Environmental Sciences

ISSN: 1569-3430 (Print) 1744-4225 (Online) Journal homepage: www.tandfonline.com/journals/nens19

Reductions of SF6 emissions from high and

medium voltage electrical equipment in Europe

S. Wartmann & J. Harnisch

To cite this article: S. Wartmann & J. Harnisch (2005) Reductions of SF6 emissions from high
and medium voltage electrical equipment in Europe, Environmental Sciences, 2:2-3, 273-281,
DOI: 10.1080/15693430500402234
To link to this article: https://doi.org/10.1080/15693430500402234

Published online: 16 Feb 2007.

Submit your article to this journal

Article views: 1115

View related articles

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=nens20
Environmental Sciences
June-September 2005; 2(2-3): 273 – 281

Reductions of SF6 emissions from high and medium voltage

electrical equipment in Europe

S. WARTMANN & J. HARNISCH

Ecofys GmbH, Nuremberg Germany

Abstract
Electrical equipment using SF6 forms an integral part of the electricity networks, providing a safe and
reliable power supply throughout the world including most European countries. SF6-emissions
potentially occur in all lifecycle-phases of electrical equipment—manufacture, use and decommission-
ing. A study summarized in this paper finds that SF6 emissions throughout the life-cycle have
substantially been reduced since the mid-nineties. Almost all available and economically feasible
reduction options have been accessed at least partly. The study demonstrates that, until 2020, further
reductions can be achieved if these reduction options are fully implemented throughout Europe. This
paper gives an overview of reduction potentials and costs, as well as the real and projected emission
development between 1995 – 2020.

Keywords: Sulphur hexafluoride, electrical equipment, emission reduction, cost

1. Introduction
SF6 electrical equipment applications are used in power systems, with switchgear being the
dominating application. Switchgear is employed to protect electrical equipment against
overload and short circuit currents as well as to connect and disconnect networks. In electrical
equipment applications, SF6 is used for insulation and/or arc-quenching. In 2003, the
industry sector of SF6 electrical equipment had a turnover of roughly e3.5 billion in the EU-
25þSwitzerland, Norway and Iceland (‘‘EU-25þ3’’), with an export rate between 60 – 70%,
depending on the type of equipment. The number of employees amounted to roughly 30,000.
The equipment installed presently in the EU-25þ3 is estimated to have a value of at least
e20 billion.
SF6 has a 22,200 times higher specific global warming potential than carbon dioxide and an
atmospheric lifetime of approximately 3,600 years compared with roughly 100 years in the
case of CO2 (IPCC 1996). For 2002, it is estimated that SF6 has contributed only 0.2% of
total EU-15 greenhouse gas emissions (EEA 2004). Apart from its status as greenhouse gas
and its long atmospheric lifetime, SF6 is generally appreciated for its complete inertness, its
lack of acute or chronic toxicity and its excellent dielectric and arc-quenching properties. A
systematic search for environmentally friendly alternatives to SF6 in the field of electrical
equipment started in the 1980s and is still ongoing within the international scientific

Correspondence: S. Wartmann, Ecofys GmbH, Nuremberg Germany. E-mail: [email protected]

ISSN 1569-3430 print/ISSN 1744-4225 online ! 2005 Taylor & Francis

DOI: 10.1080/15693430500402234
274 S. Wartmann & J. Harnisch

community. For the time being, no environmentally friendly alternatives with adequate
dielectric and/or arc-quenching capabilities have been identified.
In SF6, electrical equipment emissions potentially occur during all life-cycle phases of the
equipment: manufacture, use and decommissioning. All gas handling activities are connected
to potential emissions, e.g., filling, topping up, refilling, testing and gas recovery during
decommissioning. Additional emissions occur through leakage during use. As proper
containment of SF6 is a basic requirement, the equipment can only function properly if no or
only minimal emissions of SF6 exist. Annual leakage rates for new equipment are below 0.5%
depending on the type of equipment. Therefore, the level of emissions over the expected
lifetime of 40 – 50 years depends on the production process as well as on lifetime maintenance
requirements, equipment for SF6 handling and procedures as well as specific design elements
of the SF6 electrical equipment.

2. Approach for emission and bank estimation and projection

For the emission estimation and projection, four years were considered: 1995 as base year or
last year in which awareness about the climatic effect of SF6 had not yet become widespread,
2003 as the most recent year for which a complete dataset was available, as well as 2010 and
2020. For each of the life-cycle-phases a separate emission factor was determined for each of
the years. Emissions in manufacture were calculated based on the SF6 amount used during
the manufacturing process, emissions in use based on the banked amounts of SF6 and
emissions in decommissioning based on the SF6 amount contained in the equipment taken
out of operation.
For a simplified assessment of reduction options, costs and emission development, SF6
electrical equipment was allocated to five equipment categories (category A – E). Figure 1
shows the shares of the categories A – E to 2003 SF6-emissions from electrical equipment in
the EU25þ3.
Emissions from categories D – E contribute less than 13% to the emission total from
categories A – E. As production of category C ceased in Europe in 2000, emissions
from category D are planned to be reduced by over 70% in 2010 and the SF6-technology
for high-voltage power capacitors (category E) will be completely substituted by a SF6-free

Figure 1. Distribution of emissions over categories A – E in 2003 in the EU-25þ3.

 Reductions of SF6 emissions in Europe 275

semi-conductor-technology by 2007, the focus of the study was placed on the sealed and
closed switchgear applications, category A and category B. The latter are briefly described in
the following section.
All switchgear between 1 – 52 kV (medium voltage) in the EU-25þ3 was allocated to
category A ‘sealed pressure equipment’. Sealed pressure equipment requires gas handling
only during manufacture and decommissioning, but not during use. Typical new (state-of-
the-art) medium voltage equipment is of the sealed pressure type, while some older medium
voltage equipment which is not of the sealed pressure type exists. Category A includes gas
insulated circuit breakers and gas insulated switchgear. Sealed pressure equipment has been
in service since the early 1980s. It has a metal enclosure designed not to be opened during the
lifetime of the equipment. The design ensures very low leakage rates during operation—for
new equipment less than 0.1% annually. The average emission rate from the existing bank is
about 0.24% at present, as it also includes older equipment with slightly higher rates. About
80% of the bank has a considerable lower average emission rate of 0.14%.
Any switchgear above 52 kV, i.e., high and extra-high voltage, in the EU-25þ3 was assigned
to category B ‘closed pressure’. Unlike sealed pressure equipment, closed pressure equipment
requires maintenance including gas handling during use. For equipment designed since the
mid-nineties, this is commonly required every 20 to 25 years. The category contains gas
insulated circuit breakers, gas insulated switchgear, gas insulated lines and outdoor gas-
insulated instrument transformers. Design-related leakage rates for new equipment are
commonly far below 0.5% p.a., with an overall emission rate—including design-related
leakage, emissions from small leaks and handling losses—for the existing EU-25þ3 banks of
1.8% in 2003.

3. Bank and emission estimation

Bank data for the major countries were taken mainly from country sales data and other
country-related information available from main suppliers and users of SF6 electrical
equipment, as well as from official reporting, as far as available, and finally harmonized
between experts. Through this approach, the best available quality of data was achieved for
each country, while ensuring consistency of the overall dataset. The most detailed information
was available for more than 75% of the total EU-25þ3 bank. Data for the remaining countries
was derived similarly: partly—particularly for the ten new EU-member countries—data were
also extrapolated using indicators such as population, GDP and electricity demands, but
including whenever possible expert knowledge of the distribution and market introduction of
the SF6 technologies in high and medium voltage.
Decommissioning rates of equipment were estimated based on the lifetime of the
equipment and the distribution of the different equipment generations of sealed and closed
pressure equipment in the existing banks.

3.1. Development of consumption in manufacture

About 70% of switchgear production in high-voltage and 55 % in medium voltage is currently
exported outside the EU-25þ3. The present growth rate of production of 2.5% p.a. is
assumed to last till 2010 and is driven by demands from overseas—mainly the Middle-East
and Asia—and to a certain extent from the eastern European markets. After 2010, it is
assumed that the market demand in the EU-25þ3 will tend towards saturation. Due to
manufacturing sites being located close to the emerging markets, exports are assumed to
stabilize.
276 S. Wartmann & J. Harnisch

3.3. Development of banked amounts

Between 1995 and 2003, an annual increase of 5.8% for the EU-25þ3 can be seen (Figure 2).
The overall growth is mainly driven by sealed pressure equipment with an annual growth
rate of 13% in EU-25þ3. This is explained by the late broad implementation of this
technology in the 1990s compared to closed pressure equipment which was almost two
decades earlier. For 2003 – 2010, an overall annual growth rate of 3.9% for the EU25þ3 is
projected, declining to 1.3% for EU-25þ3 between 2010 and 2020. There is still a higher
market demand of sealed pressure equipment (3.9% p.a.) but this is reaching saturation. For
the first time, closed pressure equipment shows a net reduction of the banked amount of 0.5%
p.a in 2020, due to the increasing replacement of older and bulky closed pressure equipment
from 2010 to 2015 onwards by new very compact equipment containing considerably lower
amounts of SF6.
Decommissioning amounts are predicted to rise strongly from 2010 onwards, when larger
numbers of closed pressure equipment will start to be decommissioned. Closed pressure
equipment will contribute more than 85% of the decommissioned volumes between 2010 and
2020. As most of the sealed pressure equipment has been installed after 1995,
decommissioned quantities remain marginal until 2020.
Emission rates for manufacture, use and decommissioning were determined for 1995 and
2003 through data from industry statistics, published information and national reporting and
were, where national information was not available, extrapolated on the basis of expert
knowledge. Projections for 2010 and 2020 were made by applying reduction options to the
1995 and 2003 status.

4. Reduction of network losses

The use of SF6 technology permits a compact design of equipment and thus allows for an
optimized layout of power networks. High voltage networks can be located closer to the

Figure 2. Bank development 1995 – 2020.

 Reductions of SF6 emissions in Europe 277

consumers of electrical energy, which leads to lower transmission losses. Less energy lost
means also less CO2-emissions from power generation from fossil fuels.
Two recent lifecycle assessment (LCA) studies have compared in great detail net
greenhouse gas emissions from specific transmission and distribution networks based on
electrical equipment with and without SF6 (Krähling 1999, Mersiowsky 2003). It is
commonly discouraged to extrapolate the results of a specific LCA study to an entire sector.
In the absence of better data or models, the results of the two recent LCAs were used in the
study to estimate the contribution of SF6 technology in reducing transmission and
distribution losses of electricity in Europe as general rules of thumb: accordingly, in HV
applications 1 ton of installed SF6 would save 0.8 GWh per year and in MV applications 1 ton
of installed SF6 would save 0.2 GWh per year (applying the EU-15 generation mix value of
0.35 kg CO2/kWh (EURELECTRIC 2004)). Based on this, emission reductions from
decreased transmission and distribution losses for the installed EU-25þ3 banks in 2003 of
5000 tons of SF6 in HV equipment and 2000 tons of SF6 in MV equipment of roughly 1.7 Mt
CO2, annually were obtained. These savings therefore help to diminish the net-climate
contribution of SF6-technology in any power generation system at least partly based on fossil
fuels.

5. Assessment of reduction options and costs

Emission reductions were calculated on an annual basis. All cost figures were based on
current market prices. In case of capital goods, both interest and depreciation were taken into
account. An interest rate of 4% was applied. Depreciation times were chosen according to the
provisions of the International Accounting Standards. Transaction costs (e.g., the cost of
decision preparation and decision making) and external costs were not taken into account.
Reduction costs were calculated from the sum of annualized investment costs and annual
operating and maintenance costs divided by mean annual SF6 savings.
A reference scenario ‘Business As Usual 1995’ (BAU 1995) was set up, assuming that
awareness of the climatic effect of SF6 does not become common between 1995 and 2020,
and that no respective legislation is adopted. As a result, no specific action aiming at the
reduction of SF6 emissions for climate change prevention takes place, but only product
optimization (‘incremental technological change’) and cost reduction driven measures are
assumed to take place.
Reduction options were assessed for all life-cycle-phases. Altogether, ten measures were
considered, including nine measures within the life-cycle and one measure applying
alternative technologies (Table I). The latter included the use of SF6-free technologies in
medium voltage. The measure was restricted to the medium voltage range as fully SF6 free
technologies for the high and extra high voltage range are currently not available.
It was found that reduction measures in manufacture, use-phase and decommissioning
(‘life-cycle measures’) in 2010 offer a total reduction potential of 4.9 Mt CO2-eq.,
amounting to 70% of the annual emissions of the BAU-1995 scenario in 2010. It has to
be considered that this value includes also the reduction potential already accessed between
1995 and 2003 (same applies for 2020). With the measures ‘SF6-free technologies in
medium voltage’ a reduction potential of 1.1 Mt CO2-eq. could be achieved. Only around
50% of this reduction stems from the measure itself, the rest stems from the assumed
relocation of sealed pressure equipment manufacturing plants outside the EU. This was
assumed to happen in connection with the measure, as current export rates for sealed
pressure equipment are around 55%. Costs for relocation have not been considered as part
of the abatement cost.
278 S. Wartmann & J. Harnisch
Table I. Emission reduction measures considered.

Name Life-cycle-phase

Awareness in manufacture—Training, monitoring Manufacture

Centralized supply system Manufacture
Leakage detection with helium Manufacture
Improved Filling Procedure Manufacture
Evacuation to 50 mbar instead of 20 mbar in manufacture Manufacture
Awareness in use-phase—Training, monitoring, labelling Use-phase
Evacuation to 50 mbar instead of 20 mbar in Use-phase Use-phase
Repair and replacement Use-phase
Decommissioning infrastructure Decommissioning
Use of SF6-free technologies in medium voltage All phases

A similar picture was found for 2020. The overall reduction potential through life-cycle
measures is around 6 Mt CO2-eq., amounting to 77% of the BAU-1995 emissions in 2020,
while the measure using SF6-free technologies delivers around 1.2 Mt CO2-eq, with again
only around 50% stemming from the measure itself.
The analyses showed that in 2010 68% of the total reduction potential of 4.9 Mt CO2-eq. is
available at below 2 e/t CO2-eq., 97 % below 6 e/t CO2-eq (Figure 3). The measure ‘SF6-free
technologies in medium voltage’ with costs of 1323 e/t CO2-eq. is far beyond the price range
of the measures in the life-cycle. The picture for 2020 is quite similar to 2010. Around 86% of
the reduction volume is available at below 2 e/t CO2-eq. and around 98 % below 6 e/t CO2-
eq. The SF6-free technologies measure again shows high costs of 2144 e/t CO2-eq.

Figure 3. Emission reduction supply curve in 2010 and 2020.

 Reductions of SF6 emissions in Europe 279

6. Emission development and projections

Considerable emission reductions have already been achieved between 1995 and 2003.
Comparing 1995 and 2003 emissions, with 4.8 and 2.9 Mt CO2-eq. respectively, a 40%
emission reduction can be seen (Figure 4)—despite a bank increase by 60% in the same
timeframe. It is estimated that around 75% of the reduction potential for 2003—calculated
ex-post—has been accessed through voluntary action by the industry. This corresponds to
around 2.1 Mt. CO2-eq., being in the same order of magnitude as the reduction of
transmission losses through the use of SF6-technology in the same year, amounting to 1.7 Mt.
CO2-eq. Expert interviews showed that eight out of nine assessed life-cycle measures had at
least partly been implemented by 2003.
Additionally to the ‘Business as Usual 1995’ scenario, two further emission scenarios for
the years 2003 – 2020 were developed, differing with regard to emission reduction and control
action taken:

. The ‘Business As Usual 2003’ scenario depicts the development of emissions for the case
that no further reduction action takes place after 2003, meaning that reduction options
already implemented in 2003 remain in place, but no additional action is taken.
This scenario is of the same structure as the ‘Business as Usual 1995’ scenario, but with
2003 as reference year. As in the BAU 1995 scenario, product and cost optimization
processes lead to emission reductions to a certain extent.
. The ‘Additional Voluntary Action’ scenario assumes full implementation of all identified
reduction options during the life-cycle until 2020. The measure ‘SF6-free technologies for
medium voltage’ is not included because of its high reduction costs.

Figure 4. Emission development 1995 – 2003 compared to the development in the reference scenario.
280 S. Wartmann & J. Harnisch

Despite the reduction in specific as well as in total emissions achieved between 1995 and
2003, overall emissions in the ‘Business as Usual 2003’ scenario slightly increase around 10%
in 2010 and 2020 compared to 2003 due to the strong bank growth (Figure 5). Compared to
the ‘Business as Usual 1995’ scenario reductions of 55% and 60% are achieved in 2010 and
2020.
Additional reductions in overall emissions can be achieved if the identified reduction
options below 20 e/t CO2-eq. are fully implemented as in the ‘Additional Voluntary Action’
scenario. Compared to the ‘Business as Usual 2003’ scenario, further reductions of 1 Mt
CO2-eq. in 2010 and 1.4 Mt CO2-equ in 2020 are available. Total emissions levels of 2.2 and
1.7 Mt CO2-eq. can be achieved, equalling reductions of 70% and 75% compared to the
respective emission levels in the ‘Business as Usual 1995’ scenario.

7. Summary
The analyses showed that since 1995, a reduction of 40% of SF6 emissions from switchgear
has been achieved through voluntary actions by the European electricity industry. Most of the
potential for emission reductions has already been realized. Remaining reduction options in
the life-cycle—almost all options are available far below the allowance price in the EU-
Emissions Trading scheme of 22 EUR t CO2 (Status 30 September 2005, www.pointcar-
bon.com)—allow additional further reductions of 1 Mt CO2-eq. in 2010 and 1.4 Mt CO2-equ
in 2020.
Furthermore, life-cycle-assessments showed that the use of SF6 in electrical equipment can
reduce the overall CO2-emissions from the electricity systems due to reduced network losses.
In 2003, this effect is estimated to have avoided emissions of about 1.7 million tons of CO2
across the EU-25þ3.

Figure 5. Emission projections 1995 – 2020.

 Reductions of SF6 emissions in Europe 281

Acknowledgements
This article summarizes the findings of the study: S. Wartmann and J. Harnisch ‘Reductions
of SF6 emissions from high and medium voltage electrical equipment in Europe’,
commissioned by the Coordinating Committee for the Associations of Manufacturers of
Industrial Electrical Switchgear and Controlgear in the European Union (CAPIEL), June
2005.

References
European Environmental Agency (EEA). 2004. Annual European Community greenhouse gas inventory 1990 – 2002
and inventory report 2004. Submission to the UNFCCC Secretariat, Technical Report 2/2004.
IPCC Climate Change 1995. 1996. The second IPCC scientific sssessment; intergovernmental Panel on climate
change, world meteorological organization and United Nations Environmental Programme. Cambridge:
Cambridge University Press.
Krähling H, Römer S. 1999. Live cycle assessment study. Electricity supply using SF6 technology.
Mersiowsky I. 2003. Life cycle assessment. SF6-GIS-technology for power distribution—medium voltage. Hannover.