The Scottish Government

« Previous | Contents | Next »

Listen

Technical Appendix A5. Simulation results: Renewable Energy Supply 1 - Input-Output Analysis

(Sub-title: The economic and environmental impacts of alternative electricity generation technologies in Scotland: An Input-Output analysis)

A5.1 Introduction

Concerns about energy security and meeting environmental targets in Scotland are in the spotlight of academic, policy and public debate. As of 2000, fossil fuel (coal and gas) and nuclear technologies provided 34%, 22% and 34% respectively of the total electricity generated in Scotland. Scotland also has a history of developing electricity generation from renewable sources. A significant amount of electricity, around 9.5%, was generated by hydroelectric facilities in 2000, which were largely built in the post- WW2 years. At the same time, the last ten years have seen the development of a significant number of electricity generating facilities from other renewable sources, as well as some extension of the hydroelectric capacity. The geographical position of Scotland offers it significant renewable energy resources, including on- and off-shore wind, wave and tidal energy. A recent study for the Scottish Executive (Boehme et al, 2006) quantifies the potential scale of renewable energy resources available and extractable around Scotland. We do not seek to quantify the potential here, but to gauge the possible economic impacts of changes to the Scottish electricity generation mix.

There are likely to be significant changes to the electricity generation mix in Scotland in the coming decades. The two nuclear power stations at Hunterston B and Torness currently have lifetime licences until 2016 and 2023 respectively ²¹, while current large-scale coal facilities at Longannet and Cockenzie will come under the Large Combustion Plant Directive ( LCPD) after 2015 ²². In the case of nuclear, the Scottish Government has stated that it does not want any new nuclear facilities constructed in Scotland. The Scottish Government has also recently set out ambitious targets for renewable electricity generation. These are that by 2020, 50% of electricity generated in Scotland will come from renewable sources, with an interim target of 31% by 2011 ²³. No specific targets for any particular technology have been set for either time period, although it has been suggested that much of the renewable electricity will come from significant increases in the amount of onshore wind generation. On the other hand, recent consultations by the Scottish Government on reforms to the support for renewable energy projects have recognised the potential for Scotland to develop an indigenous marine electricity industry, and have sought to provide additional incentives through the "banding" of existing support mechanisms to the production of electricity from marine ( i.e. wave and tidal) energy devices. Total electricity generated in Scotland from all renewable sources (hydro, wind, biomass, wave and landfill gas) has grown by 40 per cent between 2000 and 2006 ( BERR). The installed capacity of renewable energy (hydro, wind/wave, landfill gas and biomass) facilities increased over the same period from 1.4 GW to 2.4 GW. Some 0.9 GW of this increase has come from the development of wind energy projects, with an installed capacity in 2006 of 946 MW, generating 2,022 GWh in 2006. Figures on the generation of electricity from different technologies in Scotland, and the capacity of renewable energy technologies, between 2000 and 2006 are given in Tables A5.1 and A5.2 below.

Table A5.1: Current (2000) shares of electricity generation by technology and four scenarios considered, %

Note: Shares may not sum 100% due to rounding.

Table A5.2: Aggregate results on GDP, employment and CO ₂ emissions

This section of the report uses Input-Output ( IO) techniques to examine the economic and environmental consequences of significant changes in the electricity generation mix in Scotland. The motivation in using IO rather than CGE analysis in this section is because of the availability of an IO model with a greater disaggregation of the electricity sector than is currently incorporate in AMOSENVI. However, at such a time as which we are able to incorporate such a breakdown to AMOSENVI model, it would be desirable to repeat the analysis in a more flexible CGE framework.

In the present analysis, we use the IO modelling framework to develop four scenarios for the Scottish electricity generation mix. In each of the scenarios we have developed, we assume that the total electricity generated in Scotland is the same as in 2000, and we vary the generation mix. In each of the scenarios, the Scottish Government's target of 50% of electricity from renewable sources is met, and we assume that there is no generation from nuclear generation technologies. The types of renewable technologies that contribute to the renewables target are different in each case, but the common modal renewable technology is wind generation. We set out details of the IO model used for this analysis and the method used in Section A5.2. We provide details on each of the four scenarios in Section A5.3. In Section A5.4 we report the results for these simulations, where we focus on the aggregate and sectoral changes in economic activity, employment and emissions of CO2. Further, in Section A5.4, we carry out some sensitivity analysis regarding our assumptions about CO2 emissions factors, particularly with reference to those assumed for coal generation.

A5.2 Method and data

A5.2.1 Basic IO system²⁴

IO is a standard method for examining the interrelationships between sectors of the economy and final demand (Miller and Blair, 1985). If certain assumptions are imposed, it provides a powerful tool for examining how changes in the final demand for products can affect the outputs of other sectors within the economy. Although IO has traditionally been used for economic impact analysis (McGregor and McNicholl, 1992), it has been subsequently extended to energy and environmental areas. In the case of Scotland, recent IO work has covered the generation and treatment of waste (Allan et al, 2007b) and CO2 (McGregor et al, 2004, 2008).

For IO analysis, the output of each sector of the economy in question is given by an equation relating total output to the demands for that sector's goods from both intermediate demand ( i.e. other industrial sectors) and final demand. Final demands include, for example, consumption, government expenditure, and exports. Imposing constant returns to scale, a passive supply side, and unchanging technology allows specification of a set of linear equations of the sort

where X _i represents the output of sector i and a _ij represents the output of sector i that is required to produce one unit of output of sector j and Y _i. The a _ij coefficients are calibrated by dividing the value of the relevant intermediate purchases by the value of industry j's output. In matrix notation, the IO system can be expressed as

This says that gross output (X) is the sum of all intermediate sales (AX) (used in the production of all other industries' outputs) and sales to final demand (Y), which are taken to be exogenous, determined wholly outwith the system. Solving for gross output (X) yields

where I is an identity matrix and the term ( (1-A) ^-1) is known as the Leontief inverse matrix. The Leontief inverse matrix can be used to examine the extent of interrelationships among sectors within an economy, showing, as it does, the degree to which one sector relies upon the other sectors within an economic space for its inputs.

The system described above is the 'open' Leontief system in which all elements of final demand are considered to be exogenous and therefore are determined entirely outwith the system. The Leontief system can be 'closed' with respect to households, where the values of the Leontief inverse include not only the direct and indirect purchases necessary to meet changes in final demand, but where induced impacts, arising from endogenous consumption demands being linked to disposable incomes, are also included. (The income from employment row and consumer expenditure column from the IO table are, in this case, incorporated into the A matrix. The induced consumption effects are thereby incorporated in the multipliers.) A key feature of this system is that consumer expenditures are linked directly to households' disposable income, rather than being treated as exogenous as in the 'open' system. As income rises, this induces households to consume more. These induced impacts reveal the wider effect of the increased incomes of workers in sectors that have experienced increased demand for their outputs. We now turn to using the features of the Leontief inverse to examine interrelationships among sectors in the Scottish economy, specifically examining the degree to which the electricity-generating sectors are embedded into the regional economy.

A5.2.2 IO multipliers

Rasmussen (1958) proposes to use the open (Type 1) Leontief inverse to estimate the direct and indirect backward linkages. These are more commonly referred to as output multipliers in that they show the additional gross output generated across an economy from an additional unit of final demand for an individual sector. They are calculated as the column sums of the Leontief inverse matrix, thus

Where a _ij identifies the element located at row i and column j in the Leontief inverse matrix. The output multiplier is defined as 'the total value of production in all sectors of the economy that is necessary to satisfy a pound's worth of final demand for sector j's output' [12]. This Type 1 output multiplier incorporates both the direct and indirect impacts of the increased demand for sector j's output while taking household consumption to be exogenous. Closing the model with respect to households implies that the induced consumption effect of extra household income associated with increasing the aggregate output of a sector is included in the Type 2 output multiplier.

Although gross output is of interest, as a measure of turnover, it says nothing about how the changes in output affect gross-value added ( GVA or GDP) and employment. These can be calculated by multiplying the Type 1 and Type 2 Leontief inverses by the GVA-output and employment-output coefficients. Thus, the open GVA multiplier, , is

where v _i is the value added to gross output ratio in sector i. The value-added multiplier gives the increase in total value-added ( GDP) resulting from a pound's worth of final demand for sector j's output.

Employment multipliers can be found in a similar way, using physical employment/output coefficients, e _i. Thus, we use a vector of employment-output coefficients ( e _i) and multiply this by the open (for Type 1) or closed (for Type 2) Leontief inverse. CO ₂ multipliers can also be derived in a similar way, using CO ₂ emissions/output coefficients ( m _i). Again, we can use these multipliers to quantify the increase in total CO ₂ emissions resulting from a pounds worth of additional final demand for sector j's output.

A5.2.3 Modelling economic impacts of alternative electricity generation mixes

In this section we set out how we use the disaggregated IO table to model the economic and environmental effects of changes in the Scottish electricity generation mix. We have seen that significant changes are expected over the next two decades, and this might be expected to have impacts on aggregate and sectoral output and employment levels where the replacement electricity generation has different linkages to the regional economy than the generation that it displaces.

Our method is to revise the A matrix of input coefficients to reflect a new pattern of purchases by the electricity supply ( i.e. non-generation) sector from the eight electricity generation technologies. The pattern of purchases by this sector is altered in line with exogenously specified scenarios for the amount of generation coming from each different technology in Scotland. We set out and detail the four scenarios that we model in Section A.5.3.

We assume that the amount of electricity produced by the generation sector in Scotland, and purchased by the electricity supply sector, remains constant at 2000 levels ( i.e. at 49.5 TWh). Boehme et al. (2006) consider that Scottish domestic demand for electricity will rise from 32.4 TWh in 2003 to 41 TWh in 2020. This would imply that the level of Scottish exports of electricity may be lower in 2020 than it is currently. However, for simplicity we assume that total final demands for electricity remains constant at levels from the year 2000 ²⁵. Keeping the same final demand values ( ), we can then examine the impacts on output across all sectors in the Scottish economy using the equation below:

where the A* matrix is constructed by adjusting the coefficients for the purchases from electricity generation by the non-generation sectors. All other coefficients remain unchanged. In this case, the new sectoral output level can be used to calculate the change in sectoral employment, GDP and CO ₂ emissions driven by the change in the pattern of electricity generation in Scotland.

A5.2.4 Data and multiplier results for Scotland

The Input-Output table used in this chapter is that which is presented in Allan et al (2007a). ²⁶ This is a thirty-one sector table for Scotland, with a base year of 2000 in which particular care has been taken to disaggregate the electricity sector between generation and non-generation activities, and then to break down generation by the technology used. This disaggregation is important as it is understood that different electricity generation technologies have different linkages with the regional economy. Further, different generation technologies will also have significant differences in their environmental impact, for instance through direct (and indirect and induced) CO ₂ emissions. Standard disaggregations of the electricity sector in IO accounts would not account for the non-generation portion of the electricity sector, instead disaggregating the whole "Electricity" by various generation technologies employed within the economy. This however, has the effect of assuming that each generation type sells directly to the end consumer of the electricity, with each generation type paying for its own transmission, distribution and supply activities. Our disaggregation allows for a more realistic treatment, albeit illustrative, of the linkages between electricity generation technologies and the consumption of electricity by industrial and final demand categories.

The IO tables for Scotland, produced annually by the Scottish Government, are the starting point for this disaggregated table. However, survey work was carried out for electricity generation facilities so as to allow the separate identification of activities within this sector. This process is necessary as in the original IO tables (as published by the Scottish Government) there is a single sector identified as "Electricity" which covers all activities carried out by firms under SIC 2003 code 85, which includes not only generation of electricity, but also intermediate stages between generation and consumption of transmission, distribution and supply. To the extent that there are activities within this sector which are not related to generation of electricity, such a disaggregation of this sector is necessary. Full details of the identification of the generation technologies within an IO framework for Scotland are given in Allan et al (2007a). The sectoral breakdown for the thirty-one sectors are given in Table A5.3 below.

Table A5.3 Sectoral breakdown of 31 sector IO table and CO2-output coefficients for 2000

To allow us to generate a set of environmental results for this IO table, we required Scottish CO2-output coefficients for the same level of sectoral aggregation. ²⁷ These were primarily obtained from the appendix of sectoral CO2-output coefficients reported in Ferguson et al (2004), but additional information was needed for the direct CO2 emissions coefficients for the eight electricity generation sectors and the non-generation portion of the electricity sector. For renewable electricity generation technologies, we make the simplifying assumption that direct emissions of CO2 are zero. This is not the same as assuming that they have zero emissions indirectly, or over their life cycle. Indeed, these sectors will have positive indirect and induced CO2-output multipliers driven by the extent to which their backward linkages support activity elsewhere in Scotland which is itself CO ₂ emitting. The remaining non-renewable technologies are nuclear, coal and gas. For nuclear generation, we again assume that there are zero direct CO2 emissions in Scotland. For coal and gas electricity generating sources we use emissions factors from the Scottish Energy Study Volume 1 ( AEA Technology, 2006). For coal, this study reports emissions factors of 0.3 kgCO ₂/kWh for Scottish coal and 0.19 kgCo2/kWh for Scottish gas generating plants. We use these factors, and the output of coal and gas electricity generating plants in Scotland in 2000 to estimate base year emissions for these electricity generating technologies. In our thirty-one sector IO model, and including the direct emissions by households, total net CO ₂ emissions from Scotland in the year 2000 are 45.2 Mt of CO ₂. This is slightly lower than published estimates for net emissions of CO ₂ in Scotland for the year 2000 ( AEA Technology, 2008) which details net emissions of 49.7 Mt of CO ₂. This might suggest that our sectoral direct CO ₂-output coefficients which we have carried over from 1999 data are, on average, lower than estimated sectoral coefficients for the year 2000. When we present changes from this base year level of CO ₂ emissions, we might therefore expect that changes in CO ₂ generation are slightly underestimated for this reason. We now set out some of the multiplier results, focusing on the extent to which electricity generation technologies differ in their backward linkages to the regional economy.

A5.2.5 Multiplier results for electricity sectors

Sectoral multiplier results from our thirty-one sector IO table have been derived. Beginning with output multipliers, these are shown, for Type 1 and Type 2 cases, in Figure A5.9 below.

Figure A5.1: Sectoral output multipliers, Type 1 and Type 2, from thirty-one sector IO table for Scotland in 2000

The calculated output multipliers for the eight electricity generation technologies are reported at the right-hand end of Figure A5.1. Allan et al (2007a) show how the pattern of intermediate purchases by the electricity generation technologies differ, and explain the differences in the output multipliers for the electricity generation sectors. These results shows that there is considerable heterogeneity among the output multipliers for the electricity generation sector, which effectively amount to a separation of the individual generating components of the overall electricity multiplier. Without disaggregation of the table, the economic impact of changes in electricity generation would be constrained to the multiplier value for the original electricity sector (2.43 and 2.84 for Type 1 and Type 2, respectively), thereby masking the striking differences between generating technologies.

Furthermore, some of the most marked differences in output multipliers are those within the fossil-fuel-based generating technologies and within renewables, so that even aggregation over either sub-sector may be highly misleading. These are clearly quite different, with a £10million reduction in coal generation resulting in a £20.5million loss of aggregate Scottish output, whereas a comparable contraction in nuclear would generate only a £12.5million reduction in aggregate output (on the basis of Type 2 multipliers). This largely reflects their differential degrees of embeddedness in the Scottish economy, with nuclear having one of smallest "knock-on" (indirect) effects.

Similarly, it would matter a great deal, on our admittedly provisional estimates, whether this loss was to be compensated for by comparable increases in the output of onshore wind (which would generate a beneficial output effect of £12.2million) or marine generation technologies (associated with an output stimulus of £24.2million). Indeed, in terms of output effects, wind is an even more limiting case than nuclear, with a negligible indirect impact on the Scottish economy. Solely from the perspective of impact effects on output, reducing nuclear and replacing the output with marine, would appear to maximize the net benefit to Scotland if these data are indicative. Of course, care needs to be taken over such a comparison. These estimates relate to variations in output at the margin assuming variable capacity: they do not take account of the costs of providing new capacity to stimulate renewables, for example, or the costs of decommissioning nuclear or coal-based generating facilities. Furthermore, they make no allowance for the qualitative difference between nuclear and marine outputs, specifically the variability of the latter.

Figures A5.2 and A5.3 give the Type 1 and Type 2 GDP-output and employment-output multipliers respectively. As mentioned above, the sectoral GDP-output multipliers can be interpreted as the additional impact on aggregate GDP of an additional £1 million of final demand for the output of each sector. An estimated (Type 2) GDP-output multiplier for the coal generation sector of 0.88 means that a £1 million increase in final demand of the coal generation sector would increase aggregate Scottish GDP by £0.88million. Sectoral employment-output multipliers can be interpreted as the additional aggregate employment generated in the Scottish economy by an additional £1 million final demand for the output of each sector. The sectoral employment-output multiplier (Type 2) for hydro generation is 23.1, implying that for an additional £1 million of final demand for the hydro generation sector, aggregate employment across Scotland is raised by 23.1 FTE jobs.

Figure A5.2: Sectoral GDP-output multipliers, Type 1 and Type 2, from thirty-one sector IO table for Scotland in 2000

Figure A5.3: Sectoral employment-output multipliers, Type 1 and Type 2, from thirty-one sector IO table for Scotland in 2000

Sectors which have high value-added to output ratios exhibit relatively high GDP-output multipliers, with nuclear and wind consequently improving their overall rankings. The top-ranked electricity generation sector in terms of GDP-multiplier values is landfill gas, which reflects a combination of high output multipliers and moderate value-added intensity. Sectors with high employment to output ratios experience a bigger employment boost, explaining the major rise in the ranking of marine and the decline in nuclear and wind when compared with the GDP-output multiplier values. Looking at the GDP-output and employment output effects, replacing nuclear and coal with hydro, landfill gas, or wind, would suggest an economic boost to GDP, whereas the employment effects would be greatest from marine, landfill gas, and hydroelectric generation. Again, the caveat that these differences relate solely to the operational stages of electricity generation applies (Proops et al, 1996; Hondo, 2005) .

We report the direct CO2-output coefficient (tonnes of CO2 per £1million of sectoral output) and the Type 1 and Type 2 CO ₂-output multipliers in Figure A5.4. While the first column reports how much is generated by £1million of output, the second and third columns here relate the amount (tonnes) of CO ₂ generated across the Scottish economy following a £1 million increase in the final demand for each sector with and without households endogenised. In the case of the electricity sectors, these results are especially heterogeneous, as may be expected. Recall firstly that renewable sectors, and the nuclear sector, were assumed to have emissions factors (CO ₂ emissions/output) of zero. There are positive Type 1 CO ₂-output multipliers for these sectors however, as an increase in the final demand for these sectors will generate additional demand through the backward intermediate linkages between this sector and other industrial sectors in the Scottish economy. Type 1 CO ₂-output multipliers for the electricity generation sectors with zero direct emissions, vary from 0 for the wind sector, to 271 for the hydro generation sector. Type 2 CO ₂-output multipliers for these electricity generation sectors will be higher than Type 1, given that we now also capture the additional CO ₂ emissions generated by the spending of increased levels of wages in these sectors being spent by Scottish households. Type 2 results for these non-directly emitting sectors are shown in the third column in each case in Figure A5.4.

Figure A5.4: Direct CO ₂-output coefficients and sectoral CO ₂-output multipliers, Type 1 and Type 2, from thirty-one sector IO table for Scotland in 2000

CO2-output multipliers for coal and gas electricity generation are the highest and second-highest values seen in Figure A5.12. These sectors direct, indirect and induced effects on CO2 emissions across Scotland are massively greater than for any other sector included in this analysis ²⁸. This represents these sectors high direct CO2-output coefficients in our base year data. An additional £1 million of final demand for the output of the coal and gas generation sectors, increases Scottish CO2 emissions (using Type 2 results) by 13,728 and 8,663 tonnes of CO2 respectively. The CO2-output (Type 1 and Type 2) multiplier for the non-generation sector is significantly higher than its direct emissions factor due to the nature of the disaggregation of the electricity sector. Each additional £1 million of final demand for the non-generation electricity sector will increase in turn the demand for each generation technology, in proportion to that technology's share in base year generation.

A5.3 Scenarios

We model the impacts of four alternative scenarios for the electricity generation mix in Scotland (see Table A5.4 below). As stated in Section A5.1, in each of these scenarios 50% of electricity comes from renewable energy sources, the majority of which comes from onshore wind. Further, in none of these scenarios is there any generation from nuclear sources in Scotland. None of these scenarios are referenced against expected or predicted changes in the pattern of electricity generation mix in Scotland, or make any assumptions about the costs or viability of any of the scenarios considered here - such as, for instance, whether each scenario provides sufficient generation to meet expected future demand or to provide appropriate margins between peak demands and supply capacity ²⁹. We use these scenarios purely to illustrate the usefulness of the IO method for estimating the economic impact of large changes in the pattern of electricity generation. We begin by briefly sketching the features of each of the scenarios considered.

Table A5.4: Current (2000) shares of electricity generation by technology and four scenarios considered, %

Note: Shares may not sum 100% due to rounding.

Scenario A: Technology mix - high marine

Under this scenario, generation from coal falls slightly compared to its base year levels, while generation from gas technologies rises slightly. Together, these technologies provide 50% of electricity generated in Scotland under this scenario. Generation from hydroelectric facilities increases by fifty per cent, up to providing 15% of electricity generated. Biomass and landfill gas increase their contribution to the Scottish electricity generation mix, rising to provide 3% and 2% of total generation in this scenario. Marine provides 10% of electricity generation capacity, with wind providing the remaining 20%.

Scenario B: Technology mix - low marine

All technologies are assumed to provide the same share of electricity generated in Scotland under this scenario, with the exception of marine and wind. Under this scenario, the proportion of electricity generation from wind is 25%, and the proportion generated from marine sources is assumed to be 5%. Such a change from Scenario A could be consistent with a less successful outcome for marine-specific support mechanisms, in terms of bring forward marine electricity generation, with wind generation dominating.

Scenario C: No Gas

Under this scenario, 50% of electricity generated in Scotland comes from renewable sources - with wind providing 30%, and hydro and marine providing 15% and 5% of total electricity generated in Scotland respectively. The remaining 50% of electricity generation is met through coal generation, with gas generation providing 0%. Output of biomass and landfill gas falls from current levels to zero.

Scenario D: No Coal

In this final scenario, renewable technologies provide the same specific and aggregate proportions of Scottish electricity generation, but rather than coal providing the remaining 50%, this is met through gas generation. By comparing Scenario C with Scenario D, we can examine the economic and environmental impacts from coal, or gas, generation providing the non-renewable portion of future Scottish electricity generation.

A5.4 Results

Table A5.5 presents the main aggregate results on GDP, employment and CO ₂ emissions for each of the four scenarios outlined above.

Table A5.5: Aggregate results on GDP, employment and CO ₂ emissions

Recall that the only difference in Scenario A compared to Scenario B is that there are higher amounts of wind and lower amounts of marine electricity generated. In Scenario B there is 25% of electricity generation from wind and 5% from marine, while in Scenario A there is 20% of electricity from wind and 10% of electricity from marine sources. The higher amount of marine generation, combined with that sector's output multiplier being significantly higher than that for wind generation, result in a greater economic boost to Scotland than in the lower wind case. The impact of an additional 5% of electricity from marine sources, rather than from wind generation, is to increase GDP by £109.55 million, and increase employment by 11813 FTE jobs with Type 1 analysis, and, under the Type 2 IO model, to raise GDP by £168.64 million and employment by 13615 FTE jobs.

The increased economic impact, and activity, generated in Scenario A compared to Scenario B, comes at the expense of a slightly smaller decline in CO2 emissions, as is reflected in the smaller reduction in the CO2 intensity of Scottish production indicator (CO2/ GDP) in Table A5.2. Under Scenario A, emissions of CO2 are 3.52% lower under Type 1 analysis, and 5.69% lower with Type 2. Under Scenario B, CO ₂ emissions are down by 3.59% and 5.89% under Type 1 and Type 2 respectively. This greater decline under Scenario B is to be expected since economic activity is greater under Scenario A (due to the additional stimulus offered by the marine generation sector) and so CO2 emissions are slightly higher - although reduced relative to the base year. This is reflected in the results for the CO2 intensity of Scottish Production, which declines by 3.9% with Type 1 and 6.28% for Type 2 under Scenario A, and by slightly less, 3.82% and 6.24% respectively under Scenario B. Under Scenario C, when it is assumed that the non-renewable 50% of electricity generation in Scotland comes solely from coal generation, the GDP and employment impact is not as large as Scenario A - an additional £287.91 million on GDP and 15.738 FTE jobs under Type 2 results. The CO2 impact however is different, with Type 1 CO2 emissions actually increased relative to the base year, and an increase in the CO2 intensity of Scottish production of 0.51%. This arises due to the assumed CO2 emitting nature of coal generation technologies. The Type 2 change in CO ₂ emissions shows a decline relative to the base year of 3.08% - a smaller fall in emissions than either Scenarios A or B - and a much smaller, 3.5%, decrease in the CO2 intensity of Scottish production. Under Scenario D, the smallest increased in GDP is observed (0.17% with Type 1 and 0.27% with Type 2) but the biggest Type 2 reduction in CO2 emissions (8.86%), which gives us the biggest Type 2 reduction in the CO2 intensity of Scottish production (9.11%). This is due largely to the absence of coal generation technologies.

These results suggest that the composition of the renewables technologies which contribute to meeting the 50% target is important. Technologies with strong backward linkages back to the Scottish economy provide the greatest possibilities for an economic gain to be realised. What is suggested by Scenarios C and Scenario D is that it matters what is assumed about the technologies which provide the other 50% of electricity generated in Scotland. Without nuclear generation, this would be likely to be met through either a combination of gas and coal technologies, or, as extreme cases, from each technology alone (e.g coal in Scenario C and gas in Scenario D). As with the wind/marine results in Scenarios A and B, the economic results for these scenarios can be explained with reference to the initial linkages of each sector. Coal generation sector has greater employment-output and GDP-output multipliers than the gas generation sector in our initial IO framework. The scenario that assumes coal technologies, rather than gas generation, provides the non-renewable element of future Scottish electricity generation sees higher economic benefits, although these are associated with smaller declines in CO2 emissions.

Sectoral results for Scenarios A and B

Recall that, in all four Scenarios, we assume that 50% of electricity generated comes from renewable sources and that there is no generation from nuclear technologies in Scotland. The demand for electricity is unchanged, so Scotland remains a net exporter of electricity to the rest of the UK. In Scenarios A and B, 30% of electricity is generated from wind and marine sources, but in Scenario A 10% comes from marine sources and 20% from wind, while in Scenario B 5% comes from marine sources and 25% from wind. All other generation technologies have the same share, so the difference in the results from Scenario A to B are solely driven by the switch from marine generation to wind generation. While the aggregate results are discussed above, we focus here on the sectoral differences in these results. Absolute sectoral changes in GDP (in £million) are shown for Scenarios A and B in Figures A5.5 and A5.6 respectively.

Figure A5.5: Absolute sectoral changes in GDP, £million, in Scenario A (high marine)

Figure A5.6: Absolute sectoral changes in GDP, £million, in Scenario B (low marine)

While the change in most sectors GDP is similar in Scenario A and B, it can clearly be seen that as well as significant changes in the wind and marine generation sectors, Scenario A sees significantly greater activity in the "Construction", "Communications, finance and business" and "Transport and other machinery". In both the Type 1 and Type 2 results, moving from Scenario A to Scenario B the change in GDP in these sectors decreases by almost fifty per cent. As seen in Section A5.2 above, these are sectors with which the marine generation sector has strong backward linkages.

The absolute change in sectoral employment in Scenarios A and B is shown in Figures A5.7 and A5.8. This shows the extent to which employment at the sectoral level is affected by the larger marine or wind generation in Scotland. The sectoral pattern of impacts may be different to that seen in Figures A5.5 and A5.6 since sectors that are GVA-intensive, are not necessarily employment intensive (as was seen in Section A5.2 above). As would be expected, in Figure A5.6, where the largest aggregate impact on employment is found, this is largely explained by the expansion of the marine sector, but also partly by the model, but significant, increase in employment in the "Construction" sector.

Figure A5.7: Absolute sectoral changes in employment, FTEs, in Scenario A (higher marine, lower wind)

Figure A5.8: Absolute sectoral changes in employment, FTEs, in Scenario B (lower marine, higher wind)

Sectoral results for Scenarios C and D

In Scenarios C and D, we assume that the non-renewable element of future Scottish electricity generation comes from two extreme possibilities - purely coal generation, and then purely gas generation. Note again, that we assume that no electricity in Scotland is generated from nuclear sources, and that the total demand for electricity us unchanged, so Scotland remains a net exporter of electricity to the rest of the UK. The renewables' share of the future electricity generation mix in both Scenarios C and D remains the same in each scenario, with 30% from wind, 15% from hydro and 5% from marine technologies. The differences in results between Scenarios C and D therefore come solely from coal generation providing the whole of the remaining 50% of Scotland's electricity generation in Scenario C, while gas generation provides this 50% under Scenario D. While the aggregate economic and environmental results are discussed above, we focus here on the sectoral differences in these results. Absolute sectoral changes in GDP (in £million) are shown for Scenarios C and D in Figures A5.9 and A5.10 respectively.

Figure A5.9: Absolute sectoral changes in GDP, £million, in Scenario C

Figure A5.10: Absolute sectoral changes in GDP, £million, in Scenario D

While the results between Scenarios C and D are approximately the same for hydro, marine and wind generation sectors, there are considerable differences among the non-renewable sectors, and also in sectors that have strong links to the non-renewable sectors. The expansion of the "Coal generation" sector in Scenario B, with results in an increase not only in the "Coal generation" sector itself, but also sees an expansion in the "Coal extraction" sector (of almost 15%) and an expansion, large in absolute terms, in the "Communications finance and business" sector. Both these sectors have links to the "Coal generation" sector in the base year IO table. The "Gas refining" sector exhibits a contraction in Scenario C and an expansion in Scenario D, as would be expected. In Scenario D GDP in the "Gas refining" sector rises by over 21%, while it falls by almost 19% in Scenario C.

The absolute changes in sectoral employment in Scenarios C and D are shown in Figures A5.11 and A5.12. This indicates the extent to which employment at the sectoral level is affected by coal or gas generation providing the non-renewable portion of future Scottish electricity outputs. As with Scenarios A and B, the biggest employment impact is in additional jobs for the expanded marine generation sector. Employment in the construction sector is higher in both scenarios, while the same negative effect as found for GDP exists for employment in the "Coal extraction" sector in Scenario D and the "Gas refining" sector in Scenario C.

Figure A5.11: Absolute sectoral changes in employment, FTEs, in Scenario C

Figure A5.12: Absolute sectoral changes in employment, FTEs, in Scenario D

A5.4.1. Sensitivity analysis

Sensitivity to emissions factors for coal generation technologies

We used figures from the Scottish Energy Study ( AEA Technology, 2006) as CO2 emissions factors (kg of CO2 produced per kWh generated). For electricity generation from gas and coal technologies respectively these were 0.19 and 0.3 kgCO2/kWh. Specifically for future coal technologies, legislation requiring "cleaner" generation makes it likely that there will be significant reductions in the quantities of CO2 produced per kWh generated from existing levels. Carbon Capture and Storage ( CCS) technology, which involves extracting CO2 emissions directly at the point of production and storing these emissions, perhaps in underground "sinks", can reduce the emissions per kWh from coal by up to 90% ³⁰. We can examine the impact of this reduction, by reducing the CO2 emissions-output coefficient for the "Coal generation" sector by 90% - from 0.3 kg/kWh to 0.03 kg/kWh.

With this adjustment to the emissions factor for the "Coal generation" sector, emissions under each scenario are given in Table A5.6

Table A5.6: Change in CO ₂ emissions, % from base year, under for scenarios with original emissions coefficients and 90% reduction in Coal generation emissions coefficients

In each of the scenarios in which coal plays a role (A-C), the reduction in the Coal CO ₂ emissions factor has a significant impact upon the resulting CO2 emissions. In Scenarios A and B the share of electricity generated from Coal sources is constant at 25%. Reducing the CO2 emissions per unit of output in the Coal generation sector by 90% significantly lowers the emissions in both of these scenarios, with emissions down by 15% and 13% under Type 1 and 13% in Type 2 results respectively. The biggest change from the original and adjusted Coal emissions coefficient unsurprisingly is in Scenario C, where we assume that Coal generation produces 50% of the total electricity generated in Scotland. Emissions of CO ₂, which rose under Type 1 analysis for the original coefficients, decline by 22% and almost 18% under Type 1 and Type 2 analysis respectively. Results for total CO ₂ emissions in Scenario D are unchanged, as in this scenario there is no electricity generated in Scotland from coal sources.

« Previous | Contents | Next »