Energy and Climate Change CommitteeWritten evidence submitted by the University of Bath (SEV32)
Addendum to “An Energy and Carbon Life Cycle Assessment of Tidal Power Case Study: the Proposed Cardiff-Weston Severn Barrage Scheme”, published May 2012
Abstract
A life cycle assessment was carried out on the “Cardiff-Weston” Severn barrage proposal as it was in 2010. The inventory analysis carried out for that work and the results of the life cycle energy and carbon analyses were published in the international journal Energy in May 2012(Kelly, 2012). That publication showed that the energy and carbon intensity of the Severn barrage is small in comparison to the National Grid mix and that the Severn barrage could contribute to meeting the UK carbon reduction target. Importantly, the operation stage was identified as both the most energy and carbon intense by a large margin. This is a notable finding as preceding studies have tended to dismiss the consequences of the barrage operation. This finding lead to a further “improvement” analysis in order to investigate the implications of operating the plant in ebb generation mode only and it was shown that the slight reduction in power output was far outweighed by the savings available in terms of energy demand and carbon emissions. This conclusion is further confirmed by the environmental impact analysis which shows that environmental impact of the operation stage is reduced so much by the removal of flood pumping that it in facts becomes the least impactful stage. When carbon and energy are examined in isolation, the operation stage is always the most dominant contributor. This does however mean that if, as is hoped, the National grid becomes more energy efficient and less carbon intensive the impact of the Severn barrage will follow suit.
1.1 Introduction
A life cycle assessment was carried out on the “Cardiff-Weston” Severn barrage proposal as it was in 2010. The inventory analysis carried out for that work and the results of the life cycle energy and carbon analyses were published in the international journal Energy in May 2012(Kelly, 2012). The inventory has since been subject to some small improvements, but the overall findings and conclusions published remain largely unchanged. This addendum presents the most up to date results for the study with regard to carbon and energy analysis. Perhaps more significantly, however, a full set of environmental impact results and interpretations are presented.
1.2 Environmental Impact Assessment Methodology
The impact assessment methodology “ReCiPe” was used to generate the results presented here. The ReCiPe 1.01(Goedkoop, De Schryver, Heijungs, Huijbregts, van Zelm, & Struijs, 2008) methodology was released in 2008 following a collaboration of RIVM, Radboud University, CML and PRé Consultants; their initials making up the capitalised letters in the methodology name. The method is largely a combination of the pre-existing Eco-indicator 99 (Goedkoop & Spriensma, Eco-indicator 99, 1999) methodology and CML method (Guinée, 2002).
1.3 Inventory Revisions
Mostly identical to that presented in the Energy publication (Kelly, 2012) but with the following minor changes:
Inclusion of some grid connection infrastructure at the construction. An adaptation of data from EcoInvent was used. Inventory data for the grid connection for 30 kW, 150 kW, 600 kW and 800 kW onshore windfarms (Burger & Bauer, 2007) was used to make a scaled estimate for the material required for the grid connection infrastructure.
Adoption of the study’s re-estimates for “on site” activities across all three barrage models ie “best”, “worst” and “initial”.
The mass of material required for the barrage turbines had been represented in both the construction and maintenance inventories by a factor of 1,000. This has been corrected, but it has been shown that this is only really significant when the model assumes that “flood pumping” is not employed, otherwise the operation stage is so dominant that this large charge is barely noticeable
1.4 Life Cycle Impact Assessment Results Interpretation
Table 1 summerises the revised carbon and energy key findings of the life cycle assessment case study.
Table 2 shows the characterised impact results for the three modelled life stages of the Severn barrage for the “initial” case and a potential error range generated by the “best” and “worst” models. The operation stage is the largest contributor in every impact category. The variation in results is also greatest at the operation stage.
Table 2 Characterised results by impact category of the life cycle stages of the Severn barrage, using Midpoint (H European) Analysis (to the nearest 6 s.f.s)
Figure 1 shows the normalised impact scores of the whole Severn barrage model according to impact category and includes error bars which depict the potential range of scores, from “best” case to “worst” case. Figure 2 shows the same data as that shown in Figure 1 but arranged according to life stage rather than by impact category so that the life stages can be compared more easily. The biggest contribution to the overall environmental impact in the normalized context is also from the operation stage, which also has the greatest range of error. This large operational impact was, however, ignored (Black & Veatch, 2007)(Woollcombe-Adams, Watson, & Shaw, 2009) or underestimated (Roberts, 1982) (Spevack, Jones, & Hammond, 2011) in the existing analyses previously reviewed (Kelly, 2012). This is probably due to a failure to acknowledge the proportion of the operational electricity demand that would not be met by the plant itself, as would normally be the case for an energy generation plant, nor the consequences of that inventory subtly. The total impact associated with decommissioning was estimated to be considerably less than that of construction. Given that the construction stage itself has been shown to be a minor contributor to the overall impact, it can now be estimated that the impact of the decommissioning stage would make negligible difference if it were included in the analysis.
1.5 Life Cycle Assessment Results Interpretation: Power In Context
The characterised results for each impact category per 1MWh of power generated are presented in Table 3.
Figure 3 compares the normalized impact score per 1MWh of the Severn barrage, with the potential range of error shown, to that of models of the UK National Grid mixes taken from the Transition Pathways Whole Systems work (Hammond, Howard, & Jones, The Energy and Environmental Implications of More Electric UK Transistion Pathways: A Whole System Perspective, In Press). It appears that the proportional spread of impacts across the suite categories included for the Severn barrage is extremely similar to all National Grid representations. This is not surprising because the overall impact of the Severn barrage is almost entirely made up of its electricity demand at the operation stage.
The relative normalized impact saving when compared to the 2050 grid is much great than the carbon saving. This is because the future scenarios developed by the Transition Pathways team (Hammond & Jones, Whole System Appraisal Feedback v 1.1, 2010) are optimised for low carbon generation rather than low impact or even low energy generation. For instance, the Market Rules scenario has a high proportion of coal fired energy generation with CCS, this explains why its specific GWP is just less than a twice that of the Severn barrage “initial” case while its specific normalized impact score is almost six times as much.
1.6 Improvement Analysis: Exclusion of “Flood Pumping”
Table 4 summarises the revised carbon and energy key findings of the life cycle assessment improvement analysis for the Severn barrage life cycle assessment under the assumption that the plant operates in ebb generation mode only.
Table 5 shows the characterised results for the lifetime environmental impact of the Severn barrage on the assumption that it operates in ebb generation mode only, the savings that are available over operating in ebb generation with “flood pumping” and the potential error range. Impact savings are available in every impact category and are 50% or more of the impact for the originally analysed system, i.e. ebb generation with flood pumping, in most instances, and are over 80% in seven out of 18 categories. It is only the category of metal depletion that sees a saving of much less than 50%; only a 3% saving is available in this category because the operation stage was already the smallest contributor, the largest being the maintenance stage which makes up 62% and 64% of the lifetime metal demand, with and without pumping respectively.
The total normalized impact score of the Severn barrage when it is assumed to operate in ebb generation mode only is 204,800,000, in a range of 325,900,000 to 173,900,000. Figure 4 compares the overall environmental impact of the Severn barrage when operating in ebb generation with flood pumping for its full lifetime to that of the Severn barrage when ebb generation only is adopted for operation, along with a possible range of error for each. The reduction in impact and in error range can be seen clearly. Importantly the maximum impact estimate for the Severn barrage when it is assumed that flood pumping is not adopted is less than the minimum impact estimate when it is assumed that it is, indicating that the excluding flood pumping will always yield a better environmental impact score, irrespective of what other decisions are made.
Figure 5 shows the normalized impact scores for Severn barrage when it is assumed the plant operates in ebb generation mode only for each impact category, with the contribution from each life stage shown and a possible range of error. When this figure is compared to Figure 1, the effect of removing the electricity demand can be seen.
Figure 6 shows the same results as Figure 5 but with the axis reversed so that the impact of the three life stages can be more easily compared. It can be seen that despite the long operational life of the barrage, the impact of the operation stage could be reduced so that it is the least impactful stage in almost all scenarios. The “worst” case operation impact score is approximately equal to the “best” case construction score, but this is the only instance where the separate life stages come close. In the “worst” case the construction stage is now estimated to be the dominant contributor to overall life impact, but in all other scenarios the maintenance stage dominates. Figure 6 also shows that the magnitude of the impact error range is now dominated by that of the range estimated at the construction stage. This indicates, if the plant operates without flood pumping, the next greatest impact reductions depend on the decisions made at the construction stage.
1.7 LiFeCyCle Assessment Results interpretation: Power In Context
It was assumed that operating the Severn barrage in ebb generation mode only would reduce its average annual output to 16 TWh and, hence, its lifetime power output to 1920 TWh (Kelly, 2012). Table 6 presents the specific characterised impact results per impact catergory. As was the case for the overall lifetime results, the removal of flood pumping offers significant savings over the originally modeled operational mode.
Figure 3 compares the specific normalized impact score of the Severn barrage, assuming ebb generation only, with that of the five representations of the UK National grid(Hammond, Howard, & Jones, The Energy and Environmental Implications of More Electric UK Transistion Pathways: A Whole System Perspective, In Press). It has already been shown that the barrage could offer significant impact savings even when flood pumping was employed, so it is not surprising that the removal of flood pumping just increases the magnitude of the savings available.
A further interesting observation can be made at this stage. When the operational impact is dominant, as is the case for ebb generation with flood pumping, increasing the operational lifetime of the plant would not significantly improve the specific impact as the overall impact would increase proportionally to the power output; however, if the “one off” activity of construction is the dominant contributor, as is estimated would be the case in the “worst” case scenario assuming ebb generation only, extending operational lifetime and hence lifetime power output would reduce specific impact and further increase the savings against the National Grid. If maintenance is the dominant contributor the effect on specific impact becomes more complicated as the maintenance regime that might be implemented after 120 years of life and its subsequent effect on power output is subject to a number of unknown, and arguably unknowable, variables.
1.8 Works Cited
Black & Veatch. (2007). Research Report 3—Severn barrage proposals. UK: Sustainable Development Commission.
Burger, B., & Bauer, C. (2007). Teil XIII: Windkraft. Switzerland: Paul Scherrer Institut Villigen, Swiss Centre for Life Cycle Inventories.
Goedkoop, M., & Spriensma, R. (1999). Eco-indicator 99. Netherlands: PRé Consultants.
Goedkoop, M., De Schryver, A., Heijungs, R., Huijbregts, M., van Zelm, R., & Struijs, J. (2008). ReCiPe. Netherlands: RIVM, CML, PRé Consultants, Radboud Universiteit Nijmegan and CE Delft.
Guinée, J. B. (2002). Handbook on LCA. Netherlands: CML, Universiteit Leiden.
Hammond, G., & Jones, C. (2010). Whole System Appraisal Feedback v 1.1. UK: Transition Pathways for a Low Carbon Economy Consortium Research.
Hammond, G., Howard, H., & Jones, C. (In Press). The Energy and Environmental Implications of More Electric UK Transistion Pathways: A Whole System Perspective. UK: Energy Policy.
Kelly, K. M. (2012). An Energy and Carbon Life Cycle Assessment of Tidal Power Case Study: the Proposed Cardiff-Weston Severn Barrage Scheme.
Roberts, F. (1982). Energy Accounting of River Severn Power Schemes. Applied Energy 11, pp 197–213.
Spevack, R., Jones, C., & Hammond, G. (2011). Technical Assessment of Two Tidal Power Barrage Schemes Across the River Severn. UK: University of Bath.
Woollcombe-Adams, C., Watson, M., & Shaw, T. (2009). Severn Barrage tidal power project implications for carbon emissions. UK: Water and Environment.
November 2012