 04] for more detailed information on underlying assumptions) – indicate that electricity could be produced and transported to the local grids at costs below 4.7 €ct/kWh, which hardly differs from the case of conventional generation today. (At gas prices in 2002 of about 2.4 €ct/kWh for industrial consumers in Germany [EC 4], electricity from newly erected combined-cycle gas power stations had already reached significantly higher at 5 €ct/kWhel.) In the resulting optimal configuration for this scenario, nearly 70 % of the power originates from wind energy produced from wind turbines with a rated power of 1,040 GW. A HVDC (High Voltage Direct Current) transmission system connects the good wind sites with the centres of demand while also powerfully integrating existing hydropower storage facilities, thus providing backup capacities that are enhanced by regional biomass power and given additional support by solar thermal electricity production. Electricity is generated from biomass at 6.6 €ct/kWhel after proceeds from heat sales have been factored in. This result lies significantly above the average price level, yet the backup capability is essential to reduce the overall cost of the entire system. About 42 % of the electricity produced is interregionally transmitted via the HVDC-System whereby the total transmission losses sum up to 4.2 of the electricity produced. Another 3.6 % loss is production which neither can be consumed at the time it is produced nor be stored for later use within the pumped storage plants and therefore is produced in excess. These two losses may be considered quite acceptable for an electricity supply only using renewable energies.
By contrast, if interregional transmission is not allowed in a restrictive decentralised scenario, excess production increases significantly to 10 % of the production, and additional backup power as well as backup energy employing other resources become necessary within individual isolated regions to meet the demand, leading to great additional expenses. In one scenario, fuel cells powered with renewable hydrogen produce electricity at about 20 €ct/kWhel (which is already quite optimistic if the hydrogen is produced from renewable energies), raising the net electricity costs to over 8 €ct/kWhel on the average. For Region 6 (Germany and Denmark), this restrictive “decentralized” (insular) strategy would lead to costs of electricity greater than 10 €ct/kWh.
The effect of cost changes for individual technologies and components was also investigated in particular scenarios. One aim was to find the costs at which PV could cost-effectively contribute to the supply. Therefore a series of scenarios has been calculated where the PV costs successively have been divided by two. As a result PV has not been chosen by the optimisation until costs have been halved three times. This major cost reduction for PV was found to enable this technology to provide a significant contribution to the electricity supply. If all other costs remained the same, the reduction to one-eighth of current PV costs would enable an economically viable 4 contribution to overall electricity generation to be provided. The generation would nevertheless be limited to the southernmost regions – particularly to regions 12,16, 17, and 18. If the cost were only one-sixteenth of present levels, PV technologies could account for about 22 of all electricity generation, reducing generation costs compared with the base case scenario by about 10 % to 4.3€ct/kWh. Even in this case, however, photovoltaic technologies would not be used in the northern regions 1, 2, 3, 6, 9, and 19, because they could not contribute to overall cost reductions.
If the costs of the mirror fields of solar thermal power plants were reduced by half – as is anticipated in the near future – solar thermal plants would already constitute about 13 % of all electricity generation. In this case, the overall electricity costs lie at 4 below those of the base case scenario. Reducing the costs of the collector array to 40 and simultaneously lowering storage costs to two-thirds of current levels (still clearly above achievable storage costs according to recent research mentioned in [Czi 04]) would increase their contribution to 28 % of the electricity produced, while the electricity generation costs would – compared to the base case scenario - fall by about 10 to 4.3 €ct/kWh. These examples illustrate that solar thermal generation presents an economically attractive perspective for the future that can be realized fairly easily in view of minimal cost regression factors.
The construction of a large hydroelectric power plant at an extremely favourable location in the Democratic Republic of Congo near Inga was also investigated for one proposed scenario (s. also [Kan 99]). The construction of a hydropower plant with a capacity of 38 GW was the decision resulting from computational optimisation. This would lower the costs of electricity by 5.3 % compared to the base-case scenario due to more economic generation and incidental system benefits. A primary reason for the low costs of the electricity produced at Inga is the high average load of the hydropower plant of about 6900 FLH and the relatively low anticipated investment costs at this very advantageous site. Two-thirds of the electricity produced at Inga is transmitted over a HVDC system with 26 GW capacity, connecting the generating station with Region 17, with the remainder conducted in equal amounts over two HVDC systems with a combined capacity of 12joining Inga with Regions 16 and 18.
In all scenarios – with the exception of restrictive and expensive insular configurations – electricity transmission is of significant importance. The necessary converter capacity (ACDC) for the HVDC grid exceeds values of over 750in some cases. (This level corresponds to about one-half of the installed generation capacity of all production facilities in the scenario regions.) The grid is used to achieve smoothing effects among different resource-dependent generation capacities using renewable energies, and to provide access to hydroelectric plants and to distributed biomass power plants both with its intrinsic storage capacity for wide-area backup applications. In the base case scenario, for instance, about 42 % of the electricity generated is transmitted via the HVDC system between the regions within the supply area. Measured against the total electricity costs the cost of the transmission system amounts to 7 of which the main part of 5 % is contributed by the transmission lines and cables. HVDC transmission has a higher intrinsic system stability than AC lines. Furthermore the transmission system of the base case scenario is highly redundant due to the fact that the thermal limit of the transmission lines is about twice the rated power and due to the fact that between almost all regions two or more systems are designed to be built parallel. But nevertheless if further redundancy was seen as desirable this could be relatively inexpensively achieved. A somewhat extreme idea would be to erect two whole systems of transmission lines in parallel. This would mean that the costs of transmission lines and cables would double but at the same time the losses would decrease due to the doubled cross section and thus the overall cost increase would only be about 3 % ensuring a degree of immunity against faults, which is by far higher than stipulated for today’s systems.
It is highly probable that these results can be transferred to other world regions, since every continent has its own renewable resources with different temporal production characteristics within a radius connectable via HVDC transmission. In some continents or regions hydropower is not exploited to the comparably high degree it is in Europe. This could negatively influence the available storage capacity. Solar energy potentials can be detected quite well with the available low resolution data used for this scenario study, showing the good conditions in many regions world wide. However, some huge regions are characterised by very rugged mountainous terrain where the detection of good wind sites is much more difficult than in smooth terrain. This means that many very good wind sites might remain hidden if meteorological data with low spatial resolution are used to search for the potentials, as done for the scenario study described here. If this technical problem were overcome, a much more positive assessment of the wind energy potentials can be expected. In the light of such an assessment it is clear that the general result – a low-cost but nevertheless totally renewable electricity supply is possible if the renewables are used in a huge powerfully interconnected supply area – holds for most areas of similar size to the European/Transeuropean example. Only the details would of cause have to be adapted to the local conditions. Furthermore there is no technical reason why, for example, southern Africa or eastern Asia should not be linked by a HVDC system to the supply area considered in the scenario study. So a future system might spread over some continents, gaining further advantages from further expansion.
The fundamental technical prerequisites for an electricity system realized entirely with renewable energies have already been fulfilled. The different scenarios show a broad range of various possibilities for a future electricity supply solely employing renewable energies and thus provide a basis for political decision. It can be deduced:
Thus the problem of converting our electricity system to one that is both globally competitive and environmentally and socially benign is therefore hardly a financial or technical issue, being instead almost entirely dependent on political attitudes and governmental priorities. There is more than enough evidence to justify a confident call for a comprehensive transition to a sustainable electricity supply, bearing in mind that a broad variety of solutions is possible. Responsible political decisions are now imperative for allocating the necessary technical, scientific and economic resources to achieve this goal.