Estimating the Norwegian circuit breaker population towards 2050

Central Office 21, rue d'Artois, 75008 Paris – France Web: www.cigre.org Accepted version Paper Information Paper number 1329 Paper title Estimating the Norwegian circuit breaker population towards 2050 Study Committee SC A3 –Transmission and distribution equipment Preferential subject PS2: Technologies supporting the power grid for energy transition to carbon neutral energy production Authors Thomas TREIDER, Salvatore D’ARCO Affiliations SINTEF Energy Research Country Norway Email address [email protected] Summary SF6-based circuit breakers (CBs) are to be replaced by SF6-free technology to reduce emissions of SF6 and comply with EU regulations. This paper aims to assess the impact of this transition in Norway in terms of the number of components required and reduction of emissions. The Norwegian CB population and its future development towards 2050 is estimated, along with the number of CBs needed to meet network expansion and replacement of older components. From this, the SF6 inventory and leaks are calculated. Two strategies for decommissioning existing CBs are compared: enforcing SF6-free operation by 2050 (“SF6-free by 2050”) or allowing CBs to reach their natural end-of-life (“Business as usual”). Assuming SF6-free technology becoming available according to timeline in EU’s regulations, “SF6-free by 2050” reduces the total emissions by 15% compared to “Business as usual”. The reduced emissions come at the expense of more CBs as “SF6-free by 2050” requires 10-50% more CBs to be commissioned by 2050, with the biggest increase seen at higher voltages where SF6-based technology is currently the norm. A delayed availability of SF6-free technology further increases the total emissions from the system and the need for new CBs to reach SF6-free operation in 2050. Keywords Circuit Breaker, SF6, Switchgear. Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 2 Accepted version 1 Introduction SF6 has been an important part of electrical switchgear due to its excellent properties for insulation and arc extinction, but its high global warming potential necessitates replacing SF6based switchgear with other technologies. Regulations by the European Union require new switchgear to be SF6-free by no later than 2032 [1], and technology providers have begun offering SF6-free alternatives for some voltage levels. In parallel with this transition, the power system is expected to grow significantly in the coming decades as a result of the partial or full electrification of industry and transportation and the transition to renewable power generation [2,3]. Together, these driving forces will result in a large number of new circuit breakers (CBs) being put into operation. Quantifying the economic and environmental impact of different rollout strategies is therefore important for TSOs, technology providers and policy makers. Some efforts have been made previously to make scenarios and projections into the future. For the overall development of the European energy system and power grid, scenarios developed by ENTSO-E and IEA are widely considered as the benchmark projections [2,3]. In [4] six European TSOs present short-term projections for new CBs and instrument transformers needed to meet the expansion of their respective grids leading up to 2032. In [5] long-term predictions were made for SF6-based components in Germany up to the year 2100, the authors of [6] developed scenarios on the same long-term horizon for the European medium voltage (MV) switchgear population and its SF6 contents, and [7] presents estimates of the present-day number of MV switchgear units in Europe. Existing estimates and scenarios focus either on single voltage levels or are limited to the near future only, lacking projections far into the future for all voltage levels. This paper aims to provide such estimates, focusing on the Norwegian CB population and its SF6 contents. To achieve this a method for estimating the present-day and future CB population has been developed and applied. The SF6 contents and emissions are then estimated based on typical SF6 contents for CBs at different voltage levels which are obtained from the scientific literature and via industry partners. The rigorous treatment of the population estimation method and its mathematical concepts is outside the scope of this paper and will be the subject of a separate publication [8]. This paper is result-oriented, presenting the important findings and industry-relevant results. The economic impact of the scenarios is not quantified as data on component costs have not been available, but the paper provides quantitative estimates on the number of components, enabling individual system operators and manufacturers to perform their own calculations as needed. The rest of the paper is organized as follows. Chapter 2 presents the methodology used, including both a summary of the population estimation method and an overview of the SF6data used as input to the calculations, while chapter 3 defines two strategies for decommissioning existing CBs. Chapter 4 presents estimated future asset needs and SF6 emissions with each of the two decommissioning strategies, demonstrating that the key point being the trade-off between the added economic cost and the reduced emissions that come from accelerating the transition to an SF6-free CB population. Chapter 5 concludes the paper. Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 3 Accepted version 2 Methodology 2.1 Estimating the present-day circuit breaker population and its future development To estimate the present-day CB population, the Open Street Map [9] is used to obtain an overview of the number of transmission lines and substations above 52 kV. Educated guesses regarding the average number of CBs associated with lines and substations at various voltage levels are used to convert the number of lines and substations to a number of CBs in the Norwegian power system. For medium voltage (MV, ≤ 52 kV), existing estimates for primary MV switchgear in Europe [7] are used as the Open Street Map database is not accurate for this voltage level. The European MV population estimate from [7] is converted to Norway directly using a factor 0.043 which has been derived based on the sizes of the two power systems (NOR/EUR = 135 TWh / 3150 TWh = 0.043). The assumed age of the CBs currently in operation is based on IEA’s estimates for the European power system [3]. Next, to estimate the future development of the CB population, scenarios from ENTSO-E’s TYNDP are used [2]. The yearly electric load and its percentage increase is used to estimate the percentage increase in the CB population, illustrated in Figure 1. In this paper the scenario “Distributed Energy” is used, which is characterized by “higher European autonomy with renewable and decentralised focus”, and a “transition initiated at a local / national level” [2]. The CBs currently in operation can be decommissioned in two ways: gradually, as they reach their natural end-of-life, or forced out sooner to accelerate the transition to SF6-free alternatives. These two decommissioning strategies are defined in more detail in section 3. The final step is to calculate the need for new CBs, which is done based on the estimated total population and the decommissioning rate of the existing CBs. 2.2 Estimating SF6 contents and emissions The CB population at each voltage level is assumed to consist of three main types: SF6-based AIS, SF6-based GIS, and non-SF6. References [4,7,10] have been used, together with input from industry partners, to obtain representative CB data: For MV it is assumed that 20% and 30% of the population is AIS and GIS, respectively (the remaining 50% are not SF6-based). For the other voltage groups, 80% and 20% are assumed to be AIS and GIS, respectively. Figure 2 summarizes the average SF6 contents assumed for AIS and GIS CBs at each voltage level as function of the CBs’ age. Figure 1 – Future circuit breaker population growth Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 4 Accepted version The emissions of SF6 are divided into operating leaks and end-of-life emissions and are also specified as function of the CBs’ age in Table 1. Both parameters are specified as a percentage of the nominal SF6 contents in Figure 2. A leakage rate of 0.1% is assumed to be the industry best-practice [5], while 1.5% is assumed best-practice for end-of-life emissions [6]. The operating leaks have been tweaked to obtain a yearly leakage rate in the range 0.15% – 0.2%, which is typical for Norway in recent years [10,11]. Table 1 –SF6 emissions from AIS and GIS as percentage of nominal SF6 contents 20+ years 10-19 years 0-9 years Future Yearly leaks 0.2% 0.15% 0.1% 0.1% End-of-life 2.5% 2% 1.5% 1.5% 3 Decommissioning strategies Two decommissioning strategies are defined: “Business as usual”, where CBs are allowed to reach their end of life naturally, and “SF6-free by 2050”, where they are forced out to remove all SF6 from the system by 2050. Table 2 summarizes the two strategies. For the “Business as usual” strategy it is assumed that the lifetime of any single CB follows a normal distribution with a mean value 𝜇 = 35 years and a standard deviation 𝜎 = 5 years. This implies that the average CB is in service for 35 years, and that 99.7% of CBs are in operation between 20 and 50 years. The “Business as usual” strategy also requires knowledge about the age of the CBs presently in operation. For this purpose, age estimates from [3] are used. For the “SF6-free by 2050” strategy it is assumed that the SF6-based CBs are decommissioned at a constant rate to reach 0 in 2050. Until SF6-free alternatives are available in unlimited quantities, and for the CBs not containing SF6, CBs are decommissioned according to “Business as usual”. SF6-free options will be assumed available in unlimited quantities in: • MV: 2026 • 52 – 75 kV: 2030 • 75 – 145 kV: 2028 • >145 kV: 2032 Figure 2 – SF6 contents per CB for AIS (left) and GIS (right) as function of voltage and component age Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 5 Accepted version Note that both scenarios assume that all new CBs are SF6-free roughly according to the timeline set forth by the EU [1]. This makes the “Business as usual” strategy more ambitious than the similarly named scenario in for example [6], where the use of SF6-based CBs is continued as today. Such a scenario is not considered to be realistic and therefore not relevant to include, but the effect of a delayed transition to SF6-free technology will be investigated later. Table 2 - Two decommissioning strategies Business as usual SF6-free by 2050 CBs are decommissioned when reaching their natural end-of-life, determined by a normal distribution (𝜇 = 35, 𝜎 = 5). New CBs are SF6-free when they become available in unlimited quantities. For MV, the share of SF6-free CBs among new CBs is gradually ramped up from today’s level described in section 2.2. Same as “Business as usual” until SF6-free CBs become available in unlimited quantities. Once SF6-free CBs are available, all SF6based CBs are decommissioned at a constant pace to obtain an SF6-free population in 2050. Old CBs are decommissioned first to maximize CB lifetime and SF6 reduction. 4 Results 4.1 Circuit breaker population and new circuit breakers needed by 2050 Table 3 shows the estimated present-day population. These are rough estimates, and for groups 2 to 6 the estimated population is in good agreement (i.e. ± 5% – 15% overall) with the numbers reported in [10] and numbers provided by the Norwegian GIS user group. Comparable estimates have not been found for group 1 and these numbers are therefore more uncertain. Voltage group 4 is vanishingly small in the Norwegian power system and is not analysed further. Table 3 - Estimated present-day CB population in Norway Group Voltage range Population Comments 1 MV (𝑽 ≤ 𝟓𝟐 𝒌𝑽) 65k Primary switchgear (See [7]) 2 𝟓𝟐 𝒌𝑽 <𝑽≤𝟕𝟓 𝒌𝑽 1500 Mostly 66 kV 3 𝟕𝟓 𝒌𝑽 < 𝑽 ≤ 𝟏𝟒𝟓 𝒌𝑽 2500 Mostly 132 kV, some 110 kV 4 𝟏𝟒𝟓 𝒌𝑽 <𝑽≤𝟐𝟒𝟓 𝒌𝑽 10 Only a handful of 220 kV components 5 𝟐𝟒𝟓 𝒌𝑽 <𝑽≤𝟑𝟎𝟎 𝒌𝑽 800 Only 300 kV 6 𝟑𝟎𝟎 𝒌𝑽 <𝑽≤𝟒𝟐𝟎 𝒌𝑽 700 Only 420 kV Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 6 Accepted version Figure 3 shows how many CBs are required by 2050 with each of the two decommissioning strategies, accounting both for replacement of old CBs and system expansion. The “SF6-free by 2050” strategy increases the number of CBs required in the five groups by approximately 10% – 50% compared to “Business as usual”. The biggest increase is seen at the highest voltages as this part of the population is currently entirely SF6-based and SF6-free alternatives become available at a later time. If the transition to SF6-free technology is delayed, implementing “SF6-free by 2050” will increase the number of new CBs required. This happens because more SF6-based CBs are added to the system before SF6-free alternatives become available (due to both system expansion and replacement of old CBs). These will then have to be decommissioned again by 2050, resulting in more CBs to be replaced prematurely than without the delayed transition. Figure 4 shows the results for “SF6-free by 2050” assuming that SF6-free alternatives are delayed by 5 years in all voltage groups (still assuming a linear ramp-up from today’s share of SF6-free CBs for MV CBs). Delaying the transition increases the number of new CBs by 5% for MV and approximately 15% for the other voltage groups. 4.2 SF6 inventory and emissions Figure 5 shows the SF6 inventory over time with each of the two decommissioning strategies. For comparison, the total inventory reported by the Norwegian GIS user group was approximately 420 tonnes in 2024, which is less than estimated here. A few factors contributing Figure 3 - Yearly and cumulative commissioning of CBs in each voltage group with “Business as usual” (top) and “SF6-free by 2050” (bottom) Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 7 Accepted version to this apparent discrepancy can be identified: the user group database does not cover all components in Norway, especially not at medium voltage. Furthermore, small variations in the assumed SF6 contents per CB, the age of the present-day population and the ratio of AIS to GIS CBs all impact the results. The estimated future inventory does not consider the gradual decrease due to the yearly leaks, but with small leaks this produces a negligible error (with 0.1% yearly leaks, only 5% of the initial gas volume leaks out over 50 years). In the case of “SF6-free by 2050” the inventory reaches 0 in 2050 by definition, while it with “Business as usual” has been reduced to approximately one third of its present-day value (it reaches 0 around 2080). In both cases the SF6 inventory decreases at a similar rate during the first few years before SF6-free alternatives become available. This decrease is mainly driven by the decommissioning of old CBs close to their expected service life, and these have significantly higher SF6 contents than their replacements. It is also evident that most of the SF6 is found in the highest voltage levels, so the delayed or accelerated availability of SF6-free alternatives here will have the biggest impact on the overall inventory. Note that the MV estimates reflect primary switchgear only, so the SF6 inventory at this level may be higher than shown here. For both decommissioning strategies the yearly leaks are simply calculated as a percentage of the total inventory that year, while the end-of-life emissions are calculated based on the number CBs being decommissioned each year. There are still SF6-based CBs left in operation in 2050 with the “Business as usual” strategy, and these will be a continued source of SF6 emissions in Figure 4 - Yearly and cumulative commissioning of CBs in each voltage group with “SF6-free by 2050” under a delayed transition Figure 5 - SF6 inventory towards 2050 with “Business as usual” (left) and “SF6-free by 2050” (right) Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 8 Accepted version the form of operational leaks and end-of-life emissions. Emissions beyond 2050 will therefore have to be accounted for to compare the total emissions with each strategy. The key difference between the two decommissioning strategies is that “SF6-free by 2050” will have lower operational leaks as the operational life of the CBs is shortened. The end-of-life emissions are also higher for this strategy if one only considers the timeline towards 2050 as all CBs, also those still functioning, are decommissioned by 2050. For “Business as usual”, the total end-of-life emissions are identical as for “SF6-free by 2050”, but they are spread out over a longer period and persist beyond 2050. The operating emissions will however be higher with “Business as usual”, resulting in higher total emissions overall with this strategy. The emissions are shown in Figure 6. If only yearly leaks are considered, then the “Business as usual” strategy has 30% higher emissions than “SF6-free by 2050” by 2100. Once the end-oflife emissions are added in – which are identical with both strategies because the same total number of CBs are decommissioned – the “Business as usual” strategy has only 15% higher emissions than “SF6-free by 2050”. Delaying the transition to SF6-free CBs greatly affects the SF6 emissions with both scenarios. Again, a 5-year delay in the transition to SF6-free CBs is assumed, and Figure 7 shows the resulting emissions by 2100. The end-of-life emissions increase by approximately 10% compared to the scenario in Figure 6 as more SF6-based CBs are introduced. The total Figure 6 - Emissions towards 2100 with “Business as usual” (left) and “SF6-free by 2050” (right) Figure 7 – Emissions with “Business as usual” (left) and “SF6-free by 2050” (right) assuming a 5-year delay on the availability of SF6-free alternatives Paper number 1329 Study committee: SC A3 –Transmission and distribution equipment Preferential subject PS2 thomas.treid[email protected] P a g e | 9 Accepted version operational leaks also increase for both strategies due the SF6 inventory remaining at a high level for a longer time. 5 Conclusions This paper has presented estimates for the Norwegian CB population and corresponding SF6 inventory, both today and towards 2050. While the estimates are based on several assumptions, they are in good agreement with the reported present-day population and SF6 inventory. Two decommissioning strategies have been compared: “Business as usual” in which existing CBs are replaced at their end-of-life, effectively phasing out all SF6 by the early 2080s, and “SF6-free by 2050” which forces all SF6-based CBs out of operation by 2050. The latter strategy requires 10-50% more CBs to be added by 2050 compared to the former, with the biggest difference seen at the highest voltage levels. The two decommissioning strategies represent the two main choices for TSOs, and for both of them the results demonstrate that a significant addition of CBs is required by 2050 due to the overall system growth and replacement of aging CBs. Ultimately, TSOs and DSOs must determine if the reduction achieved by applying the “SF6-free by 2050” strategy is worth the added cost of additional CBs. Comparing the total SF6 emissions by 2100, the “SF6-free by 2050” strategy reduces emissions by 15% compared to “Business as usual”. The added emissions from manufacturing and commissioning of this surplus of CBs are not included in the emission estimates. Such estimates would increase the emissions more for “SF6-free by 2050”, and an LCA analysis should be performed to complement the study performed in this paper. Future work will also focus on scaling up the results to the European level based on the grid characteristics of the different countries. 6 Acknowledgements This work has received funding from the European Union’s Horizon Europe research and innovation project under grant agreement no 101135484 (EU project MISSION). The authors would like to thank Nina Sasaki Støa-Aanensen, Mark Kuschel and the MISSION partners for their constructive feedback and helpful contributions to the work, and the Norwegian GIS user group for SF6 inventory and emission data.