Journal of Cleaner Production 106 (2015) 565e575

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Life cycle assessment of electricity generation in Mauritius

Ravina Brizmohun a, b, Toolseeram Ramjeawon a, Adisa Azapagic b, *
a
Faculty of Engineering, University of Mauritius, R
eduit, Mauritius
b
School of Chemical Engineering and Analytical Science, The Mill, Sackville Street, The University of Manchester, Manchester M13 9PL, UK

a r t i c l e i n f o a b s t r a c t

Article history: Electricity demand in Mauritius is growing rapidly but its environmental implications are as yet un-
Received 15 October 2013 known. This is the topic of the current paper which presents for the rst time the life cycle environ-
Received in revised form mental impacts of electricity generation in Mauritius aiming to inform electricity generators and policy
8 November 2014
makers on how the impacts could be reduced. The majority of country's electricity is generated from
Accepted 9 November 2014
Available online 15 November 2014
fossil fuels, with coal contributing 40% and fuel oil 37%; the rest is from sugarcane bagasse (19%) and
hydro-power (4%). The results suggest that electricity from oil has the highest impacts for six out of ten
categories considered compared to the other three sources: acidication, freshwater, terrestrial and
Keywords:
Electricity generation
human toxicity, ozone layer depletion and photochemical oxidants. The remaining four impacts
Life cycle assessment (depletion of resources, global warming, eutrophication and marine toxicity) are highest for coal. The
Environmental impacts lowest impacts are found for electricity from hydro-power. For example, the global warming potential
Mauritius (GWP) of electricity from coal is estimated at 1444 kg CO2 eq./MWh and for oil 754 kg CO2 eq./MWh,
while for bagasse and hydro-power this impact is several orders of magnitude lower (29 and 8.6 kg CO2
eq./MWh, respectively). Oil and coal are the main contributors to the overall impacts from electricity in
Mauritius (88%e99%). The contribution of bagasse is small (<1%e12%) and that from hydro-power
negligible (<0.1%). The GWP of the electricity mix is estimated at 868 kg CO2 eq./MWh. This is equiva-
lent to the annual GWP of 2.22 Mt CO2 eq. in 2012, an increase of 16% since 2007. To reduce its carbon
emissions, Mauritius should consider reducing the share of fossil fuels through increased use of re-
newables such as solar PV and wind as well as improving the efciency of the fossil power plants and
reducing energy demand.
2014 Elsevier Ltd. All rights reserved.

1. Introduction As illustrated in Fig. 2, the main sources of electricity in

Global energy consumption is projected to double by 2040
compared to the demand in 2010, with the largest increase ex-
pected in the developing world (EIA, 2013). Being a rapidly devel-
oping country, energy demand in Mauritius is also growing, in
particular its electricity consumption, which increased by 66% in
the period from 2001 to 2012 (see Fig. 1). The largest consumer of
electricity in Mauritius is the commercial sector, including tourism,
information and communication technologies, nancial services
and real estate (36% in 2012), followed by households (33%) and
industry (30%) (CEB, 2013; CSO, 2012). It is predicted that, with the
growth and diversication of the economy, the electricity demand
will continue to increase (CEB, 2013).
* Corresponding author. Tel.: 44 (0) 161 306 4363.
E-mail address: adisa.azapagic@manchester.ac.uk (A. Azapagic).
Mauritius are coal (~40%), fuel oil (~37%), sugarcane bagasse (~19%)
and hydro (~3.6%), with their contribution remaining pretty con-
stant since 2007. Around 60% of national demand is supplied by the
Central Electricity Board (CEB, 2008), the sole national electricity
supply corporation responsible for electricity generation, trans-
mission and distribution. The rest is provided by Independent Po-
wer Producers (IPPs) who are associated with the sugar sector and
use bagasse to generate electricity during sugarcane crop season.
Given the high contribution of fossil fuels to the electricity mix
(~77%) and the expected future demand growth (CEB, 2013), it is
important to understand environmental implications of different
electricity options for Mauritius to help identify sustainable path-
ways for meeting the demand. However, although the Mauritius
Government has pledged to reduce the environmental impacts and
particularly greenhouse gas (GHG) emissions from the energy
sector by 30% by 2025 by increasing the share of renewables to 35%
(GOM, 2007; MREPU, 2009), the full environmental implications of
current electricity generation are unknown.

http://dx.doi.org/10.1016/j.jclepro.2014.11.033
0959-6526/ 2014 Elsevier Ltd. All rights reserved.
566 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575

that, the following section details the methodology used in the

study, together with the data and the assumptions. This is followed
by the discussion of results in Section 3 and conclusions in Section
4.

2. Methods

The LCA methodology applied in this work follows the ISO

14040 and 14044 standards (ISO, 2006a; b). The LCA software
SimaPro 7 (PRe  Consultants, 2008) has been used to model the
system. The impacts have been estimated according to the CML 2
Fig. 1. Electricity consumption in Mauritius from 2001 to 2012 (based on data from Baseline 2001 method (Guine e et al., 2001). This is a problem-
CSO (2012)). oriented method, often referred to as a midpoint approach,
because it considers environmental burdens at an intermediate
point between the point of intervention (extraction of resources
Therefore, this paper sets out to estimate the life cycle envi-
or emissions to the environment) and the ultimate damage caused
ronmental impacts of electricity in Mauritius and provide baseline
by that intervention. This method has been chosen here as one of
data for monitoring progress (or otherwise) in reducing the impacts
the most commonly used to allow comparison of our results with
from the sector. Although many life cycle assessment (LCA) studies
other relevant studies.
of electricity have been carried out for different countries, as far as
The impact categories included in the CML method and
we are aware, there are no studies of electricity generation in
considered in this study are: global warming potential (GWP),
Mauritius. The only exception is a study by Ramjeawon (2008) but
abiotic depletion potential (ADP), acidication potential (AP),
that only considered electricity from bagasse rather than the whole
eutrophication potential (EP), human toxicity potential (HTP),
electricity system.
freshwater aquatic ecotoxicity potential (FAETP), marine aquatic
The LCA studies of electricity generation for other countries
ecotoxicity potential (MAETP), ozone layer depletion (ODP),
include those for Japan (Hondo, 2005), Singapore (Kannan et al.,
photochemical oxidants creation potential (POCP) and terrestrial
2007), Belgium and Spain (Foidart et al., 2010), Mexico (Santoyo-
ecotoxicity potential (TETP). The impacts are calculated based on
Castelazo et al., 2011), Nigeria (Gujba et al., 2011), the UK
the global impact (characterisation) factors for most categories,
(Stamford and Azapagic, 2012) and Turkey (Atilgan and Azapagic,
except for the AP and POCP which refer to European conditions.
2015), to name but a few. Many studies have also been carried
Although the latter is a limitation, no other method relates these
out for individual electricity technologies as opposed to an elec-
impacts specically to Mauritius or Africa.
tricity mix. Those relevant to the technologies present in the
Mauritius electricity mix include studies of coal and oil by Hondo
(2005), Santoyo-Castelazo et al. (2011) and Peiu (2007); sugar- 2.1. Goal and scope denition
cane bagasse by Ramjeawon (2008), Renouf et al. (2011) and Lopes
Silva et al. (2012); and hydro-power by Gagnon et al. (2002), The goal of the study is to estimate the life cycle environmental
Go ralczyk (2003) and Suwanit and Gheewala (2011). The results impacts of electricity generation in Mauritius. As illustrated in
obtained in some of these studies are compared to the results ob- Fig. 3, the scope of the study is from cradle to consumer, consid-
tained in the current study later in the paper (Section 3.2). Prior to ering the following stages:

Fig. 2. Electricity mix in Mauritius over the period 2007e2012.

Source: CSO (2012).
 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575 567

Fig. 3. The life cycle of electricity generation in Mauritius.

 extraction and production of fuels; 2.2. Data quality and assumptions

 electricity generation;
 electricity transmission and distribution to the point of use;
 all transport steps in the system; and
 all waste management steps except for decommissioning of
power plants.
Construction of coal and oil power plants is excluded from the
system boundaries as its contribution is typically small over their
lifetime; this is also the reason why decommissioning is not
considered. Furthermore, construction of the plants generating
electricity from bagasse is also excluded because the main purpose
of the plant is to produce sugar, with electricity being a co-product.
However, construction of hydro-power plants is considered
because all the impacts are related to this life cycle stage
(Ecoinvent, 2007); for consistency with the other types of plant,
decommissioning is not included.
The study is based on the functional unit dened as 1 MWh of
electricity delivered to the consumer.
To ensure the highest-quality data are used in the study, the
following guiding principles were followed in selecting the
data:
i. Age: The base year chosen for this study is 2007 as the data
available were most complete for that year. In any case, as
can be seen in Fig. 2, the electricity mix in Mauritius has
remained approximately constant from 2007 to 2012.
Moreover, there have been no major changes in the type of
fuel and technology over this period.
ii. Source: Primary (foreground) data for different electricity
technologies and the grid were collected from industry and
government, including from the Central Electricity Board
(CEB, 2007a,b; 2008) and the Central Statistics Ofce (CSO,
2008, 2011). The background life cycle inventory (LCI) data
were sourced from the Ecoinvent V2 database (Ecoinvent,

Table 1
Data and assumptions for electricity from coal in the base year (2007).

Life cycle stage Data and assumptions Source

Coal mining and processing Coal type and origin: Hard coal from South Africa MoE (2009)
Background data: Ecoinvent Ecoinvent (2007)
Transport Coal shipping from South Africa to the coal storage area in the harbour in Mauritius by S. Sookhraz, CEB, Pers. comm.,
43,000 t tankers 7 April 2010
Average distance from South Africa to Mauritius: 2827 km Own estimate
Coal transport from storage to power plant by 40 t diesel lorries Golder Asssociates (1997)
Total distance between all coal power plants and their coal storage areas: 155 km Own estimate
Power generation Number of plants: 5 CEB (2008)
Installed capacities: 25, 35, 37, 70 and 87 MW CEB (2008)
Total annual generation: 993.6 GWh/yr CSO (2008)
Amount of coal: 552,632 t/yr CSO (2008)
Net caloric value: 25 MJ/kg MoE (2009)
Sulphur content: 0.62% MoE (2009)
Ash content: 14%; 1/3rd y ash (25,789 t/yr); 2/3rd bottom ash (51,560 t/yr) MoE (2009)
Average power plant efciency: 25.8% CEB (2008)
Air emissions (CO2, SO2, CO, CH4, N2O, NOx, NMVOCa) calculated using IPCC factors (no IPCC (2006)
emission controls are used)
Background data: Ecoinvent Ecoinvent (2007)
Transmission and distribution Electricity losses: 9.7%b CEB (2007)
a
Non-methane volatile organic compounds.
b
Across the electricity grid.
568 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575

Table 2
Data and assumptions for electricity from fuel oil in the base year (2007).

Life cycle stage Data and assumptions Source

Fuel renery Fuel oil origin: India CSO (2008)

Background data: Ecoinvent Ecoinvent (2007)
Transportation Fuel oil shipped to Mauritius by 30,000 t tankers S. Sookhraz, CEB, Pers. comm., 7 April 2010
Shipping distance from India to Mauritius: 4650 km Own estimate
Transport within Mauritius to power plants by pipelines Ecoinvent (2007)
Background data: Ecoinvent Ecoinvent (2007)
Power generation Number of plants: 3 CSO (2008)
Installed capacities: 44, 113 and 138 MW CSO (2008)
Total annual generation: 885 GWh/yr CSO (2008)
Amount of fuel oil: 186,403 t/yr CSO (2008)
Net caloric value: 40 MJ/kg Mott MacDonald (2003)
Sulphur content: 2.4% and 3.0%a Mott MacDonald (2003)
Average power plant efciency: 42.7% CEB (2008)
Air emissions (CO2, SO2, CO, CH4, N2O, NOx, NMVOCb) IPCC (2006)
calculated using IPCC factors (no emission controls are used)
Background data: Ecoinvent Ecoinvent (2007)
Transmission and distribution Electricity losses: 9.7%c CEB (2007)
a
Two types of fuel oil are used with different sulphur content.
b
Non-methane volatile organic compounds.
c
Across the electricity grid.

Table 3
Data and assumptions for electricity from bagasse in the base year (2007).

Life cycle stage Assumptions Source

Sugarcane cultivation 1 ha 64.9 t of sugarcane 6.3 t of sugar MSIRI (2007)

and harvesting Cultivation area: 69,000 ha; with a seven-year planting cycle MSIRI (2007)
(based on 1 ha) Machinery usage for eld preparation: 3.35 h/ha (123 kW capacity) Cahoolessur (2002)
Diesel consumption by machinery: 7.84 l/ha Cahoolessur (2002)
Background data for energy used by machinery: Ecoinvent Ecoinvent (2007)
Irrigation water: 12,000 m3/ha.yr Balloo, Pers. comm., 12 August 2008
Average electricity usage for irrigation: 243 kWh/ha.yr Balloo, Pers. comm., 12 August 2008
Amount of herbicides used: 7.8 kg/ha (active ingredients) Agreco Consortium (2007)
Amount of fertilisers used: 138 kg N/ha, 50 kg P2O5/ha and 175 kg K2O/ha Ramjeawon (2008)
Machinery usage for fertiliser spreading: 0.91 h/ha (90 kW capacity) Cahoolessur (2002)
Background data for energy used by machinery: Ecoinvent Ecoinvent (2007)
Fertiliser surface runoffs: 0.7 kg P/ha, 5.5 kg NO3/ha Agreco Consortium (2007)
Sugarcane area burnt: 1608 ha/yr (2.33% of cane area). MSIRI (2007)
Machinery usage for harvesting: 0.19 h/ha Cahoolessur (2002)
Usage of mechanical loader: 0.92 h/ha Cahoolessur (2002)
Emissions estimated using the US EPA methodology EPA (1995)
Transport Average distance from sugarcane eld to factory: 7 km Own estimate
Average distance for transport of fertilisers: 20 km Own estimate
All transport by 16 t trucks with fuel consumption of 0.075 L/t.km Ecoinvent (2007)
Sugar processing and Amount of sugarcane crushed: 4.24 Mt/yr MSIRI (2007)
electricity generation Amount of sugar produced: 435,972 t (0.103 t sugar/t of sugarcane) MSIRI (2007)
Amount of bagasse produced: 1.44 Mt MSIRI (2007)
(0.3 t bagasse/t of sugarcane)
Amount of molasses produced: 0.27 t/t of sugar MSIRI (2007)
(0.03 t of molasses/t of cane)
Bagasse composition: 50% bre, 48% water and 2% sugars Ramjeawon (2008)
Net caloric value of bagasse at 48% moisture: 8 MJ/kg. Ramjeawon (2008)
Amount of bagasse used for electricity generation: 1.052 Mt/yr CSO (2008)
Water usage: 9.6 t/t of sugarcane (most of which is recycled) Cahoolessur (2002)
Wastewater generated: 10.2 kg/t sugarcane MSIRI (2007)
Wastewater pollution: COD: 2.068 kg/t sugarcane; BOD: 0.715 kg/t sugarcane; Cahoolessur (2002)
TSS: 1.365 kg/t sugarcanea
Slag and ash: 3.72 kg/t of sugarcane Cahoolessur (2002)
Electricity and steam generated: 110 kWhel/t sugarcane and 500 kg of CEB (2008)
steam/t sugarcane
Economic allocation at the sugar processing plant: 34.9% for bagasse, Own estimate
53.9% for sugar and 11.2% for molasses (based on the commercial value based on data from Ramjeawon (2008)
of Rs14,000/t of sugar, Rs7/kWh for bagasseb and Rs1000/t of molasses
Mass allocation at power plant: 76% for process steam and electricity, Ramjeawon (2008)
24% for electricity exported to the grid
Transmission and distribution Electricity losses: 9.7%c CEB (2007)
a
COD: chemical oxygen demand; BOD: biological oxygen demand; TSS: total suspended solids.
b
The sugar renery is paid per kWh generated from bagasse rather than per tonne of bagasse produced.
c
Across the electricity grid.
 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575
Table 4
Data and assumptions for electricity from hydro-plants in the base year (2007).
Type of plant
Reservoir
Power generation
Run-of-river
Power generation
Transmission and distribution
a
Across the electricity grid.
2007). The emissions from the coal and oil power plants
were calculated following the IPCC methodology (IPCC,
2006), based on fuel composition and plant operation data.
iii. Geographical origin: The primary data for the fuels and
technologies are specic to Mauritius. The background LCI
data sourced from Ecoinvent have been adapted to the
Mauritius conditions as far as possible (e.g. by changing the
technology efciency, scale, etc.).
iv. Completeness: Most data were available either from a spe-
cic source or have been calculated as mentioned under ii.
v. Consistency and reliability: To ensure consistency, data were
sourced from a select number of reliable sources, such as
industry, government and Ecoinvent (as mentioned under ii.).
Therefore, the overall data quality is deemed to be high. More
detail on the data and assumptions for each source of electricity is
given in the next sections.
2.2.1. Electricity from coal and fuel oil
The data and assumptions for electricity from hard coal and fuel
oil are summarised in Table 1 and Table 2, respectively. There are
ve coal plants in Mauritius with the installed capacity of 25, 35, 37,
70 and 87 MW and an average electrical efciency of 25.8% (CEB,
2008). Coal is imported from South Africa and, after being ship-
ped to Mauritius, it is transported by road to the power plants
(Table 1). Around 552.6 kt of coal were used in the base year (2007)
to generate 993.6 GWh of electricity (CSO, 2008).
There are three fuel oil power plants in Mauritius with the
installed capacity of 44, 113 and 138 MW. In the base year, 885 GWh
were generated from 186.4 kt of fuel oil (CSO, 2008) which is im-
ported from India (Table 2).
2.2.2. Electricity from bagasse
Mauritius produces around 436 kt of sugar annually from
4.24 Mt of sugarcane (MSIRI, 2007). Bagasse is one of the by-
products of sugarcane processing (together with molasses) which
sugar producers use to generate energy for the plant and the excess
electricity is sold to the grid (see Table 3). As illustrated in Fig. 3, the
following life cycle stages are considered in the study:
 sugarcane cultivation and harvesting;
 transport of sugarcane to sugar reneries;
 sugarcane processing to produce sugar and co-products
(molasses and bagasse);
 electricity generation from bagasse; and
 electricity transmission and distribution to the user.
569
Data and assumptions Source
Number of plants: 4 CEB (2008)
Installed capacity: 4, 10, 11.1 and 30 MW CEB (2008)
Total annual production: 75.6 GWh/yr CSO (2008)
Average capacity of plants in Ecoinvent: 176 MW Ecoinvent (2007)
Efciency: 78% Ecoinvent (2007)
Lifetime: 150 years Ecoinvent (2007)
Background data: Ecoinvent Ecoinvent (2007)
Number of plants: 4 CEB (2008)
Installed capacity: 0.9, 1, 1.2 and 1.2 MW CEB (2008)
Total annual production: 8.2 GWh/yr CSO (2008)
Average capacity of plants in Ecoinvent: 49 MW Ecoinvent (2007)
Efciency: 82% Ecoinvent (2007)
Lifetime: 80 years Ecoinvent (2007)
Background data: Ecoinvent Ecoinvent (2007)
Electricity losses: 9.7%a CEB (2007)
At the sugar processing stage, the impacts have been allocated
between sugar and the co-products using an economic basis as
detailed in Table 3. System expansion was not feasible as the co-
products cannot be produced in an alternative system. Mass and
energy allocation were also inappropriate because the relative
masses and energy contents of sugar, molasses and bagasse are
disproportionate and do not reect the main reason for the exis-
tence of the plant, which is sugar production. At the power plant,
the allocation between the electricity and heat used for sugar
production and the electricity exported to the grid was carried out
in proportion to their respective amounts (see Table 3), following
the approach in Ramjeawon (2008).
2.2.3. Electricity from hydro-plants
There are eight hydro-plants in Mauritius, half of which are
reservoir and the other half run-of-river plants. The installed ca-
pacities for the former four are 4, 10, 11.1 and 30 MW and for the
latter four 0.9, 1, 1.2 and 1.2 MW (CEB, 2008). In the base year they
generated 75.6 MWh and 8.2 MWh, respectively.
Since primary data for hydro-plants were not available, Ecoin-
vent data have been used but adapted to reect the capacity of the
Mauritius plants. For these purposes, the Ecoinvent data for the two
types of plant were scaled based on the economies of scale method
used for scaling the capital costs of process plants as follows
(Sinnott, 1996):
C2 C1  c2 =c1 0:6 (1)
where C1 and C2 are costs of the larger and smaller plant respec-
tively, here representing the materials and energy being used for to
construct the plant; c1 and c2 are capacities of the larger and
smaller plant, respectively; 0.6 is the economy of scale factor. This
approach is more appropriate than the linear scaling of equipment
normally applied in LCA, particularly because the impacts from
hydro-power are largely related to power plant construction. A
similar approach has been used in other LCA studies (e.g. Greening
and Azapagic, 2012; Whiting and Azapagic, 2014).
To calculate the impacts of electricity generation, the lifetime
for the reservoir plants is assumed to be 150 years and for the
run-of-river installations 80 years (Ecoinvent, 2007).
3. Results and discussion
The results are presented rst for each electricity source and
then for the overall Mauritius electricity mix.
570 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575

Fig. 4. Environmental impacts per MWh by electricity source. [The values have been scaled to t. To obtain the original values, multiply the values shown in the graph with the
factors shown on the x-axis against relevant impacts. GWP: Global warming potential; ADP: Abiotic depletion potential; AP: Acidication potential; EP: Eutrophication potential;
FAETP: Fresh water aquatic ecotoxicity potential; HTP: Human toxicity potential; MAETP: Marine aquatic ecotoxicity potential; ODP: Ozone layer depletion potential; POCP:
Photochemical oxidants creation potential; TETP: Terrestrial ecotoxicity potential].

3.1. Environmental impacts by electricity source
The environmental impacts for each electricity source in Fig. 4
suggest that coal and oil have the highest and hydro-power the
lowest impacts per MWh of electricity produced. For example, at
1444 kg CO2 eq./MWh, coal is the worst option for the GWP. By
contrast, the GWP of fuel oil is around half that value (754 kg CO2
eq.) while for bagasse and hydro-power this impact is several or-
ders of magnitude lower (29 and 8.6 kg CO2 eq., respectively).
In addition to the GWP, coal has three other impacts higher than
any other option considered: the ADP with around 10 kg Sb eq./
MWh, the EP with 661 g PO4 eq./MWh and the MAETP at 657 t DCB
eq./MWh. Again, these are several times lower for bagasse and
hydro-electricity (Fig. 4). However, for electricity from fuel oil, six
impacts are the highest: the AP (29.2 kg SO2 eq./MWh), FAETP
(26.7 kg DCB eq./MWh), HTP (297 kg DCB eq./MWh), ODP
(102.8 mg CFC-11 eq./MWh), POCP (1.17 kg C2H4 eq./MWh) and
TETP (6.17 kg DCB eq./MWh).
For both coal and oil power plants, the majority of the impacts
are from combustion of the fuels in the power plants (see Fig. 5
and Fig. 6). The exceptions are the ADP which is almost entirely
due to fuel extraction and the ODP which in the case of coal is
largely (81%) due to the emissions of halons and CFCs in the life
cycle of transport and in the case of oil due to bromo-tri-uoro-
methane released during oil production (99%). Another excep-
tion is FAETP for coal electricity, of which more than a half (54%)
is due to coal mining and emissions of metals such as vanadium,
nickel and beryllium to fresh water; the rest is due to water
emissions of barite and barium related to the life cycle of coal
transport.

Fig. 5. Contribution to impacts of different life cycle stages for electricity from coal. [The values have been scaled to t. To obtain the original values, multiply the values shown in
the graph with the factors shown on the x-axis against relevant impacts. For impacts nomenclature, see Fig. 4].
 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575 571

Fig. 6. Contribution to impacts of different life cycle stages for electricity from fuel oil. [The values have been scaled to t. To obtain the original values, multiply the values shown in
the graph with the factors shown on the x-axis against relevant impacts. For impacts nomenclature, see Fig. 4].

Fig. 7. Contribution to impacts of different life cycle stages for electricity from bagasse. [The values have been scaled to t. To obtain the original values, multiply the values shown
in the graph with the factors shown on the x-axis against the impacts. For impacts nomenclature, see Fig. 4].

For electricity from bagasse, the production of bagasse and
electricity generation contribute equally the environmental im-
pacts (Fig. 7), except for the AP, EP and POCP which are mainly due
to the slag and ash remaining after the combustion of bagasse. By
contrast, all impacts from hydro-power are from the construction of
the plants (Ecoinvent, 2007).
3.2. Comparison of results with literature
The study results are compared with literature in Fig. 8e10 for
the four electricity options that make up the Mauritius electricity
mix. To enable comparison, only the studies that used the same life
cycle impact assessment method as here (CML 2001) are
considered.
As can be seen from Fig. 8, the results for coal are generally
comparable and within the range found in literature. However, the
ADP, AP and the EP are at the top of the range owing to a lower
efciency of the power plants in Mauritius and the lack of the
abatement of the emissions of sulphur dioxide and nitrous oxides.
Furthermore, the sulphur content in the coal is higher compared to
that assumed in the other studies.
The results for electricity from fuel oil are also comparable to
literature, except for the AP and EP (Fig. 9). Similar to coal, this is
because of the high sulphur content in the fuel oil and the lack of
emission controls at the power plants.
572 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575

Fig. 8. Comparison with literature of environmental impacts from coal electricity. [Literature sources: Hondo (2005), Ecoinvent (2007), Weisser (2007), Peiu (2007), Santoyo-
Castelazo et al. (2011), Mallia and Lewis (2013). Note that only the minimum and maximum values reported in the literature are shown with a number of points laying in be-
tween but not shown. Some impacts have been scaled to t. To obtain the original values, multiply the values shown in the graph with the factors shown on the x-axis against
relevant impacts. For impacts nomenclature, see Fig. 4.].

Fig. 9. Comparison with literature of impacts from fuel oil electricity. [Literature sources: Hondo (2005), Ecoinvent (2007), Kannan et al. (2007), Peiu (2007), Santoyo-Castelazo et al.
(2011). Note that only the minimum and maximum values reported in the literature are shown with a number of points laying in between but not shown. Some impacts have been
scaled to t. To obtain the original values, multiply the values shown in the graph with the factors shown on the x-axis against relevant impacts. For impacts nomenclature, see
Fig. 4.].

Fig. 10. Comparison with literature of impacts from bagasse and hydro-electricity. [Literature sources: Bagasse: Ecoinvent (2007), Ramjeawon (2008), Renouf et al. (2011), Lopes
 ralczyk (2003), Ecoinvent (2007), Peiu (2007), Suwanit and Gheewala (2011), Santoyo-Castelazo et al. (2011). Note that only the
Silva et al. (2012). Hydro: Gagnon et al. (2002), Go
minimum and maximum values reported in the literature are shown with a number of points laying in between but not shown. Some impacts have been scaled to t. To obtain the
original values, multiply the values shown in the graph with the factors shown on the x-axis against relevant impacts. For impacts nomenclature, see Fig. 4.].
 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575
Fig. 11. Environmental impacts of electricity generation in Mauritius. [Some impacts have been scaled to t. To obtain the original values, multiply the values shown in the graph
with the factors shown on the x-axis against the impacts. For impacts nomenclature, see Fig. 4.].
The number of studies of electricity from bagasse is limited,
with only three identied in the literature: one in Mauritius
(Ramjeawon, 2008), one in Australia (Renouf et al., 2011) and
another in Brazil (Lopes Silva et al., 2012). Owing to a lack of
information in these studies, only three impacts can be
compared: the GWP, AP and EP. As indicated in Fig. 10, for the
GWP and EP the values estimated in this work lie within the
range reported in these studies. However, the AP is twice as high
(0.9 kg SO2 eq./MWh) as the maximum value as reported in the
Australian study (0.45 kg SO2 eq./MWh). Arguably, however,
these results are close enough given different climatic conditions
and farming practices in Australia and Mauritius, as well as
various other differences in the assumptions across the life cycle
in the two studies.
It can also be observed from Fig. 10 that the impacts estimated in
the current study for hydro-power tend to be closer to the mini-
mum values or in some cases lower than what is reported in the
literature. This could be due to the different size of the plants and
assumptions on their lifetimes. For example, the study by Suwanit
and Gheewala (2011) considered run-of-river plants with capacities
from 0.2 to 6 MW and a lifetime of 50 years. By comparison, the
capacity of the same type of plant in this study ranges from 0.9 to
1.2 MW and the lifespan is 80 years (see Table 4).
3.3. Environmental impacts of electricity in Mauritius
The life cycle environmental impacts of the electricity mix in
Mauritius for the base year (2007) are shown in Fig. 11 with the
contributions of different sources given in Fig. 12. The latter shows
that the impacts from coal power contribute most of the ADP and
GWP (67% each), EP (59%) and MAETP (78%) with the remaining
impacts being largely (>70%) from fuel oil. Although the share of
electricity from bagasse in the electricity mix is 19%, its contribu-
tion to the total impacts is small, ranging from 1% for the GWP,
ADP, MAETP and TETP to 11% for the EP. The contribution of hydro-
power to the total impacts is insignicant (<0.1%) as it makes up
only 3.4% of the electricity mix.
The following discussion provides further details on each of the
impacts shown in Fig. 11.
573
Global warming potential: The total GWP of electricity gener-
ation in Mauritius is estimated at 868 kg CO2 eq./MWh. By
comparison, the average European and UK GWP values are 471 kg
CO2 eq./MWh and 582 kg CO2 eq./MWh (Ecoinvent, 2007),
respectively.
As mentioned above, electricity from coal contributes 67% to the
total impact, followed by 32% from fuel oil. The large majority of
this is from fuel combustion, with 551 kg CO2 eq./MWh (64%) from
coal and 243 kg CO2 eq./MWh (28%) from fuel oil. Carbon dioxide
accounts for around 96% of the total GWP.
Considering electricity consumption in Mauritius in 2012 and
the GWP estimated here, the annual life cycle emissions of GHG
emissions from electricity generation reached 2.22 Mt CO2 eq./year
(Fig. 13). This represents a 16% increase on the base year (2007)
emissions (Fig. 13). Therefore, if Mauritius is to meet its GHG
emissions target for the energy sector of 30% reduction by 2025,
these trends would need to be reversed as soon as possible.
Abiotic depletion potential: Around 6 kg of Sb eq. is depleted per
MWh electricity. The percentage contributions from coal and oil are
the same as for the GWP (67% and 32%, respectively). Although the
contribution of coal and oil to the electricity mix is roughly equal,
electricity generation from coal is less efcient, hence its higher
depletion of fossil resources per MWh (see Fig. 4).
Acidication potential: This impact is estimated at 15.5 kg SO2
eq./MWh, 70% of which is from fuel oil electricity. As mentioned
earlier, this is due to the high sulphur content (2.4e3%; see Table 2)
and the absence of emission controls in the power plants. The main
contributors to this impact are the emissions of sulphur dioxide
(90%), nitrous oxides (9%) and ammonia (0.3%).
Eutrophication potential: Of the total EP of 447 g PO4 eq./MWh,
59.5% is from coal electricity, 29.8% from oil and 10.7% from bagasse.
Combustion of coal and oil is responsible for 50% and 18% of this
impact, respectively, while agriculture is the main contributor to
electricity generation from bagasse.
Freshwater aquatic ecotoxicity potential: An estimated 11.2 kg of
1,4 dichlorobenzene (DCB) eq. is emitted per MWh with 88%
coming from oil combustion. The main contributors to this impact
are emissions of heavy metals such as vanadium (86%), nickel (9%)
and cobalt (2%).
574 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575
Fig. 12. Contribution of different sources to the environmental impacts of electricity in Mauritius. [For impacts nomenclature, see Fig. 4.].
Human toxicity potential: This impact is equivalent to 150 kg DCB
eq./MWh, of which 73% is from oil and 25% from coal combustion.
Emissions to air of nickel, vanadium and hydrogen uoride
contribute 21%, 12% and 12% respectively; the other contributors
are heavy metals and polycyclic aromatic hydrocarbons.
Marine aquatic ecotoxicity potential: Out of 338 t DCB eq./MWh,
78% is from coal and 21% from oil power plants. The main
contributor to MAETP is hydrogen uoride (78%).
Ozone layer depletion potential: The ODP is estimated at 41 mg
CFC-11 eq./MWh, with 93% being from oil electricity. The large
majority of that (99%) is due to bromo-tri-uoro-methane released
during oil production.
Photochemical oxidants creation potential: As for the ODP, fuel
oil is also the main contributor (73%) to this impact estimated at
594 g C2H4 eq./MWh; a quarter is from coal combustion and the
small remainder from bagasse. The main contributors are sulphur
dioxide (around 94%), carbon monoxide and methane.
Terrestrial ecotoxicity potential: Around 2.7 kg DCB eq. is dis-
charged to land per MWh electricity. Similar to the FAETP, the
emissions of heavy metals to air from combustion of oil are the
main contributor (80%) with vanadium contributing 78%.
Fig. 13. Estimated increase of GWP with electricity consumption from 2007 to 2012.
[Based on the estimated GWP of electricity generation of 868 kg CO2 eq./MWh and the
amount of electricity consumed in each year over the period as shown in the gure.].
4. Conclusions
This paper has considered for the rst time life cycle environ-
mental impacts of electricity in Mauritius aiming to inform elec-
tricity generators and policy makers on the current impacts and
environmental hotspots in the electricity sector. The results suggest
that electricity from oil and coal has the highest impacts, with oil
being the worst option for six (AP, FAETP, HTP, ODP, POCP and TETP)
and coal for four (ADP, GWP, EP and MAETP) out of ten impacts
considered. The majority of the impacts are from the combustion of
fuels in power plants. The lowest impacts are found for electricity
from hydro-power, followed by bagasse. For example, the global
warming potential of electricity from coal is estimated at 1444 kg
CO2 eq./MWh and for oil 754 kg CO2 eq./MWh, while for bagasse
and hydro-power this impact is several orders of magnitude lower
(29 and 8.6 kg CO2 eq./MWh, respectively).
With respect to the electricity mix, oil and coal are the main
contributors to the environmental impacts (89%e99%) of electricity
generation in Mauritius; the contribution of bagasse is small (<1%e
11%) and that from hydro-power negligible (<0.1%).
The GWP of electricity generation in Mauritius is estimated at
868 kg CO2 eq./MWh. Based on the electricity consumption in 2012,
this is equivalent to 2.22 Mt CO2 eq./year, a 16% increase on 2007.
Therefore, if Mauritius is to reduce its carbon emissions, these
trends must be reversed. This could be achieved by reducing the
share of fossil fuels in the electricity mix through an increased use
of renewables as well as by improving the efciency of the fossil
power plants and stimulating a reduction in energy demand.
With respect to the renewables, the potential of bagasse is
limited e its contribution to the electricity mix over the past few
years has been gradually decreasing and this trend is expected to
continue in the future because of the decreasing sugar exports.
However, other renewables such as solar PV and wind could play a
role in the future given that they are abundant in Mauritius.
However, other sustainability aspects will also need to be consid-
ered, including the costs of these technologies, subsidies that the
government may need to provide to stimulate their uptake as well
as social acceptance. These and other sustainability issues should
 R. Brizmohun et al. / Journal of Cleaner Production 106 (2015) 565e575 575

be explored and understood fully to warrant a sustainable future Guinee, J.B., Gorre
e, M., Heijungs, R., Huppes, G., Kleijn, R., van Oers, L., Wegener
Sleeswijk, A., Suh, S., Udo de Haes, H.A., de Bruijn, H., van, Duin, R.,
electricity system in Mauritius.
Huijbregts, M.A.J., 2001. Life Cycle Assessment: an Operational Guide to the ISO
Standards. Kluwer, Dordrecht, The Netherlands.
Acknowledgements Gujba, H., Mulugetta, Y., Azapagic, A., 2011. Power generation scenarios for Nigeria:
an environmental and cost assessment. Energy Policy 39, 968e980.
Hondo, H., 2005. Life cycle GHG emission analysis of power generation systems:
This study was funded by the Commonwealth Scholarship Japanese case. Energy 30, 2042e2056.
Commission and the UK Engineering and Physical Sciences IPCC, 2006. IPCC Guidelines for National Greenhouse Gas Inventories. IGES, Japan.
Research Council (EPSRC, grant no. EP/K011820/1). The authors ISO, 2006a. Environmental Management - Life Cycle Assessment - Principles and
Framework. ISO 14040. ISO, Geneva.
gratefully acknowledge this funding. ISO, 2006b. Environmental Management - Life Cycle Assessment - Requirements
and Guidelines. ISO 14044. ISO, Geneva.
References Kannan, R., Leong, K.C., Osman, R., Ho, H.K., 2007. Life cycle Energy, emissions and
cost inventory of power generation technologies in Singapore. Renew. Sustain.
Energy Rev. 11, 702e715.
Agreco Consortium, 2007. Implementation of the Multi-annual Adaptation Strategy
Lopes Silva, D., Delai, I., Miranda, M., Montes, M., Ometto, A., 2012. LCA application:
for the Mauritian Sugarcane Cluster (2006e2015), Strategic Environmental
the case of the sugar cane bagasse electricity generation in Brazil. In:
Assessment. European Commission and Government of Mauritius.
Dornfeld, D.A., Linke, B.S. (Eds.), Leveraging Technology for a Sustainable World.
Atilgan, B., Azapagic, A., 2015. Life cycle environmental impacts of electricity
Springer Berlin Heidelberg, pp. 85e90.
from fossil fuels in Turkey. J. Clean. Prod. 106, 555e564. http://dx.doi.org/
Mallia, E., Lewis, G., 2013. Life cycle greenhouse gas emissions of electricity
10.1016/j.jclepro.2014.07.046.
generation in the province of Ontario, Canada. Int. J. Life Cycle Assess 18 (2),
Cahoolessur, K., 2002. Life Cycle Assessment of Cane Sugar Production on the Island
377e391.
of Mauritius. University of Mauritius.
MoE, 2009. Technical Advisory Committee Report on Coal Ash Management. Min-
CEB, 2007. Central Electricity Board. Annual Report.
istry of Environment and National Development Unit, Mauritius.
CEB, 2008. Central Electricity Board. Annual Report.
Mott MacDonald, 2003. St. Louis Power Station Redevelopment e Environmental
CEB, 2013. Integrated Electricity Plan 2013e2022. Central Electricity Board.
Impact Assessment Report.
CSO, 2008. Digest of Statistics. Central Statistics Ofce, Mauritius. www.gov.mu/
MREPU, 2009. Long-term Energy Strategy 2009-2025. Ministry of Renewable En-
portal/site/cso.
ergy and Public Utilities, Republic of Mauritius. www.sids2014.org/content/
CSO, 2012. Digest of Statistics. Central Statistics Ofce, Mauritius. www.gov.mu/
documents/68Energy%20Strategy.pdf.
portal/site/cso.
MSIRI, 2007. Annual Report, Mauritius Sugar Industry Research Institute.
Ecoinvent, 2007. Ecoinvent V 2.0 Database. Swiss Centre for Life Cycle Inventories,
Peiu, N., 2007. Life cycle inventory study of the electrical energy production in
Dbendorf, Switzerland.
Romania. Int. J. Life Cycle Assess. 12 (4), 225e229.
EIA, 2013. World Energy Outlook. U.S. Energy Information Agency. www.eia.gov/  Consultants, 2008. SimaPro 7. PRe  Consultants, Amersfoort, The Netherlands.
PRe
forecasts/ieo/world.cfm.
Ramjeawon, T., 2008. Life cycle assessment of electricity generation from bagasse in
EPA, 1995. State Workbook Methodologies for Estimating Greenhouse Gas Emis-
Mauritius. J. Clean. Prod. 16, 1727e1734.
sions. EPA 230-B-95e001, second ed. US Environment Protection Agency,
Renouf, M.A., Pagan, R.J., Wegener, M.K., 2011. Life cycle assessment of Australian
Washington D.C. Ofce of Policy, Planning and Evaluation.
sugarcane products with a focus on cane processing. Int. J. LCA 16, 125e137.
, J., Gasol, C.M., Gabarrell, X., Rieradevall, J., 2010. How important
Foidart, F., Oliver-Sola
Santoyo-Castelazo, E., Gujba, H., Azapagic, A., 2011. Life cycle assessment of elec-
are current energy mix choices on future sustainability? Case study: Belgium and
tricity generation in Mexico. Energy 36, 1488e1499.
Spaindprojections towards 2020e2030. Energy Policy 38, 5028e5037.
Sinnott, R., 1996. Coulson and Richardson's Chemical Engineering. Pergamon Press.
Gagnon, L., Be langer, C., Uchiyama, Y., 2002. Life-cycle assessment of electricity gen-
Stamford, L., Azapagic, A., 2012. Life cycle sustainability assessment of electricity
eration options: the status of research in year 2001. Energy Policy 30, 1267e1278.
options for the UK. Int. J. Energy Res. 36, 1263e1290.
Greening, B., Azapagic, A., 2012. Domestic heat pumps: life cycle environmental
Suwanit, W., Gheewala, S.H., 2011. Life cycle assessment of mini-hydro-power
impacts and potential implications for the UK. Energy 39, 205e217.
plants in Thailand. Int. J. LCA 16, 849e858.
Golder Asssociates, 1997. Environmental Impact Assessment report for Centrale
Weisser, D., 2007. A guide to life-cycle greenhouse gas (GHG) emissions from
Thermique de Belle Vue. Golder Asssociates, Mauritius.
electric supply technologies. Energy 32, 1543e1559.
Go ralczyk, M., 2003. Life-cycle assessment in the renewable energy sector. Appl.
Whiting, A., Azapagic, A., 2014. Life cycle environmental impacts of generating
Energy 75 (3), 201e211.
electricity and heat from biogas produced by anaerobic digestion. Energy 70,
GOM, 2007. Outline of Energy Policy 2007-2025: towards a Coherent Strategy for
181e193.
the Development of the Energy Sector in Mauritius. Government of Mauritius.