**3. Results and Discussion**

This section describes the evolution in the sustainability indicators as calculated for the Spanish electricity system since 1990 and also for the alternative scenarios projected for 2030 and 2050. The year 2015 has been used as a reference for the current situation. As described above, the sustainability indicators cover three dimensions (environment, economic and socio-economic), which are structured into the following three sub-sections.

#### *3.1. Environmental Sustainability Assessment*

Figure 8 illustrates the evolution in the environmental sustainability of the Spanish electricity system in six categories. The analysis of the global warming category (top left) shows a rapid increase in total GHG emissions since 1990 (7.9 <sup>×</sup> 1010 t CO2/year) to reach a maximum in 2007 of 1.72 <sup>×</sup> <sup>10</sup><sup>11</sup> t CO2/year. This is caused primarily by the progressive growth in the economic activity of the country which demands increasing consumptions of electricity for industrial and domestic applications, as illustrated in Figure 2. During this period, the carbon intensity of the electricity system (GHG

emissions per unit of power generated) remains rather stable at values between 0.50–0.57 t CO2/MWh, due to the strong contribution of fossil fuels, primarily coal and oil.

**Figure 8.** Historic environmental performance of Spain's electricity system (1990–2015): (**a**) climate change, (**b**) fossil depletion, (**c**) ozone layer depletion, (**d**) terrestrial acidification, (**e**) human toxicity, (**f**) photochemical smog.In the first stage of the ST scenario (2030), the results show a notable reduction (44%) in the emissions per unit of power, due to a decline in the contribution of coal, the complete elimination of oil and the increasing weight of renewables (mainly PV). In the second stage of the ST scenario (2050), the higher impact per unit of power compared with 2030 is associated to the prevalence of natural gas in the power mix and the not so decisive upsurge in the penetration of renewables.

The carbon footprint of the power system is progressively reduced after 2008 reaching a minimum of 9.6 t CO2/year in 2014. This situation may be associated with the economic recession that the Spanish economy suffered between 2011 and 2013 due to the global financial crisis. By looking at the carbon intensity, the results evidence a progressive reduction in the emissions per unit of power between 2007 (0.56 t CO2/MWh) and 2014 (0.38 t CO2/MWh). This is attributable to the implementation of the national strategy endorsing the use of natural gas (burnt in higher efficiency and lower carbon intensity combined cycles) and the onset of ambitious national plans for the promotion of renewables. These two approaches progressively relegated the use of the more carbon intensive coal and oil power plants. A detailed description of the aggressive policies implemented by the Spanish Government in that period to favour the deployment of renewables may be revised in [40,41]. The progressive recovery of the Spanish economy and a renewed interest in the use of national coal for power generation explains the gradual upturn in GHG emissions after 2014, despite the increasing contribution of renewables.

A similar approach may be used to evaluate the evolution in the environmental performance of the Spanish electricity system on the other five categories considered in this investigation. In all cases, this performance may be related to changes in the overall demand and technology mix of the system. As explained in Figures 2 and 3, this is conditioned by the economic and political situation of the county, with a rapid economic growth and electrification based on fossil fuels between 1990 and 2000, and a more progressive increase in demand (typical of a more mature economic situation) based primarily on natural gas from 2008. From this date until the present, there is a gradual reduction in power generation partially attributable to the global economic crisis and also to the reinforcement of the tertiary sector and delocalization of more energy intensive economic activities. This period is also characterized by a progressive but strong public support for renewables. In order to avoid extending this section excessively, the discussion on the remaining environmental categories is more restrained, focusing primarily on general trends and key issues.

Thus, in the category describing the depletion of fossil resources (top middle), the results show a profile very similar to that of fossil fuel utilization. The small contribution of hydropower to this category is related to fossil fuel utilization and the consumption of other natural resources during the construction phase of the plants (mainly reservoirs). Regarding the ozone layer depletion category, the emissions follow a pattern strongly affected by the use of nuclear and natural gas power. This is due to the emission of halogenated hydrocarbons (Freon, Halon, CFCs, HCFCs, etc.) used as refrigerating and fire suppressing agents.

The results show a solid correlation between terrestrial acidification (bottom left) and fossil fuel utilization, primarily coal. Thus, the progressive reduction in coal contribution between 1990 and 2010 results in lower impact values in this category. Changes in the Spanish policies regarding the promotion of national coal or its substitution for natural gas and renewables, due to European commitments related to climate change, are responsible for the fluctuations in the acidification impact observed after 2010.

In terms of human toxicity, the assessment of the Spanish electricity system illustrates a progressive reduction in the emissions of 1.4 DB eq per unit of power between 1990 and 2010, later followed by a certain degree of stabilization. The first period may be explained by a progressive penetration of natural gas (at the expense of coal and oil) and the second by the significant contribution of renewables to this category. Regarding the photochemical smog category, the results evidence the strong contribution of coal and biomass plants, or rather, the comparatively smaller contribution to this category of all other power generation technologies. Therefore, the results show an increase in NMVOC emissions per unit of power generated between 1990 and 2005 due to the incorporation of biomass power plants. This is followed by a rapid reduction between 2006 and 2011 due to the smaller contribution of biomass and coal to the power mix, followed by an upturn after 2012 due to an increase in power demand and the incorporation of additional biomass capacity.

Figure 9 describes the same environmental profiles generated by the Spanish electricity system as determined in the four scenarios of 2030 and 2050. Regarding the climate change category, the results evidence a reduction in GHG emission in all cases, which is less marked in the stagnation (ST) scenario due to the prevalence of fossil fuels in the electricity mix.

**Figure 9.** Environmental performance of future projections of Spain's electricity system: 2030–2050: (**a**) climate change, (**b**) fossil depletion, (**c**) ozone layer depletion, (**d**) terrestrial acidification, (**e**) human toxicity, (**f**) photochemical smog.

On the opposite side of the spectrum, the decarbonisation scenario (DC) shows a very significant reduction (78%) in GHG emissions by 2030 due to complete elimination of coal and oil from the mix, and the strong contribution of renewables. This trend continues up until 2050 where all the power is generated by renewables (mainly PV and wind, with a small contribution from hydropower), resulting in a very reduced overall GHG output, both in total terms (2.4 <sup>×</sup> 1010 t CO2 eq) and per unit of power generated (0.037 t CO2 eq/MWh).

The other two scenarios (CP and AT) show similar patterns to each other in terms of overall GHG emissions. The impact on climate change generated in 2030 is expected to be less severe than that calculated in the DC scenario, due to the comparatively higher contribution of natural gas, which is used to smooth the transition towards the elimination of fossil fuels. In the longer term (2050), the three scenarios (DC, CP, AT) generate a similar impact in the climate change category. However, since the AT assumes a higher power demand, this results in a smaller impact per unit of power (0.033 t CO2 eq/MWh), which is achieved by permitting a certain contribution of nuclear up until 2050.

A simplified analysis of the situation regarding the other five impact categories and the four projected scenarios is provided below. Regarding the fossil depletion category, the results show a reduction in each scenario, except for ST, due to the substitution of fossil fuels for renewables. The benefits are more markedly observed in 2030 in the DC scenario, due to the more ambitious stance on renewables, compared to CP and AT. In 2050, the impact generated on this category was insignificant in the three scenarios that opted for renewables (DC, CP and AT), but even higher than the present situation in the ST scenario, due to its strong dependence on natural gas.

With regards to ozone layer depletion, the results show a strong dependence on the use of nuclear power and natural gas. These technologies are favoured in all four future scenarios in 2030, which is why impact values on these categories are not reduced in this time horizon. However, in the longer term (2050), the results evidence a significant reduction in the scenarios that assume the closure of nuclear power and the elimination of natural gas (DC and CP). In contrast, the stagnation (ST) scenario maintains a very high impact due to its reliance on natural gas.

Regarding terrestrial acidification, the results show a strong alleviation in this impact category in each scenario, except ST, due to elimination of coal from the mix. This effect is more marked in the DC scenario due to the more decisive penetration of renewables and elimination of sulphur containing fuels (mainly coal but also oil and natural gas). In terms of human toxicity, the results show a notable reduction both in gross emissions and per unit of power generated. These benefits are less marked in the ST scenarios due to the strong contribution of coal in 2030 and natural gas in 2050. The reduced impact in this category observed in 2030 in the DC scenario (compared against CP and AT) is due to the limited contribution of natural gas. The total impact on this category is still noticeable in 2050 due to toxic emissions associated with the life cycle of renewables, primarily in their fabrication stage.

Finally, the impact generated in the photochemical smog category is significantly reduced in each scenario due to the elimination of biomass and coal power plants (except ST in 2030). The use of natural gas to smooth the transition into renewables in the CP and AT scenarios is responsible for the higher impact on this category in 2030 and that is also why ST scenario shows a higher impact in this category of the ST scenario, compared to the other three. Of the other three, AT showed the lowest impact per unit of power in 2050 due to the higher power demand assumed and the utilization of nuclear stations.

The economic sustainability of the Spanish electricity system has been evaluated using the LCOE as the indicator. This analysis only covers the period 2010–2016 due to lack of information regarding the installation of additional capacity in earlier years. Figure 10 shows that the total cost of power generation in Spain in 2010 is dominated by fossil and nuclear technologies. The generation of renewables grows progressively in this period and so does their economic contribution to the electricity system. Overall, the total cost of electricity in Spain during this period (2010–2016) does not change significantly, due to the fact that the renewables with a higher contribution (wind and PV) are attributed similar costs to fossil technologies, while the contribution of higher cost renewables (CSP, biomass and biogas) is still marginal (see Figure 6).

#### 3.1.1. Electricity Costs in Historic and Present Data

Figure 11 shows a breakdown of the total cost of power generation in Spain during the reference year 2015, indicating the contribution of different technologies and life cycle stages.

The results show that most of the LCOE in the Spanish electricity system correspond to the capital cost of building the infrastructures (52%), followed by fuel costs (37%) and to a lesser extent the operation and management of the plants (11.1%). The main contributors to the capital costs are the nuclear energy and the renewable technologies (primarily wind). It is also remarkable the relatively high contribution of CSP to overall power costs, despite its limited generation share. Regarding fuel costs for 2015, the results show a large contribution of natural gas and a smaller proportion of coal, oil and nuclear. The contribution of biomass costs is very limited but still far greater than what should correspond to its limited generation capacity. The contribution of other renewables to this life cycle stage is obviously null.

**Figure 10.** Historic values of power generation costs estimated as a summation of LCOE of all contributing technologies.

**Figure 11.** Contribution of different technologies to the aggregated LCOE of Spain in 2015

#### 3.1.2. Electricity Costs in Future Projections

Figure 12 illustrates the cost of power generated in the four scenarios projected for 2030 and 2050. The results show a cost reduction per unit of energy generated (€/MWh) in each projected scenario when compared to the reference year of 2015 (69.12 €/MWh). Cost cuts grow larger between 2030 and 2050 in each scenario. These cuts are greater in the scenarios dominated by renewables (DC = 29.76 €/MWh; CP = 30.63 €/MWh; AT = 39.77 €/MWh in 2050) due to the cost reductions envisaged for wind and PV. The cost differences between the scenarios dominated by renewables and fossil technologies are less marked in the short term (2030) but become remarkable in the long term (2050) scenarios. Despite being slightly lower than that of 2015, the cost per unit of power of the scenarios with the strongest contribution of renewables (ST) is by far the highest of all in the long run (57.74 €/MWh in 2050).

Comparing the overall cost of the electricity systems is less apparent due to differences in the power demand considered in each scenario (see Figure 4). The results evidence a progressive cost reduction in the decarbonization (DC) and current policy (CP) scenarios when compared to the situation in 2015. This is so despite the significantly higher generation values considered in the future scenarios (DC = 477,073 GWh and CP = 416,698 GWh in 2050, compared to the 279,600 GWh for 2015). This is due to the strong penetration of renewables and the cost reductions envisaged for wind and solar. The higher overall cost generated by the AT scenario is due to the strong power demand associated with this case (581,930 GWh in 2050, almost double of that in 2015). Despite assuming the lowest generation values (ST = 352,507 GWh in 2050) the overall cost of the stagnation scenario was one of the highest due to the high economic intensity of fossil fuels and the limited contribution of renewables.

**Figure 12.** Power generation costs (LCOE) and contribution of different technologies in the four scenarios considered for 2030 and 2050 (DC = decarbonization; AP = maintaining current policies; AT = Advanced technologies; ST = stagnation).
