Beruflich Dokumente
Kultur Dokumente
ASSESSMENT
for Power Distribution
– Medium Voltage –
Summary
Solvay
Fluor und Derivate
Life Cycle Assessment
SF6-GIS Technology
for Power Distribution
– Medium-Voltage –
Summary
1 Pages
SF6-GIS Technology for Power Distribution – Medium Voltage –
Introduction
In autumn 2002, the German Ministry of the Environment published a policy paper
on the »Implementation of the National Climate Protection Programme regarding
Fluorinated Greenhouse Gases (HFC, FC, SF6)«. This policy paper addressed,
among other sectors, the use of SF6 in certain electrical switchgear, claiming appli-
cation of SF6 in medium-voltage switchgear to be dispensable and suggesting a
ban of SF6-insulated in favour of air-insulated medium-voltage switchgear.
An earlier life cycle assessment (LCA) on high-voltage (HV) switchgear had shown
power supply systems using SF6-insulated switchgear to be advantageous or at
least competitive from an environmental point of view. The present LCA was com-
missioned by:
This LCA was conducted in accordance with ISO 14040–43 standards and in-
cluded stakeholder participation. The Critical Review according to ISO 14040 was
conducted by TÜV NORD CERT. The independent advisory board comprised:
Page 2
SF6-GIS Technology for Power Distribution – Medium Voltage –
The goal and scope of the present study is to investigate the environmental profile
of power distribution in the medium-voltage range (Figure 1). This is accomplished
by life cycle assessment (LCA) of representative product mixes and grids in the
medium-voltage (MV) range, with a comparison between on the one hand air-
insulated switchgear (AIS) and on the other hand gas-insulated switchgear (GIS).
The prerequisite of applying LCA methods is that the systems under comparison
are functionally equivalent. In this case, the functional unit is the distribution of a
certain amount of electricity on the medium-voltage level during one year. Hence,
the comparison cannot be made switchgear by switchgear, but must rather en-
compass representative product mixes and grid designs.
Figure 1: Simplified scheme of electricity supply – transfer (high-voltage level, 110 kV and above) and distribution
(medium-voltage level, 10/20 kV).
Page 3
SF6-GIS Technology for Power Distribution – Medium Voltage –
Losses of electric power due to ohmic resistance were differentiated into constant
and load-dependent in order to facilitate a subsequent variation of load. The default
load of all switchgear was assumed to be 50 %. This was later on varied between a
minimum of 30 % (rural) and up to 80 % (urban).
Table 2: Investigated switchgear and their electrical data based upon calculations from single-line diagrams. Fig-
ures were aggregated into ZVEI Product Mix according to Table 5.
Abbreviations: PL,50 — ohmic loss at 50 % load; PL,100 — ohmic loss at 100 % load (nominal current); PL∗ — constant ohmic loss
due to voltage transformers. RMU — ring-main unit. S/S — substation.
Ohmic resistance of each distinct part of switchgear were taken from type testing
documents. Single-line diagrams were used to determine actual flows of electric
Page 4
SF6-GIS Technology for Power Distribution – Medium Voltage –
currents and calculate losses according to ohmic losses accordingly (Table 2). Cur-
rent and voltage transformers as well as fuses were appropriately considered.
Apart from electrical data, the survey also comprised amounts of material based
upon disassembly analysis by each equipment manufacturer (Table 3). Amounts of
concrete for buildings were calculated with an empirical formula based upon di-
mensions of the switchgear and required aisles.
Based upon operators’ experiences and expectations, the total lifecycle was as-
sumed to be 30 years for AIS, and 40 years for GIS. This assumption was supple-
mented by a variation between 30 and 50 years.
Table 3: Investigated switchgear and their material data based upon disassembly analysis. Figures were aggre-
gated into ZVEI Product Mix according to Table 5.
The functional unit of power distribution requires that a certain product mix, i. e.
specific numbers of units from each type of switchgear, be considered. Two differ-
ent approaches for obtaining such representative product mixes were employed.
♦ Firstly, an average product mix was calculated based upon figures from a ZVEI
statistical survey of units sold in Germany. This covered the entire range of MV
switchgear equipment for utilities as well as industrial and infrastructure cus-
Page 5
SF6-GIS Technology for Power Distribution – Medium Voltage –
tomers, both 10 and 20 kV. However, this did not allow a consideration of ca-
bles and lines in grids.
♦ Secondly, two sample utility distribution grids were designed, one for an urban
(10 kV, Figure 2) and one for a rural area (20 kV, Figure 3). These grids are
representative of Germany in that their topology was derived from a medium-
sized city. Based upon the grid design study, the lengths of cables and lines
were given (Table 4).
Table 4: Topology and quantity structures of sample urban and rural utility grids [ABB 2003]
Voltage / kV 20 10
Load / MW 25.3 36.1
Grid area / km² ≈ 51 ≈ 12
Load density / MW/km² 0.5 3.0
Ohmic losses per year / GWh/a 1.12 1.52
Page 6
SF6-GIS Technology for Power Distribution – Medium Voltage –
Page 7
SF6-GIS Technology for Power Distribution – Medium Voltage –
For the two sample utility distribution grids, ohmic losses for cables and lines, and
for HV/MV transformers were computed from actual load and considered as con-
stant. Since the presupposition was an identical grid topology in case of AIS and
GIS equipment, this did not impinge upon the comparison. Conversely, medium- to
low-voltage (MV/LV) transformers were by definition beyond the system boundary
of MV power distribution; therefore, their substantial ohmic losses were not in-
cluded.
The two sample grids represent municipal utility structures, i. e. do not comprise
industrial grids connected by consumer substations. Therefore, RMUs are even
more abundant. Aside from the lack of industrial sub-grids, however, the quantity
structures of the two sample grids are sufficiently comparable with the ZVEI prod-
uct mix described above. Table 5 provides a comparative overview of all medium-
voltage product mixes thus obtained.
Table 5: Investigated product mixes for medium-voltage switchgear. ZVEI Product Mix is derived from ZVEI sta-
tistical survey of medium-voltage units sold in Germany. Figures are units, normalised with reference to
one double-busbar transformer substation 24 kV. Figures for urban and rural utility grids are derived
from representative topology and grid planning, cf. Figure 2 and Figure 3, respectively [ABB 2003].
Types of switchgear are specified in Table 1.
Type of Switchgear ZVEI Product Mix Urban Grid (Utility) Rural Grid (Utility)
The life cycle inventory covered acquisition of raw materials, pre-production of all
relevant materials, manufacture and operation of switchgear, and electricity gen-
eration. SF6 emissions from GIS included losses during manufacture, operation,
and end-of-life management. Discounted SF6 emissions during manufacture and
operation were calculated to be 0.14 % p. a., based upon ZVEI surveys. Losses
during end-of-life management were assumed to be 2 %, provided that SF6 ReUse
is employed. This figure was varied across a wide range.
Page 8
SF6-GIS Technology for Power Distribution – Medium Voltage –
Moreover, parameter variations and scenario analyses covered the following rele-
vant factors: SF6 emission rates, length of switchgear use phase, load of switch-
gear, shares of AIS and GIS equipment in grids, and electricity generation mix. The
latter scenarios addressed varying CO2 loads per kWh due to different national
electricity generation mixes.
First, the results are presented on switchgear level, i. e. not considering other grid
components, such as cables, lines, and transformers. This allows for a more de-
tailed comparison. In a subsequent step, however, a dominance analysis will show
the rather minor contribution of switchgear compared with those other grid compo-
nents. This will put the results on switchgear level into perspective of the grid level.
All results for environmental impact categories are shown as relative figures, nor-
malised to the AIS reference system. The results obtained for global warming po-
tential (GWP) as the critical impact category – with the smallest differences be-
tween AIS and GIS systems – are commented individually. Results for the other
impact categories – primary energy demand, acidification potential (AP), and nutri-
fication potential (NP) – show even larger advantages for the GIS systems.
Page 9
SF6-GIS Technology for Power Distribution – Medium Voltage –
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
ZVEI AIS ZVEI GIS ZVEI AIS ZVEI GIS
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
ZVEI AIS ZVEI GIS ZVEI AIS ZVEI GIS
Figure 4: Results for environmental impact categories calculated for the ZVEI Product Mix (switchgear level, i. e.
other grid components, such as cables, lines, and transformers were not considered). Relative figures
normalised to the reference system AIS = 100 %.
Page 10
SF6-GIS Technology for Power Distribution – Medium Voltage –
Table 6: Results for environmental impact category indicators calculated for the ZVEI Product Mix
Materials and Ohmic Losses Ohmic Losses Ohmic Losses SF6 Emissions of
Manufacture of of Transformer of Ring-Main Units of Customer Switchgear
all Switchgear Substations (RMU) Substations
Primary Energy Demand [MJ/a]
AIS 443,625 263,450 872,130 234,700 —
GIS 135,721 202,000 595,680 234,160 —
Global Warming Potential GWP100 [kg CO2-Äq./a]
AIS 37,855 14,888 49,284 13,263 0
GIS 12,029 11,415 33,662 13,232 19,556
Acidification Potential AP [kg SO2-Äq./a]
AIS 456.0 18.3 60.7 16.3 —
GIS 130.6 14.1 41.4 16.3 —
Nutrification Potential NP [kg PO4-Äq./a]
AIS 12.59 2.80 9.26 2.49 —
GIS 3.37 2.14 6.32 2.49 —
Urban Grid
Since the topology of the urban grid remains identical in case of AIS or GIS equip-
ment (cf. Figure 2 and Table 4), only the results for the underlying switchgear
product mix are given here; the influence of other grid components is shown in the
dominance analysis below.
The global warming potential GWP100 calculated for the urban grid amounts to
1,220 tonnes of CO2 eq./a with AIS switchgear and approx. 1,190 tonnes
of CO2 eq./a with GIS switchgear. However, ohmic losses of cables, lines and
MV/LV transformers are identical in both cases and account for 91 % (AIS) and
92 % (GIS) of these figures, respectively. Whereas the contribution of switchgear is
only 10.4 % (AIS) and 7.4 % (GIS), respectively. With GIS switchgear, SF6 emis-
sions contribute 1.4 % to the total GWP100 of the system. Using GIS switchgear,
the advantage regarding GWP100 is 3.2 %, as compared to AIS. Figure 5 shows
only the comparison by switchgear, excluding the dominant and identical contribu-
tion by cables, lines, and HV/MV transformers. Results for all environmental impact
categories in absolute figures calculated for the urban grid are shown in Table 7.
Page 11
SF6-GIS Technology for Power Distribution – Medium Voltage –
100%
100%
80%
80%
60%
60%
40% 40%
20% 20%
0% 0%
Urban AIS Urban GIS Urban AIS Urban GIS
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
Urban AIS Urban GIS Urban AIS Urban GIS
Figure 5: Results for environmental impact categories calculated for the urban grid (switchgear level, i. e. other
grid components, such as cables, lines, and transformers were not considered). Relative figures normal-
ised to the reference system AIS = 100 %.
Page 12
SF6-GIS Technology for Power Distribution – Medium Voltage –
Table 7: Results for environmental impact category indicators calculated for the urban grid
Rural Grid
Since the topology of the rural grid remains identical in case of AIS or GIS equip-
ment (cf. Figure 3 and Table 4), only the results for the underlying switchgear
product mix are given here; the influence of other grid components is shown in the
dominance analysis below.
The global warming potential GWP100 calculated for the rural grid amounts to
930 tonnes of CO2 eq./a with AIS switchgear and approx. 890 tonnes of CO2 eq./a
with GIS switchgear. However, ohmic losses of cables, lines and HV/MV trans-
formers are identical in both cases and account for 87 % (AIS) and 91 % (GIS) of
these figures, respectively. Whereas the contribution of switchgear is only 12.6 %
(AIS) and 8.7 % (GIS), respectively. With GIS switchgear, SF6 emissions contribute
1.6 % to the total GWP100 of the system. Using GIS switchgear, the advantage re-
garding GWP100 is 4.2 %, as compared to AIS. Figure 6 shows only the comparison
by switchgear, excluding the dominant and identical contribution by cables, lines,
and MV/LV transformers. Results for all environmental impact categories in abso-
lute figures calculated for the rural grid are shown in Table 8.
Page 13
SF6-GIS Technology for Power Distribution – Medium Voltage –
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
Rural AIS Rural GIS Rural AIS Rural GIS
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
Rural AIS Rural GIS Rural AIS Rural GIS
Figure 6: Results for environmental impact categories calculated for the rural grid (switchgear level, i. e. other grid
components, such as cables, lines, and transformers were not considered). Relative figures normalised
to the reference system AIS = 100 %.
Page 14
SF6-GIS Technology for Power Distribution – Medium Voltage –
Table 8: Results for environmental impact category indicators calculated for the rural grid
Figure 7 illustrates this relation for the urban grid; the results for the rural grid are
comparable. Furthermore, this shows that RMU in turn constitute the predominant
contribution among switchgear, i. e. if cables, lines and transformers are ignored.
Moreover, in case of GIS equipment, the contribution of SF6 emissions to the GWP
of these sample grids is only 1.4 % (urban) and 1.6 % (rural), respectively. This
confirms that with respect to climate protection, restrictions on SF6 emissions from
switchgear would reduce total GWP only insignificantly.
Page 15
SF6-GIS Technology for Power Distribution – Medium Voltage –
Transformer Stations
1%
Others
10% Ring-main Units (RMU)
9%
Transformer Stations
1%
Others
8%
7%
Ring-main Units (RMU)
7%
Figure 7: Dominance analysis for sample urban utility grids using AIS or GIS equipment, respectively. Contribu-
tions of grid components to global warming potential (GWP100) are shown. Grid components other than
switchgear include cables, lines, and transformers. (Totals may differ from 100 % due to rounding im-
precision.)
Page 16
SF6-GIS Technology for Power Distribution – Medium Voltage –
On grid level, being relevant for planning decisions, the predominant contribution to
climate impact (GWP) of power distribution grids are ohmic losses from compo-
nents other than switchgear. It was shown that restrictions on a particular switch-
gear technology, and SF6-GIS in particular, would fail to be environmentally effec-
tive. On the contrary, GIS technology permits locating the RMU close to consum-
ers, thus further minimising substantial losses arising from MV/LV transformers. In
a conservative approach, this effect has not been quantified in this LCA.
In the following, the major factors determining the results of this LCA are discussed
as for their influence (derived from parameter variation and scenario analysis):
♦ Since ohmic losses are the single most relevant factor determining the GWP of
the distribution grid, the specific CO2 load of the electricity mix plays a signifi-
cant role. However, variations of the specific CO2 load of the electricity mix are
fairly limited, unless energy generation undergoes a fundamental change. For
instance, a realistically moderate increase of either wind energy or power pur-
chases from other European countries does not substantially influence the re-
sults of this LCA. Conversely, a fundamental change in energy generation (for
instance, with decentralised power plants) would also imply completely different
transmission and distribution grids.
Page 17
SF6-GIS Technology for Power Distribution – Medium Voltage –
Transferability of Results
The basic results of this LCA apply to industrial grids as well. Apart from topology,
the main difference is that RMUs are the most prevalent type of switchgear in utility
grids; whereas in industrial grids, there are larger numbers of transformer and cus-
tomer substations. Hence the advantages of GIS technology that were shown to be
greater for RMUs than for substations will make less of a difference. Since ohmic
losses of other components are predominant, this will not substantially affect the
expected results for industrial grids. On the contrary, the load is rather higher in in-
dustrial grids, thus shifting the balance further in favour of GIS equipment.
With a similar rationale, the fundamental conclusions of this LCA can also be ap-
plied to other European countries. As ohmic losses of other components are the
major influence on GWP, the specific national product mix will not affect the out-
come very much. This was shown by a scenario analysis (Figure 8); using the
ZVEI Product Mix, the resulting advantages in GWP are shown in relation to the
AIS reference scenario. Even for countries with a lower CO2 load of the electricity
generation mix – due to either nuclear or renewable energy sources – the relative
GWP advantage of employing GIS equipment remains greater.
100%
Relative GWP Advantage vs. Basis Scenario AIS [%]
80%
60%
40%
Electricity Mix
Czech Rep.
20% Electricity Mix
France
0%
-20%
Electricity Electricity Mix
Mix Norway Germany/Wind +10%
-40%
-100%
0,0 0,2 0,4 0,6 0,8 1,0 1,2
CO2 Load of Electrity Mix [kg CO2 eq./kWh]
Figure 8: Scenario analysis examining the influence of the CO2 load of the electricity generation mix on results for
global warming potential GWP100 – calculated for ZVEI Product Mix
Page 18
SF6-GIS Technology for Power Distribution – Medium Voltage –
maining identical for AIS and GIS, the differences in grid design in various coun-
tries will not have a substantial bearing on overall results.
When assessing the environmental profile of power distribution, the system ap-
proach, implying a consideration on grid level, is indispensable to obtain meaning-
ful results. Consequently, if ohmic losses from cables, lines, and transformers are
taken into consideration, the differences between MV switchgear technologies AIS
and GIS become all but negligible. This life cycle assessment (LCA) indicates that
electrical power losses due to ohmic resistance of cables, lines, and HV/MV trans-
formers constitute the predominant contribution to the GWP of distribution grids.
Page 19
SF6-GIS Technology for Power Distribution – Medium Voltage –
♦ The results of this LCA may in principle be transferred to other European coun-
tries. The primary energy mix for electricity supply, being the most important re-
gional factor, showed only a minor influence on the results.
References
The full report of this LCA study is available on request – please contact the au-
thor: Dr.-Ing. Ivo Mersiowsky, Solvay Management Support
1
Document DG ENV.C.2/JD D(2003)421117 dated 12/06/2003.
Page 20
Validity Statement by TÜV NORD CERT