Transportation Engineering 2 (2020) 100019

Contents lists available at ScienceDirect

Transportation Engineering
journal homepage: www.elsevier.com/locate/treng

Life cycle inventory of bitumen in South Africa

Sheldon A. Blaauw a,b,∗, James W. Maina a, Louis J. Grobler c
a
Department of Civil Engineering, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa
b
Ove Arup & Partners Limited, 1st Floor City Gate West, Tollhouse Hill, Nottingham NG1 5AT, United Kingdom
c
Dinokeng Engineers, Office@Nature, 500 Botterklapper Street, Die Wilgers, Pretoria 38602, South Africa

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

Keywords: Road pavements are constructed of bulk raw materials such as aggregate, cement, bitumen and water. The en-
Life cycle assessment vironmental, social, and economic impacts of the materials and processes should be quantified and monitored
Life cycle inventory towards the optimisation of pavement design. At present, no such protocol is in place in South Africa. This paper
Pavement sustainability
proposes a framework for the development of a pavement life cycle assessment model, starting by documenting
Carbon dioxide equivalent
the life cycle inventory for bitumen, one of the leading environmental and social burdensome materials used for
Bitumen
pavement development.
This inventory acts as the first building block in the development of a life cycle assessment model by evaluating
and delineating primary flows (inputs of materials and energy and outputs of pollution) related to the supply
chain of bitumen in South Africa. The primary flows are represented by indicators which measure their quantita-
tive impacts. The inventory provides impact category indicators for environmental and social related emissions,
energy- and water-use and currently excludes other indicators such as emissions to water, waste generation, jobs
creation and economic transformation, amongst others. These indicators are omitted due to lack of quality data at
present and difficulty in the quantification of impacts, but recognition is given to their relevancy and importance.
The approach is envisioned to function as the first version of a living protocol that will be improved on through
further research. Although the primary target audience is South African road authorities, the approach can be
adapted for use in any country.

1. Introduction For bitumen, the background processes involve extraction processes

whereas for energy, they may include delivering the crude oil to the
Road pavements are constructed of bulk raw materials such as aggre- place of refining. Due to the influence of background processes, it is im-
gate, cement, bitumen and water. The environmental, social, and eco- portant to develop inputs to an LCI regionally and consider the typical
nomic impacts of the materials and processes should be quantified and practices and available resources of that region. It is often challenging
monitored towards the optimisation of the pavement design. At present, collecting data from regions which do not voluntarily publish data re-
no such protocol is in place in South Africa. The need was identified to garding the processes or resources used to produce pavement materials.
develop a holistic life cycle assessment (LCA) model of the effects of al- In these situations, published data sets (either commercial sources or
ternative pavement designs and technologies in terms of construction, from literature) are often utilised. The issues related to the use of proxy
maintenance and rehabilitation. A life cycle inventory (LCI) is the first data were observed in the early years of pavement LCI development
step in the development of an LCA. [2] when studies were rapidly published and had numerous flaws. These
The LCI analysis phase consists of the tracking of flows (inputs of flaws commonly manifested as poor-quality data, gaps in research be-
materials, resources and energy and outputs of waste, pollution and co- ing filled by proxy data and a general lack of consideration for location
products) which are represented by indicators for the system being stud- specific evaluations.
ied [1]. For a typical asphalt plant, the indicators are listed in the in- Certain inventories are often viewed as ‘golden black boxes’ [3] and
ventory in terms of energy use, the use of materials such as aggregates, applied without much appreciation for inherent limitations in the inven-
bitumen and water and the production of greenhouse gasses. These are tories. Considering the widely cited Eurobitume inventory, originally
referred to as the primary data sources for an LCI, which are to be traced published in 2012 [4] and revised in 2020 [5]. The inventory has vary-
back to their origin, known as the background data processes [1]. ing embodied carbon factors applied to certain elements of the inventory

∗
Corresponding author.
E-mail address: sheldon.blaauw@arup.com (S.A. Blaauw).

https://doi.org/10.1016/j.treng.2020.100019
Received 14 July 2020; Received in revised form 1 September 2020; Accepted 1 September 2020
2666-691X/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

Fig. 1. System boundaries of the assessed system.

between the two data sets, with up to 50% differences in some instances. omitted due to lack of quality data at present and difficulty in the quan-
This highlights the impacts of lack of quality data on predictions. An ad- tification of impacts, but recognition is given to their relevancy and im-
ditional flaw present in LCI development models is the inconsistency in portance. Indicators related to the provision of infrastructure required
selecting impact categories; where energy and carbon emissions are typ- for the extraction, transportation, refinery and storage of bitumen are
ically focused on with other important indicators such as water use often excluded from this study. Although the primary target audience is South
being omitted [6]. Focus on a holistic inventory which incorporates both African road authorities, the approach can be adapted for use in any
environmental and social indicators, as demonstrated by Bressi et al. [7], country.
is required.
Supporting the need to quantify these indicators, Giunta et al. 2.2. Functional unit and system boundaries
[8] demonstrates the contributions of various pavement life cycle phases
(excluding the use phase) to the total energy and carbon footprint esti- The functional unit, representing the reference unit used to quantify
mates for a typical pavement project. Giunta et al. found that the pro- indicator performance, is either one tonne crude oil bitumen depending
duction phase generally accounted for more than 50% of the total energy on the relevant phase considered. This study comprises all the processes
and emissions related to pavement development, with bitumen produc- for the production and transportation of bitumen for a cradle-to-gate
tion being the leading material contributor. Methods, such as the use approach. In detail, the system boundaries include the following steps,
of recycled polystyrene to substitute bitumen, are available to reduce shown in Fig. 1:
these impacts [9].
The inventory detailed in this study is specific to South Africa but 1 Crude oil extraction, which includes the requirements for extraction
can be modified to be representative of any country. This approach fol- of crude oil from oil reserves in Africa and the Middle East and prepa-
lows the notion that sustainability is spatially dependant and influenced ration for transportation.
by local boundaries [10]. The research highlights the typical processes 2 Crude oil transportation, which includes the processes required for
bitumen undergoes during production and the common constraints ex- transporting crude oil by ship and pipeline to the relevant South
perienced in the development of an inventory for penetration-grade bi- African refineries.
tumen. Furthermore, the need to develop country specific inventories is 3 Bitumen refinery and storage, which includes the processes required
demonstrated through data validation and sensitivity analyses, to high- to process, refine and store paving-grade bitumen in South Africa.
light key factors affecting estimates and support conclusions and recom-
mendations. 2.3. Impact category indicators

2. Method: life cycle inventory For this study, the following impact category indicators are used:

- Carbon emissions (kg CO2 e/t)

2.1. Goal and scope - Energy use (MJ/t)
- Water use (l/t)
The goal of this study is to provide an inventory for the production - Sulphur Dioxide emissions (kg SO2 /t)
of straight-run penetration grade bitumen in South Africa. International - Nitrogen Oxides emissions (kg NOx/t)
Organisation for Standardisation (ISO) standard 14040 [11] was used - Particulate Matter emissions (kg PM10 /t)
to develop this inventory. The inventory is intended to act as a building - Volatile Organic Compounds emissions (kg VOC/t)
block on which to improve and develop further life cycle inventories
where penetration grade bitumen is used with the ensuing intention Allocation procedures specified in ISO [11] are not performed in this
to develop a holistic LCA model for pavement infrastructure in South study.
Africa.
The scope of the study extends from cradle-to-gate for the production 2.4. Emissions
of penetration grade bitumen at any refinery in South Africa. The inven-
tory provides indicators for environmental and social related emissions, In this study, two separate types of emissions are focused on, namely
energy- and water-use but currently excludes other indicators such as environmental and social relevant emissions. Environmental emissions
emissions to water and waste generation. Social related indicators in- are “greenhouse” gasses (GHGs) affecting climate, covered under the
cluding jobs creation, local economic development, and economic trans- Kyoto protocol, such as carbon dioxide and methane [12]. Social rele-
formation are also excluded from consideration. These indicators are vant emissions are those that predominantly affect human health such
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

as Sulphur Dioxide (SO2 ), Nitrogen Oxides (NOx), Particulate Matter Table 1

(PM) and Volatile Organic Compounds (VOCs) [13]. Weighted average indicator factors for crude oil extraction per tonne.

Indicators Unit Average∗ [18–26]

2.4.1. Environmental emissions
Environmental Related Indicators
To measure the environmental impact of the provision of pavement
Carbon Emissions kg CO2 e/t 157.38
infrastructure, it is important to understand the various terms used to Energy Use MJ/t 1450.28
define it and its effect on the environment, such as “greenhouse gases”, Water Use l/t 290.92
“CO2 ”, “CO2 e”, and “carbon” as these terms may often be used inter- DQI 75%
Social Related Indicators European average[5]
changeably, and their meaning may become confusing [14].
Sulphur Dioxide Emissions kg SO2 /t 0.4
A GHG is any gas in the atmosphere which adds to the greenhouse Nitrogen Oxides Emissions kg NOx/t 0.25
effect by absorbing and re-emitting heat, increasing the temperature of Particulate Matter Emissions kg PM10 /t 0.02
the planet above what it would normally be. Carbon dioxide is the most Volatile Organic Compounds Emissions kg VOC/t 0.33
common GHG emitted from human activities in terms of quantity and DQI 72%
total global warming potential (GWP). The GWP indicates the amount ∗
Weighted average for Africa and the Middle East.
of warming a gas causes over several years (typically 100 years). CO2 e
is a term used to describe the collection of greenhouse gases in a com-
mon unit. For any quantity or type of GHG, CO2 e signifies its equivalent ports (mainly African and the Middle Eastern countries) and the emis-
CO2 and the GWP impact on the atmosphere. Embodied carbon refers to sions so produced are specific to the extraction country and company.
the CO2 e associated with the non-operational phase of the project. This The International Association of Oil and Gas Producers (IPIECA) pub-
includes the emissions caused by the extraction, manufacturing and pro- lished guidance on voluntary sustainability reporting for oil and gas pro-
duction of material, transportation, construction, maintenance, rehabil- ducers worldwide. This is a parallel program to the Voluntary National
itation, deconstruction and disposal and end of life aspects which make Review developed by the United Nations (UN) and allows interested
up the materials of a pavement. The whole life carbon of pavement in- companies to publish data for certain indicators related to, amongst
frastructure is both the embodied carbon and the CO2 e associated with others, the environmental and social impacts of their operations. Using
the use (i.e. operational) phase of the pavement [12,14]. these data, together with Eurobitume [5] estimates, typical values for
the environmental and social emissions emitted and energy- and water-
2.4.2. Social emissions use related to crude oil extraction may be calculated, shown in Table 1.
Emissions from pavement infrastructure development which pre- These indicator factors account for all background processes typically
dominantly affect human health include SO2 , NOx, PM and VOCs. When required for crude oil extraction including preparation for transporta-
these emissions are breathed in, they may cause noise, throat and airway tion. These estimates have been obtained from various sources and do
problems, exacerbate asthma and in some cases cause cancer [13]. not necessarily represent the view or findings of the respective compa-
nies referenced. Data quality indicators (DQIs) are used in this study to
2.4.3. Legislation governing emissions in South Africa assess the quality of data collected according to the applicable ISO guid-
The main legislation which governs these emissions in South Africa is ance [11]. To do this, an adapted data quality matrix [17] is used and
the National Environmental Management: Air Quality Act (AQA) [15]. applied to all data in this study. The data quality matrix is provided as
The act forms the framework for the control of air pollution and defines supplementary information.
the air emissions limits for specific activities and processes. According
to section 29 of the AQA, any production process which involves CO2 e
emission over 0.1 Megatonnes annually is required to provide a pol- 2.7. Transport
lution prevention plan to the Minister for approval. The processing of
crude oil at a refinery requires an Atmospheric Emission License. De- 2.7.1. Shipping
fined emission limits (taken as daily averages) are provided for refineries Crude oil tankers are available in a variety of sizes, namely PANA-
in South Africa under the AQA. These emission limits are 20 tonne/day MAX, AFRAMAX, SUEZMAX, VLCC and ULCC. In this study, crude oil is
for SO2 , 4 tonne/day for NOx and 1 tonne/day for PM. assumed to be transported using an AFRAMAX vessel size of between 80
- 120 Dead Weight Tonnes (DWT). This vessel size is popular amongst
2.5. Life cycle inventory of bitumen in South Africa oil companies [27] and assumed to be a conservative representation of
typical vessel size from all regions to South Africa.
South Africa does not have any crude oil reserves and imports the Typical AFRAMAX vessel specifications are used for this study [28].
raw material to produce bitumen. The major contributors to crude oil The fuel consumption is calculated on the assumption that the ship is
imports data as provided by the South African Department of Energy fully loaded for the trip to South Africa, and the return trip is not consid-
[16] are divided into two regions; Africa (52%) and Middle East (48%). ered as part of the overall journey. An adjustment is made in that 28% of
Bitumen is shipped to predominantly two ports in South Africa, namely: South Africa’s bitumen production is exported [29]. It is assumed then
Durban (77%) and Cape Town (23%). South Africa has four refineries: that the ship would be loaded for its return trip from South Africa and
•
emissions related to the return trip are embodied in export products. The
Natref (Sasol and Total) – Sasolburg,
•
data for the marine engine emissions produced by the shipping vessel
Enref (Engen) – Durban,
•
is obtained from the European Environment Agency (EEA) [30]. Con-
Sapref (Shell and BP) – Durban), and
•
sumption of fuel to load and discharge crude oil or petroleum products
Astron (old Chevron or Caltex) – Cape Town.
from the vessel are vessel- and port-specific. Accurate data are not avail-
These refineries are responsible for the supply of bitumen to the able for South African ports and typical conservative values [5] are used
South African market; however, they are not always able to guarantee for the analysis. To determine transportation distance, a port to port
stable bitumen supplies and refined bitumen is often imported. distance calculation tool [31] is used. Missing information is supple-
mented with distance measurements. EEA [30] provides estimates for
2.6. Extraction social-related emissions for typical marine fuel used. It is noted that the
type of marine fuel used affects these estimates. Using this information,
Oil extraction is the first process in bitumen production. The oil ex- typical indicator factors may be calculated for shipping one tonne of
traction process occurs in other countries from which South Africa im- crude oil from all regions, shown in Table 2.
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

Table 2 Table 4
Weighted average indicator factors for shipping one-tonne crude oil. Average refinery and storage environmental indicator factors for one-
tonne bitumen production.
Indicators Unit Weighted Average
Indicators Unit Average
Environmental Related Indicators
Carbon Emissions kg CO2 e/t 11.12 Environmental Related Indicators
Energy Use MJ/t 140.15 Carbon Emissions [38,39] kg CO2 e/t 53.09
Water Use l/t 48 Energy Use MJ/t 610
DQI 86% Water Use l/t 342.5
Social Related Indicators DQI 75%
Sulphur Dioxide Emissions kg SO2 /t 0.13 Social Related Indicators
Nitrogen Oxides Emissions kg NOx/t 0.51 Sulphur Dioxide Emissions kg SO2 /t 0.6
Particulate Matter Emissions kg PM10 /t 0.01 Nitrogen Oxides Emissions kg NOx/t 1.75
Volatile Organic Compounds Emissions kg VOC/t 0.02 Particulate Matter Emissions kg PM10 /t Not reported
DQI 86% Volatile Organic Compounds Emissions kg VOC/t 0.28
DQI 76%

Table 3
Average pipeline transport indicator factors for one-tonne crude oil for coastal
and inland scenarios [33,34]. 2.8. Bitumen refinery and storage
Indicators Unit Coastal average Inland average
To produce bitumen, a complex refinery is required to conduct a
Environmental Related indicators straight-run distillation process on crude oil which produces a range of
Carbon Emissions kg CO2 e/t 0.30 11.77
petroleum products. Bitumen is a major product of this process, esti-
Energy Use MJ/t 1.01 39.6
Water Use l/t 0.19 7.26 mated at 28.5% bitumen yield by mass [5].
DQI 74% The energy used for the distillation process is not well reported, how-
Social Related Indicators ever, Eurobitume [5] estimates that 510 MJ/t is required and includes
Sulphur Dioxide Emissions kg SO2 /t 0 0.1
consideration of various back-ground processes such as crude oil han-
Nitrogen Oxides Emissions kg NOx/t 0 0.05
Particulate Matter Emissions kg PM10 /t 0 0
dling, desalting, flaring, loading area, general heating and lighting. A
Volatile Organic Compounds kg VOC/t Not reported Not reported further 100 MJ/t is estimated for the storage of bitumen. CONCAWE
Emissions [35] data highlights that of the total energy used in a refinery, 96% is
DQI 76% obtained from burning bi-products (84% refinery gas) and heavy fuel
oil (12%). The remaining 4% is from the electrical grid. Eurobitume
[5] disregards the energy obtained from the electricity grid. Alternative
These estimates are based on a range of assumptions and simplifica- opinions [3] state that South African refineries may use heavy fuel oil
tions. This applies for the Eurobitume [5] database which only considers as the primary fuel type. Based on this information, it was decided to
one shipping vessel specification. More accurate results may be obtained use heavy fuel oil (90%) as the main fuel type, supplemented with esti-
if better quality data is acquired reflecting the actual shipping specifi- mated distributions for refinery gas (6%) and electricity (4%) from the
cations of South Africa’s oil imports. It is further worth noting that the electrical grid.
relatively long shipping distances to South Africa significantly increase Each of these energy sources was further investigated. Using calorific
the environmental and social burden of imported oil compared to most values of 40.4 MJ/kg for heavy fuel and 49.5 MJ/kg for refinery gas
countries. [36], the consumption of energy for the production of bitumen by
straight-run distillation may be calculated. Social related emissions for
typical refinery processes in South Africa are obtained from Sapref [37].
2.7.2. Pipeline Table 4 shows the indicator factors for one-tonne bitumen production.
South Africa makes use of various pipelines to transport crude oil The total emissions of 53.09 kg CO2 e/t for the refinery and stor-
from the ships to the four refineries. For instance, ships offloading at age of bitumen may be compared to the estimate by Eurobitume [5] of
Durban Port are required to offload at an oil terminal roughly 2.5 km 25.55 kg CO2 e/t. The increase in emissions is related to the differences
offshore of the Bluff. Oil is pumped from the terminal through to the in energy source distributions and higher emissions factors for South
Enref and Sapref refineries, and further to Natref in Sasolburg with sim- African electricity production, amongst others. It is further worth not-
ilar operations used for Astron in the Cape. Pipelines transport 100% of ing that the emissions related to bitumen refinery between the Euro-
crude requirements to refineries in South Africa [32]. Pipeline lengths bitume [4] (46.03 kg CO2 e/t) and Eurobitume [5] (25.55 kg CO2 e/t)
are obtained from Transnet [32]. represent a reduction of roughly 46% and highlights the requirement
The energy required to pump crude oil to the different refineries is for further research and the need for quality data.
calculated using conservative data provided by Eurobitume [5]. It is
noted that the energy use from the Eurobitume [5] database is 260%
higher than that from the Eurobitume [4] database. This reflects the 3. Results
assumptions made and the effect different pipeline specifications and
topography have on the energy consumption. Eskom [33] provides esti- The results of the analysis are provided for the emissions, energy-
mates for social-related emissions for electricity obtained from the elec- and water-use for bitumen production in South Africa considering both
trical grid. Emissions are provided for two scenarios, namely coastal coastal and inland scenarios, shown in Table 5.
(average for Astron, Sapref and Enref) and inland (Natref). Typical fac- An analysis of the phases which contribute most to the carbon emis-
tors are shown in Table 3 for the transportation of one-tonne bitumen sions of bitumen in South Africa was conducted. Fig. 2 shows the con-
via pipelines. tribution of each phase.
More accurate data is required for the pipelines in South Africa, It is seen that the extraction phase contributes to the majority of the
considering the energy efficiency of pump stations, the effect of to- carbon emissions of bitumen production for both coastal and inland sce-
pography (with Sasolburg being 1500 m above sea level, for instance), narios, followed by the refinery and storage and transportation phases.
and the typical specifications of the various pipelines used to transport Similar results are obtained when evaluating the contributions for other
crude oil. indicators.
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

Table 5
Indicator factors for bitumen in South Africa.

Scenario Unit Extraction Transportation Refinery and storage Total

Carbon Emissions
Coastal kg CO2 e/t 157.38 11.42 53.09 221.89
Inland kg CO2 e/t 22.89 233.36
Energy Use
Coastal MJ/t 1450.28 141.15 610 2201.40
Inland MJ/t 179.75 2240.00
Water Use
Coastal l/t 290.92 48.19 342.50 681.61
Inland l/t 55.26 688.68
Sulphur Dioxide Emissions
Coastal kg SO2 /t 0.40 0.13 0.60 1.13
Inland kg SO2 /t 0.23 1.23
Nitrogen Oxides Emissions
Coastal kg NOx/t 0.25 0.51 1.75 2.51
Inland kg NOx/t 0.56 2.56
Particulate Matter Emissions
Coastal kg PM10 /t 0.02 0.01 Not 0.03
Inland kg PM10 /t reported
Volatile Organic Compounds Emissions
Coastal kg VOC/t 0.33 0.02 0.28 0.68
Inland kg VOC/t

Fig. 2. Contribution of phases to the carbon emissions of

bitumen in South Africa.

It is noted that the results presented in this study have made use of 3.1.1. Sensitivity analysis
proxy data in the absence of accurate local data. As described in litera- To determine the parameters and assumptions which most greatly
ture, the use of proxy data to fill gaps in research is a common approach influence the carbon emission values of bitumen production in South
implemented in LCA studies where the driver of the study is a relative Africa, a sensitivity analysis was conducted. The results of this analysis
comparison and not quantifying absolute numbers. However, as bitu- highlight the areas which predominantly require additional research to
men has been identified as a leading material contributing to emissions ensure accurate estimates are available.
[8], the need for accurate estimates is emphasised. The region and typical associated extraction processes contribute sig-
The use of proxy data was especially required for the refinery nificantly to the carbon emissions of bitumen production. In the sensitiv-
and storage phases and is related to the energy requirements for bi- ity analysis, typical extraction embodied carbon values for the Former
tumen processing. Participation of local refineries could enhance the Soviet Union and South America [5] were used as alternative scenarios.
quality of available data and remove the need to rely on external Using these estimates, a 25% reduction in extraction embodied carbon
estimates. estimates per tonne crude oil was obtained.
The relatively low shipping estimate is due to the type of vessel con-
3.1. Data validation sidered in the calculations and is not a true reflection of the shipping
reality. This statement is however also applicable to the shipping esti-
To validate the results, a comparison may be made to similar life cy- mates of Eurobitume [5]. In the sensitivity analysis, the shipping speci-
cle inventories related to penetration grade bitumen production, shown fications from the Eurobitume [5] analysis were used as the alternative
in Table 6. scenario and a 12.5% increase in carbon emissions was calculated for
The carbon emissions for bitumen production from crude oil in South crude oil shipping per tonne.
Africa are generally higher than in other regions in the world. Various The energy estimate required to pump crude oil greatly affects the
factors affect these estimates, detailed in the sensitivity analysis shown environmental burden of transportation. The lengthy pipeline used to
in the following section. pump crude oil to Sasolburg may have a range of pipeline diameters,
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

Table 6
Comparison of carbon emissions for straight-run bitumen production from crude oil.

Life Cycle Inventory Unit Extraction Transportation Refinery and storage Total

Hammond and Jones [17] kg CO2 e/t – – – 191.00

Eurobitume [4] kg CO2 e/t 99.14 30.08 45.03 174.25
Eurobitume [5] kg CO2 e/t 90.14 21.25 25.41 136.80
South Africa – Import Coastal kg CO2 e/t 157.38 11.42 53.09 221.89
South Africa – Import Inland kg CO2 e/t 157.38 22.89 53.09 233.36

Table 7 • The inventory provided in this study is based on reasonable quality

Key indicators for reporting. data and provides rational estimates for certain environmental and
Key Indicators Unit social impacts for bitumen production in South Africa. This inven-
1
tory may be used to enhance prediction estimates for assessments
Crude oil carbon emissions kg CO2 e/t
where paving-grade bitumen is utilised.
Comprehensive energy consumption per tonne of bitumen2 MJ/t
• There is a drive within South African roads authorities to promote
Freshwater consumption per tonne of bitumen l/t
Total water consumption per tonne of bitumen l/t sustainable technologies and practices. However, targeted efforts are
Emissions per tonne of bitumen3 kg gas/t required from roads authorities to focus on the supply chain of bitu-
Emissions to water per tonne bitumen4 ppm/t
men and further to asphalt production. These authorities would do
Waste generation per tonne bitumen kg/t
Job creation in the bitumen production sector Jobs/year
well to support academic and industry research aimed at developing
Educational outreach a database of quality data.
Women empowered5 • There is a need to create an online platform to promote sustainable
1 practices. This online platform may take the form of a SANRAL web-
dependant on source and transportation method of crude oil to the re-
finery. page that is open to the public to provide a knowledge base for the
2
Including energy source distributions. pavement community including contractors, government agencies,
3
Key emissions include CO2 e, SO2 , NOx, PM and VOCs, amongst others. private consultants as well as future industry members. Experiences
4
Key emissions include wastewater, hazardous waste (e.g. phenol and with Pavement Interactive, a similar online platform, has shown
other aromatic compounds) and oil effluent, amongst others. great reach and success [40].
5
May be assessed qualitatively using a typical empowerment index and • Certain social indicators are not yet well understood, which may
monitored yearly to evaluate performance. carry little meaning to the practitioner. Further research is required
to develop valid quantification and impact models.
• Development of a standard pavement LCA is required for a cradle-to-
pump stations specifications, various topographical variations and as
grave scenario. This LCA needs to be developed specifically to South
a result of a different energy requirement to pump crude oil. Carbon
Africa.
emission estimates are directly related to energy consumption and a • Holistic sustainability is only achieved if all three tenets of sustain-
change in energy distribution proportionally affects the embodied car-
ability (i.e. economic, environmental, social) are confidently evalu-
bon of crude oil transport via pipelines. Using Eurobitume [4] estimates
ated. Further research is required to develop a model that quantifies
as the alternative scenario, a 65% decrease in embodied carbon esti-
holistic sustainability.
mates is calculated.
What further influences the accuracy of the results are basing refin- Declaration of Competing Interest
ery and storage estimates on typical international values. However, it
is worth noting that estimates for the refinery and storage phases may The authors declare that they have no known competing financial
be reduced if accurate energy distributions are known. Using an energy interests or personal relationships that could have appeared to influence
distribution change favouring heavy fuel oil as the main energy source the work reported in this paper.
for the alternative scenario reduced the carbon emissions of bitumen
refinery and storage by 9.7%. This highlights the need for the industry Supplementary material
to partake by voluntarily releasing information related to production
factors. Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.treng.2020.100019.
3.2. Guidance for refineries and road authorities in South Africa
References
It is proposed that refineries monitor, measure and voluntarily pub- [1] FHWA, Federal highway administration, Life Cycle Assessment of Pavements, Tech
lish data on the following key indicators of bitumen production, shown Brief, Washington, D.C., USA, 2014 FHWA-HIF-15-001.
in Table 7. [2] N. Santero, E. Masanat, A. Horvath, Life cycle assessment of pavements. Part I: crit-
ical review, Resour. Conserv. Recycl. 55 (9–10) (2011) 801–809.
It is further proposed that refineries use carbon emission estimates [3] J. Harvey, A. Saboori, M. Dauvergne, W.JvdM. Steyn, A. Jullien, H. Li, Comparison
for crude oil to benchmark the carbon footprint for crude they use. This of new pavement construction GHG and energy impacts in different regions, Papers
may assist overall decision making and on where to source crude from. From the International Symposium on Pavement Life Cycle Assessment, University
of California, California, USA, 2014 ISBN 978-0-692-29357-7.
There also exists a need for road authorities such as the South African
[4] Eurobitme, The Eurobitume Life-Cycle Inventory for Bitumen, Version 2, European
National Roads Agency SOC Limited (SANRAL) to promote voluntary Bitumen Association, Brussels, Belgium, 2012.
monitoring and publication actions and provide a platform for the col- [5] Eurobitme, The Eurobitume Life-Cycle Inventory for Bitumen, Version 3.1, European
Bitumen Association, Brussels, Belgium, 2020.
lection and delineation of data from industry and researchers.
[6] T.M. Gulotta, M. Mistretta, F.G. Practicò, A life cycle scenario analysis of differ-
ent pavement technologies for urban roads, Sci. Total Environ. 673 (18) (2019)
Conclusions 585–593.
[7] S. Bressi, G. D’Angelo, J. Santos, M. Giunta, Environmental performance analysis of
bitumen stabilized ballast for railway track-bed using life cycle assessment, Constr.
The following conclusions were drawn from the study: Build. Mater. 188 (2018) 1050–1064.
S.A. Blaauw, J.W. Maina and L.J. Grobler Transportation Engineering 2 (2020) 100019

[8] M. Giunta, M. Mistretta, F.G. Practicò, M.T. Gulotta, Environmental sus- [24] Noble Energy, 2018. Sustainability report. Noble Energy. Houston Texas. Available
tainability and energy assessment of bituminous pavements made with at Nblenergy.com.
unconventional materials, in: Proceedings of the 5th International Sympo- [25] Tullow Oil, 2019. Sustainability Report. Tullow Oil. London, UK. Available at Tul-
sium on Asphalt Pavements and Environment (APA), 2020, pp. 123–132. lowoil.com.
https://doi.org/10.1007/978-3-030-29779-4_12. [26] PDO, Sustainability Report 2018 – Building a Sustainable Future, OPD, Muscat, Sul-
[9] M. Vila-Cortavitarte, P. Lastra-Gonzalez, M.A. Calzada-Perez, I. Indacoechea-Vega, tanate of Oman, 2018 Available at Pdo.co.om.
Analysis of the influence of using recycled polystyrene as a substitute for bitumen [27] US EIA, 2014. United Stated Energy Information Administration. Hamilton,
in the behavior of asphalt concrete mixtures, J. Clean. Prod. 170 (2018) 1279– M.T. London Tanker Brokers’ Panel. US EIA. Viewed 05/05/2020. Available at
1287. Eia.gov/todayinenergy.
[10] U. Berardi, Clarifying the new interpretations of the concept of sustainable building, [28] Wärtsilä, 2020. WSD 42 111K Data Sheet. Wärtsilä. Viewed 01/04/2020. Available
Sustain. Cities Soc. 8 (2013) 72–78. at Wartsila.com.
[11] ISO, International Organisation for Standardisation, ISO 14040, Environmental Man- [29] SABITA, South African bitumen association, Best Practice Guide for the Procure-
agement - Life Cycle Assessment - Principles and Framework, ISO, Geneva, Switzer- ment and Importing of Bitumen, Edition 1, SABITA, Cape Town, South Africa, 2013
land, 2006. Available at Sabita.co.za.
[12] IPCC, Summary for policymakers, Contribution of Working Group II to the Fourth [30] EEA, European Environment Agency, Air Pollutant Emission Inventory Guidebook.
Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge Technical Guidance to Prepare National Emission Inventories, EEA, Luxembourg,
University Press, Cambridge, England, 2007. Luxembourg, 2019 ISSN 1977-8449.
[13] WHO, World Health Organisation, Air Quality Guidelines Global Update: Partic- [31] SeaRates, 2020. Port to port distance calculation tool. Searates. Viewed 21/04/2020.
ulate Matter, Ozone, Nitrogen Dioxide and Sulphur Dioxide, WHO, Copenhagen, Available at Searates.com
Denmark, 2005 EUR/05/5046029. [32] Transnet, 2019. Pipelines - Information Sheet. Transnet, Johannesburg, South Africa.
[14] M. Brander, Greenhouse gases, CO2 , CO2 e, and carbon: what do all these terms Available at Transnet.net.
mean? Ecometrica (2012) Edinburgh, Scotland. Available at Ecometrica.com. [33] Eskom, Environmental implications of using or saving electricity, Eskom Holdings,
[15] AQA, National Environmental Management: Air Quality Act. Act No. 39, Department Sandton, South Africa, 2014 Available at Integratedreport.eskom.co.za.
of Environmental Affairs, Pretoria, South Africa, 2004. [34] R. Inglesi-Lotz, J.N. Blignaut, Electricity intensities of the OECD and South Africa: a
[16] SA-DoE, Annual Report 2018/2019, DoE, Pretoria, South Africa, 2019 Available at comparison, Renew. Sustain. Energy Rev. 16 (7) (2012) 4491–4499.
Energy.gov.za. [35] CONCAWE, Sulphur Dioxide Emissions From Oil Refineries in Europe. Report no.
[17] Hammond, G.P., Jones, C.I., 2019. Embodied energy and carbon – the inventory for 5/18, CONCAWE, Brussels, Belgium, 2018.
carbon and energy database. Available at Circularecology.com [36] UN – ESY, United Nations- Energy Statistics Yearbook, Statistic Division. UN, Brus-
[18] CNOOC, Environmental, social and governance report, Investor Relations Depart- sels, Belgium, 2017.
ment of CNOOC Limited, Beijing, China, 2018 CNOOCAvailable at Cnoocltd.com. [37] Sapref, 2017. Sustainability report - A report for our stakeholders. Sapref. Durban,
[19] Eni, Sustainability report. Eni, Rome, Italy, 2018 Available at Eni.com. South Africa. Available at Sapref.com
[20] Equinor, Sustainability report, Equinor. Stavanger, Norway, 2019 Available at [38] IMO, International Maritime Organisation, Third IMO Greenhouse Gas 2014 Study,
Equinor.com. IMO, London, UK, 2015.
[21] Occidental, 2018. Occidental petroleum corporation - annual summary, environ- [39] Volker-Quaschning, Specific Carbon Dioxide Emissions for Various Fuels, Volk-
mental data. Occidental petroleum. Houston, Texas. Available at Oxy.com. er-Quaschning, 2015 Viewed 28/05/2020. Available at Volker-quaschning.de.
[22] Chevron, 2018. Sustainability report. Chevron. California, USA. Available at [40] S.T. Muench, Y.Y. Lin, S. Katara, A. Armstrong, Roadprint: practical pavement life
Chevron.com. cycle assessment (LCA) using general available data, in: International Symposium
[23] BP, 2019. Sustainability report. BP, London, UK. Available at Bp.com. on Pavement LCA, Davis, California, 2014, pp. 249–262.