The future of the European electricity system and the impact of fluctuating renewable energy – A scenario analysis

Highlights

• Different paths for the transition of the European energy system until 2050 are analyzed.

• Infeeds from fluctuating RES are a main challenge – their market share in the EU reaches 29–44%.

• We use a stochastic energy system model with endogenous cost-resource curves for RES.

• A main driver in the different scenarios is the development of demand.

• A coordinated policy approach is important to integrate stochastic infeeds from RES.

Abstract

The ongoing transformation of the European energy system comes along with new challenges, notably increasing amounts of power generation from intermittent sources like wind and solar. How current objectives for emission reduction can be reached in the future and what the future power system will look like is, however, not fully clear. In particular, power plant investments in the long run and power plant dispatch in the short run are subject to considerable uncertainty. Therefore an approach is presented which allows electricity market development to be assessed in the presence of stochastic power feed-in and endogenous investments in power plants and renewable energies. To illustrate the range of possible future developments, five scenarios for the European electricity system up to 2050 are investigated. Both generation investments and dispatch as well as utilization of transmission lines are optimized for these scenarios and additional sensitivity analyses are carried out.

Introduction

In 1992, the United Nations Framework Convention on Climate Change defined the objective of stabilizing greenhouse emissions on an adequate level. Based on this, the European Union derived a temperature target to limit global temperature increase to 2 °C or below (1939th EU Council meeting, 1996). Later, this target was adopted on a global level (UN, 2010). According to the IPCC (2007), this means a reduction of CO₂-equivalent emissions between 50% and 80%. Especially the electricity sector has large potential for emission reduction, with scenarios of up to 100% electricity production from renewable energies (see ECF, 2010). Many of these concepts give the impression that target scenarios are developed neglecting any potential risk and imponderables along the envisaged way (see Laughton, 2007). But often the triangle of energy policy targets, consisting of security of supply, sustainability and economic efficiency, is not balanced. For example, de Jager et al. (2011) and Ragwitz et al. (2011) estimate that current investments have to double just to reach the 20% renewable energy production target of the European Union until 2020 (Directive 2009/28/EC). In the context of the global economic crisis and high national debts, these targets are ambitious and might be revised in the future in favor of economic growth.

In general, governments and societies may choose among four groups of approaches to reduce CO₂ emissions (Knopf et al., 2010) and fulfill the aforementioned targets. A first option is to focus on putting a price on greenhouse gas emissions. This can be done by taxes or via an emission trading system as applied in the European Union since 2005 (covering approximately 40% of Europeans emissions). Such an approach internalizes externalities for emissions, as Pigou (1920) has already shown, in principle. Difficulties occur when trying to determine a suitable price (or quantity) level.

A second approach is to foster technological progress to reduce emissions. This can be done via subsidies or tax reductions. In nearly all the countries of Europe, governments heavily support renewable energy sources (RES). Already at the end of 2011, there was wind power capacity of about 94 GW and solar capacity of about 50 GWp (EWEA, 2011). In 2011 alone, European solar capacity increased by 20 GWp. Italy (9 GWp) and Germany (7.5 GWp) were the main drivers of this growth due to high incentives. In the past also Spain saw high investments, but with a cut in subsidies market growth was reduced (EPIA, 2011). It is expected that PV will soon reach grid parity compared with household prices in many European countries due to learning curve effects if this has not yet happened, as in Italy. Actually, governments prefer direct interventions instead of setting the required framework. This is especially true for the feed-in of fluctuating energy sources like wind and solar energy. Different support instruments can be found all over Europe (Klessmann et al., 2011, Kitzing et al., 2012).

A third approach is “command and control”. This means that policy makers or regulators define a technology's specific upper limit for emissions. Of course, this only works when an appropriate compliance system is implemented. Upper limits for CO₂ emission of new vehicles are one example.

Life style changes are a fourth approach to reach the target. This means on the one hand the reduction of energy demand and on the other hand an increase of public acceptance of changes that come along with new technologies.

Especially the latter aspect indicates that the different approaches should not be handled separately. Rather, governments will probably apply a bundle of measures from all four groups to reach the objectives – yet the question is to what extent they will follow the three paths of the triangle of policy targets.

Setting different foci in the triangle of policy targets is analyzed in this paper and impacts on the development of the European power system are shown. Besides the description of target visions, also different paths to reach the final target are compared. Therefore the instruments described above are used to different extents.

Nevertheless, grid operators have to integrate the huge capacity of RES into the system. Notably technologies with intermittent production make system planning more difficult, because their capacity credit is limited compared with conventional generation and their stochastic behavior has an impact on both the dispatch of plants and also the long term investment planning in power systems (i.e. Möst and Fichtner, 2010, Swider and Weber, 2007, Tuohy et al., 2009). But this is not only a European problem. MacGill (2010) presents approaches to how to integrate fluctuating wind energy in the Australian power system where a renewable energy target has existed since 2001. Liu et al. (2011) have shown the limitations of the Chinese power system to integrate more than about 25% of wind energy and how this will affect grid stability. Baldick (2011) analyzes the integration of wind energy in Texas. He balances the effect of wind energy against carbon dioxide emissions and the related costs of wind extensions.

Especially the infeeds of PV and wind decrease the current price level and conventional power plants may be put out of business. On the other side, controllable power plant capacity is needed to provide system services and to cover demand when fluctuating renewable energies are not available. Hence, a suitable model framework is necessary which can handle all these influencing factors.

In order to consider the impact of renewable energy fluctuation, it is not sufficient to use deterministic planning tools as they were established previously, because these do not properly consider volatile generation (i.e. Tuohy et al., 2009). Several models have been developed to determine unit commitment and dispatch, taking into account the stochastic behavior of wind generators (i.e. Pappala et al., 2009, Ruiz et al., 2009, Garcia-Gonzalez et al., 2008), but these models are not designed for assessing long-term developments with endogenous investments, nor do they include the fluctuating behavior of other intermittent technologies like photovoltaic in particular. Thus, Nagl et al. (2012) present a stochastic linear system modeling approach for Europe in which they consider the uncertainty of having a year with high or low infeeds from wind and solar. But they neglect the uncertainty of the respective hourly dispatch decision by having a perfect forecast within one scenario branch (year). Notably Swider and Weber (2007) present such an approach including short-term uncertainties in the long-term investment decisions. There they use recombining trees to cover short-term uncertainties in wind infeeds and hydro inflows.

In the paper at hand, we combine the modeling of uncertainties in power plant dispatch and the inclusion of endogenous investments in renewable energies. Hence, we present a stochastic power system market model that takes the intermittent characteristics of wind and solar into account and is capable of modeling the whole European power market in order to evaluate future power system developments. We use the model to assess the influence of intermittent production of renewable energies on future power markets based on several scenarios. The scenarios reflect different overall objectives and a subsequent choice of instruments among those defined above in order to reach the general objectives.

The remainder of this article is organized as follows: first the applied model and enhanced methodology to consider stochastic inputs are described in Section 2. The investigated electricity system and scenarios are reviewed in Section 3. In Section 4 we present model results and discuss their implications. The article ends with brief conclusions on the achieved results.