Accepted Manuscript

New sustainability indices for product design employing environmental impact and
risk reduction: Case study on gasoline blends
Mohammad Hossein Ordouei, Ali Elkamel, Maurice B. Dusseault, Ibrahim Al-Hajri
PII:

S0959-6526(15)00859-8

DOI:

10.1016/j.jclepro.2015.06.126

Reference:

JCLP 5785

To appear in:

Journal of Cleaner Production

Received Date: 3 March 2015

Revised Date:

16 June 2015

Accepted Date: 26 June 2015

Please cite this article as: Ordouei MH, Elkamel A, Dusseault MB, Al-Hajri I, New sustainability indices
for product design employing environmental impact and risk reduction: Case study on gasoline blends,
Journal of Cleaner Production (2015), doi: 10.1016/j.jclepro.2015.06.126.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Word Count: 6,992

New sustainability indices for product design employing environmental impact and risk
reduction: Case study on gasoline blends

Department of Chemical Engineering, University of Waterloo, 200 University Avenue West

SC

RI
PT

Mohammad Hossein Ordouei*a, Ali Elkamela, Maurice B Dusseaultb, Ibrahim Al-Hajric

M
AN
U

Waterloo, Ontario, N2L 3G1, Canada

Department of Earth and Environmental Sciences, University of Waterloo, 200 University

Avenue West, Waterloo, Ontario, N2L 3G1, Canada

Department of Chemical Engineering, Kuwait University, P.O. Box 5969, Safat, Kuwait

AC
C

EP

TE
D

*Author to whom correspondence may be addressed:

E-mail: mhordoue@uwaterloo.ca
Phone: 1-(519)-888-4567 (Ext. 33004)
Fax: 1-(519)-888-4347

ACCEPTED MANUSCRIPT
LIST OF ABREVIATIONS

Description

AHP

Analytical Hierarchy Process

Empirical Kinetic Coefficient for Intensity of

Intermolecular Interactions

Ci

Concentration of Hydrocarbons/Oxygenates

CI

Consistency Index

CR

Consistency Ratio

Di

Dipole Moments

E0

Fossil Gasoline; No Ethanol, No Blend

SC

M
AN
U

E5

Blend: 5% Ethanol and 95% Gasoline

E10

Standard Blend: 10% Ethanol and 90% Gasoline

EtOH

Ethanol

Frequency of Accidents for Chemical i

TE
D

fi
Hi

Hazard Effects of Chemical i,

I5

Blend: 5% Isooctane and 95% Gasoline

EP

KPI

AC
C

Mj
M5

RI
PT

Abbreviation

Key Performance Indicators

Chemicals Inventory (tonne)

Blend: 5% Methanol and 95% Gasoline

MeOH

Methanol

Empirical Kinetic Coefficient for Intensity of

Intermolecular Interactions

Number of metrics

ON

Octane Number

PEI

Potential Environmental Impacts

RI

Random Index

ACCEPTED MANUSCRIPT
Risk Index Associated with Product Streams

(R.I)T

Sum of Risk Index of Product & Waste Streams

(R.I)W

Risk Index Associated with Waste Streams

RON

Research Octane Number

xi, j

Mass Fraction of Component i in Stream j (i, j = 1, 2, )

WAR Algorithm

WAste Reduction Algorithm

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

(R.I)P

ACCEPTED MANUSCRIPT
Abstract
Product life-cycle design is a pathway from concept to research and development, from
manufacturing to marketing and distribution. In the current approach to product design two
objectives are broadly of stakeholders and customers interest: product specification and
price. Sustainable product design, however, requires that environmental, energy, and safety

RI
PT

issues, among others, be addressed. This paper introduces a new methodology employing
three sustainability indices related to environment, energy and safety for a variety of
applications including product design. The indices can be used individually or together, and it
is possible to add other metrics depending on the nature of products and processes. To test the

SC

indices, blends of a typical gasoline as a product design case study were investigated.

Gasoline blends are good examples of sustainable product design. Different gasoline blend

M
AN
U

products were modeled by Aspen HYSYS in order to estimate the octane numbers. New
sustainability indices involving metrics such as octane number, mileage loss, potential
environmental impacts, and safety risk were then applied on the blends and compared to fuel
economy using the Analytical Hierarchy Process (AHP) methodology to select the best
product. The range of each metric varies from one to another depending on the nature of the
metrics, applications, regulations etc. however, AHP is used for ranking purposes in the range

TE
D

of 0 to 100%.

Key words: new sustainability measurement methodology; sustainable gasoline blend

design; environmental impacts assessment; safety risk reduction; profitability analysis;

AC
C

EP

analytical hierarchy process.

1. Introduction

Sustainability is the path of continuous improvement, wherein the products and services
required by society are delivered with progressively less negative impacts upon the Earth
(Cobb et al., 2007). The sustainability has a wide variety of important perspective ranging
from molecular level (Manley et al., 2008) to the estimation of conventional environmental,
economic, and social theoretical frameworks for teaching purposes (Carew and Mitchell,
2008).

ACCEPTED MANUSCRIPT
A sustainability framework for corporations (Amimi and Bienstock, 2014) may include the
outcomes of integration of: (a) sustainability into corporate strategy and external
communications to corporate stakeholders; (b) organizations supply chain to sustainability;
(c) economic, ecological-environmental, and equity-social concerns in strategic decisions and
design processes to name a few.

RI
PT

Few sustainability measurements appeared over the years. The Dow Jones Sustainability
Indices (2013), for instance, results in high levels of aptitude in different areas such as
strategy, financial, customer and product, governance and shareholders, and human
resources. The FTSE4Good Environmental Leaders Europe 40 Index (2015), a European-

SC

based index, provides a guideline for investors who look for European companies
demonstrated for best practice environmental management. It is used for retail and

M
AN
U

institutional investment products and actively encourages companies to be more responsible.

The AIChE Sustainability Index (SI) is based on a set of new metrics that form a consolidated
Sustainability Index or SI (Cobb et al., 2007). The data sources of this index include
companys annual sustainability report, industrial performance rankings, governments
pamphlet and newsletters. The metrics are scaled from 0 - 7 and depicted on a spider chart.

TE
D

The above mentioned sustainability indices have several advantages such as improving
companys profits; however, they are either concerned with society or have a business
approach and therefore, they are based on individual companys performance. In addition,
there are some missing key sustainability considerations in those indices, which limit their

EP

applications to chemical processes. AIChEs SI has overcome this shortcoming; however, it

needs a large number of data, which vary from a company to another based on business

AC
C

performance. Above all, none of the existing indices can be used for prototype chemical
products.

Chemical engineering has a substantial talent to achieve sustainability (Batterham, 2006).

Thus, the objective of a sustainable chemical process is to satisfy stakeholders, regulators and
society by its profitability and social responsibility while generating lower potential
environmental impacts and wastes without compromising products quality. Such designs use
minimum raw materials and minimum energy.

ACCEPTED MANUSCRIPT
This paper introduces a new set of indices for sustainability measurement of chemical
products; e.g. solvents, fossil fuels, biofuels, yet unlike afore-mentioned methodologies
require minimum available data.
Combination of sustainability with product economy will result in designing sustainable
products such as sustainable vehicles (van Lante and van Til, 2008). The idea of product

RI
PT

design does not involve design and manufacture of a commodity that already exists; it is
generally applied to an innovative product improvement or an entirely new product. The
design of a product has several objectives; to assure organizational profit, to meet market
needs, to use corporate resources optimally, to meet physical property goals, to achieve

SC

environmental protection, to reduce waste, to improve safety (user and producer), and so on.

M
AN
U

The product design process has four general steps (Reid and Sanders, 2012):

Idea development: The design of a product starts with an idea as a desirable solution to a need.

Product screening: When it comes to a number of product ideas the second step is the evaluation
of the ideas to opt the supreme likelihood to thrive.

Preliminary design and testing: At this step the first product is designed and manufactured

TE
D

followed by market and prototype analysis.

Final design: After successful marketing and commercialization of test product, the final design of
the product is made and publicized.

EP

Hence, the second step of the product design stipulates a powerful screening tool. This paper
pioneers a new methodology to estimate basic metrics such as potential environmental impact

AC
C

(PEI) of material (Cabezas et al., 1997, 1999; Young and Cabezas, 1999; Young et al., 2000),
PEI of energy (Ordouei et al., 2014a), risks to plant safety (Ordouei et al., 2014b), octane
number (Aspen Technology Inc., 2006; , Albahri, 2002; Rao, 2007; Alexandrovna and
Tuyen, 2010), mileage loss and economic analysis of a variety of different gasoline blended
with agents such as methanol (MeOH), ethanol (EtOH) and isooctane. It turns out to be
difficult to establish a clear trade-off between these metrics and economic return, so a
methodology called Analytical Hierarchy Process (AHP) was employed (Saaty, 2008;
Alonso, 2006) to estimate the total value of the above metrics for each blend as Key
Performance Indicators (KPI). Cost-benefit (in terms of KPI) analysis was carried out to rank
the blends, and the results show that methanol blended gasoline has the highest KPI and the

ACCEPTED MANUSCRIPT
lowest price among the product array, meaning that it maximizes environmental, safety and
economical factors.
A mixture of an alternative chemical, such as methanol or bioethanol, to a conventional one,
such as gasoline, by various fractions is called blend. Blends in low proportions are currently
being used by customers at pumps in large scale, which is a preliminary step to higher

RI
PT

integration prospect.

The US Environmental Protection Agency (EPA) maintains the use of gasoline blends across
the US provided that the blend is made in accordance with Clean Air Act Criteria.

SC

Californias special blend shows a decrease in emissions from vehicles compared with
regular gasoline; e.g. VOCs (volatile organic compounds) up to 29%, NOx by 6% (US

M
AN
U

Governmental Accountability Office, 2005).

There are different types of blends: gasoline and ethanol, low-level biodiesel, biodiesel (B20
and above), hydrogen and natural gas etc. (US Energy Efficiency and Renewable Energy,
2014).

TE
D

2. The impacts of gasoline blends on Octane Number (ON)

An important specification of a gasoline fuel is octane number (ON). Compressed gasolineair mixtures in an internal combustion engine should burn smoothly, but they tend to ignite

EP

prematurely during compression, creating engine knock - rattling or pinging sounds in one or
more cylinders - which can eventually damage the engine. Octane number (ON) is a metric of

AC
C

the antiknock quality of a gasoline. In the past, tetraethyl lead, Pb(C2H5)4, was added at about
2.5 grams per gallon (0.66 kg per m3) of gasoline to increase ON; however, with the shift
from leaded gasoline, more expensive compounds such as aromatics and highly branched
alkanes were added to maintain high ON values. ON is a metric indicating the resistance of a
motor fuel to pre-ignition knock (Computer Support Group Network, 2012). It ranges from
100 for isooctane (a highly branched alkane with minimal knock) to zero for heptane (a
straight hydrocarbon chain with bad knock). The ON of a straight-run gasoline is 70
(Minnesota Department of Commerce, 2004). Other octane facts include (Speight, 2007):

A gasoline with relatively high ON burns slowly, preventing engine knock.

ACCEPTED MANUSCRIPT

Engine knock is not necessarily because of the wrong gasoline formulation; there may be
problems with electronic control systems, ignition timing, exhaust gas recirculation, and
so on.

In general, branched isomers are less susceptible to premature ignition; therefore these
molecules can serve as anti-knock agents.

RI
PT

The following typical cracking reaction (among many others) takes place in a refinery:

C13H28 (alkane, liquid)
C8H18 (alkane, liquid) + C2H4 (alkene, gas) + C3H6 (alkene, gas).
During this process, isomerization reactions can be favored in order to generate more pre-

SC

ignition resistant branched isomers to serve as anti-knock agents in the gasoline.

M
AN
U

ON is an arithmetic average of the research octane number and the motor octane number,
determined by ASTM D2699 and ASTM D2700 standards, respectively. Different methods
for predicting ON of a fuel and fuel-additive mixtures have been introduced based on the
chemical properties of the mixture components. For instance, Albahri (2002) introduced a
structural group based method for ON calculation by tabulating the functional groups into
paraffins, olefins, cyclics, and aromatics, giving 33 different groups, and assigning an ON

TE
D

value to each group. Then, the following empirical equation was used to calculate a value for

EP

ON.

Rao (2007) asserted that the ONs of hydrocarbon mixtures are not additive, and therefore

AC
C

they cannot be calculated by summation of the ON of individual hydrocarbons. He reported

that the ON of a hydrocarbon is influenced by other properties such as boiling point, density,
molecular structure, the number of branches in an isomer, and molecular weight.
Alexandrovna and Tuyen (2010) introduced the following model based on different naphtha
process streams (reformates, isomerates, straight-runs and alkylates):

where

ACCEPTED MANUSCRIPT
Bi = . Din

(3)

Ci is the concentration of hydrocarbons and oxygenates, and and n are empirically

determined kinetic coefficients representing the intensity of intermolecular interactions due to
dipole moments Di.

RI
PT

Aspen HYSYS (Aspen Technology Inc., 2006) has a proprietary method for the calculation
of ON, which was used for this study.

SC

3. The impacts of blending ethanol and gasoline on mileage

There are different commercial choices of gasoline blends - regular, premium, mid-grade,

M
AN
U

super, ethanol blends in various percentages, and so on. The blends with higher octane
numbers are more expensive, by as much as 12-13% (Minnesota Department of Commerce,
2004).

Organically-sourced ethanol produced from plants, typically sugar cane or corn, is called bioethanol. With no taxes, royalties, and perhaps even with subsidies, it may be economic and

TE
D

suitable for blending with gasoline. Blending of ethanol and gasoline in the US was studied
by NREL (National Renewable Energy Library, 2008) and the following results were
published:

EP

a) Over 70% of gasoline at the pump in the US contains ethanol.

b) The standard E10 blend contains 10% ethanol and 90% gasoline and is sold as regular
gasoline.

AC
C

c) Renewable Fuel Standards legislation passed as an amendment to the Energy

Independence and Security Act (EISA) demands E10 as the minimum ethanol grade
gasoline in the US until 2022.
d) Up to 20% of ethanol blended (E20) into gasoline apparently presents no problems for the
current vehicle fleet or fuel dispensing devices.
e) Substitution (blending of ethanol and gasoline) has adverse effects on mileage because of
the lower energy content of ethanol, approximately 67% the energy density of gasoline.
E10, E15, and E20 blends have mileage reductions of 3.9%, 5.0% and 7.7%, compared
with E0. However, ethanol addition decreases the retail gasoline price in the US by
almost 17% compared with E0.

ACCEPTED MANUSCRIPT
f) US energy policy encourages using blended gasoline through mandates and subsidies,
whereas fossil fuels have additional taxes, different in each state.

RI
PT

4. The effects of ethanol, methanol, isooctane on the octane number of gasoline

blends
The blending of bio-ethanol and gasoline is economically beneficial for oil companies and
reduces the fossil fuel consumption. In this section, a typical gasoline composition is adapted
from Lin and Chou (1995), and the ON of E10, E15 and E20 was estimated through the

SC

proprietary model introduced by the Aspen HYSYS process simulator to be 73.8, 73 and 72,
respectively (please see the following section). The ON of the reference chemical agent

Information Administration, 2002).

M
AN
U

(isooctane) equals 100, whereas a more precise number is apparently about 98 (US Energy

In this research, methanol (MeOH) and isooctane were probed as substitutions to EtOH at a
flow rate of 5 kg/h blending with gasoline at different flow rates from 85 kg/h to 95 kg/h.
Figure 1 shows the results of calculations of octane numbers.

TE
D

The rationale underlying the study of effects of 5% of chemical agents on the blends is that in
Canada bio-ethanol and gasoline are blended with the same ratio (E5) by law (Financial Post,
2012).

EP

At 5 wt% of each chemical agent blended with 95 wt% gasoline, MeOH blend (M5) has the
highest ON among EtOH (E5) and isooctane (I5) blends. The small differences between all
three blends in ON are due to using 5% chemicals for each. This means 95% fossil gasoline

level.

AC
C

for each blend wouldnt make significant variations in the result at almost 95% confidence

Historically, up to 2010, the isooctane production rate in the USA was about 2,210 m3
(13,900 barrels) per stream day, but it has dropped dramatically down to 254 m3 (1,600
barrels) per stream day in the space of two years followed by another plunge to 32 m3 (200
barrels) per stream day in the years 2013 and 2014 (EIA, 2015). In such a huge shortage of
isooctane, methanol and ethanol are key alternatives due to their relatively cheap price and
availability in large scale from petrochemical industries and from waste recovery plants
where biomethanol and bioethanol are produced from agricultural wastes.

10

ACCEPTED MANUSCRIPT
5. The impacts of E5, M5 and I5 on heating value, mileage and price
Comparison of mileage loss of the blends (i.e. E5, M5 and I5) with pure gasoline was made
based on their heating values. Table 1 lists the heating value of the pure chemical, the blends,
and the mileage loss corresponding to each blend. As can be seen, M5 has the least heating
value and the highest mileage loss. This means that a certain amount of the pure gasoline that

RI
PT

would drive a car 100 km, E5 and M5 would achieve 98.2 km and 97.4 km, respectively.
The economic evaluation is presented in Table 2 (EIA, 2014; Business Analytic Center,
2014; ICIS, 2014), which compares the chemical prices on the basis of the following

M
AN
U

SC

equation:

where subscript i denotes the blend components and xi represents the mass fraction of
component i.

Thus, the blends price (I5, E5 and M5) are 1,012 US$/m3, 956 US$/m3 and 948 US$/m3,
respectively (Table 3). In this study, the average prices of pure chemicals have been used for

TE
D

calculations.

The price comparison of the blends is shown in Figure 2, where M5 is seen to have the lowest
price. At this point, one may suggest that M5 is the best choice among other options due to its
higher octane number and lower price, but other factors arise; i.e. the sustainability of the

EP

new product design, M5. In the succeeding sections, the new sustainability indices and multi-

AC
C

criteria decision-making process are discussed in detail.

6. Impacts of E5, M5 and I5 on environment in Potential Environmental Impacts

(PEIs)
Perhaps most important, an energy impact index and safety risk index, as well as the material
impact index from the WAste Reduction (WAR) algorithm, should be applied to the blends in
order to estimate total energy and material impacts on the environment and to assess the
safety risks of the blends. The WAR algorithm estimates the environmental impacts of
materials and employs eight impact categories under two main groups (Cabezas et al., 1997,
1999; Young and Cabezas, 1999; Young et al., 2000):

11

ACCEPTED MANUSCRIPT

Local Toxicological Impacts: Human Toxicity Potential by Ingestion (HTPI),

Human Toxicity Potential by Exposure (HTPE), Terrestrial Toxicity Potential (TTP),
Aquatic Toxicity Potential (ATP).

Global Atmospheric Impacts: Global Warming Potential (GWP), Ozone Depletion

Potential (ODP), Photochemical Oxidation Potential (PCOP), Acidification Potential

RI
PT

(AP).
Ordouei et al. (2014a) have introduced a simple methodology for the estimation of the
impacts of the energy consumption on the environment due to fossil fuel combustion in terms
of mass CO2, SO2 and NO2 gases per unit time. In this study, Aspen HYSYS process
gasoline with afore-mentioned chemicals (Figure 3).

SC

modeling simulator (Aspen Technology Inc., 2006) has been used to model the blending of

M
AN
U

In this model, the impact of power/energy consumption, Ie, can be ignored as all chemicals
and blends use almost the same energy for transportation, meaning that the difference in the
energy impacts of blends on the environment is negligible (i.e. not applicable in this case):
Ie NA (Not Applicable)

(5)

TE
D

Ie represents the impacts of energy consumption within a process on the environment. Hence,
we will focus only on the impacts of the blends on the environment. These are calculated in
PEI/h using CHEMCAD (Cabezas et al., 2009) and are shown in Figure 4. As can be seen, M5

EP

has the lowest impact on the environment.

AC
C

7. The impacts of E5, M5 and I5 on safety risk

The risks associated with each blend were estimated in Number of Affected People per
Year based on the methodology originally presented by Al-Sharrah et al. (Al-Sharrah et al.,
2007) and improved later by Ordouei et al. (2014b). A chemical process plant may have
multiple product streams; therefore, the risk of all product streams can be estimated as
follows:

(. ) =

   ,

(6)

Similarly, the risk of all waste streams can be calculated by the following equation:

12

ACCEPTED MANUSCRIPT

(. ) =

   ,

(7)

R.I is an abbreviation for Risk Index, superscripts P and W denote product and waste streams,
respectively. (R.I)P and (R.I)W express the impacts of the calculated risks in number of
affected people per year, which is the maximum potential risks attributed to the total product

RI
PT

streams and the total waste streams, respectively. Subscripts i and j designate the chemicals
within the streams and the streams within the process, respectively. Mj stands for the mass
in tons of chemicals released to the environment (inventory), which can be calculated from the
design basis. Hi denotes the hazard effects of chemical i, in number of people affected per ton
of chemical released to the environment. xi, j is the mass fraction of component i in stream j (i,

SC

j = 1, 2, ). Finally, fi represents the frequency of accidents for chemical component i in

M
AN
U

number of accidents per year.

The total risk, (R.I)T, is the sum of the equations (6) and (7):

(R.I)T = (R.I)P + (R.I)W

(8)

The toxicity and frequency of all compositions of each blend from EPAs databanks and the

TE
D

associated inventory have been used to calculate R.Is. More details about the methodology
and databanks have been provided by Ordouei et al. (2014b). The results of the calculation of
risk indices are shown in Figure 5. We note that the safety risk indices for both M5 and I5 are

EP

the same, whereas there is a decrease in risk by about 5% for E5.

AC
C

8. Selecting the best blend through the Analytic Hierarchy Process (AHP)
methodology
The results from the previous sections are summarized in Table 4. Clearly, isooctane is not
the best substitute for increasing ON only. Price, environmental impact and safety have to be
taken into account. Thus, isooctane can be replaced by better alternatives and fortunately,
such alternatives do exist as discussed in Section 4.
Table 4 contains factors which have positive or negative influences on the blends qualities.
For instance, a higher ON is considered a positive factor affecting the life time of a car
engine, whereas a higher PEI value has adverse environmental consequences. Moreover, PEI
values are much higher than the value of other factors so that a method of including all of the

13

ACCEPTED MANUSCRIPT
impact factors in an appropriate manner is needed. Thus, the objective of this section is to
rank the factors in Table 4 in an appropriate way. The Analytic Hierarchy Process (AHP) is a
mathematical methodology for weighting and ranking of the metrics as well as multi-criteria
decision making (Saaty, 2008). The AHP is a comparison based methodology in the range
from 1/9 to 9 where 1/9 is assigned for least importance and 9 for outright superiority. All

RI
PT

other odd and even numbers fall in between. The inputs can be either of actual measurement;
e.g. price, weight, and subjective opinion such as preference. The outputs; however, may be
ratio scales from Eigenvector method and/or consistency index from Eigenvalue method.
There are several applications of AHP including evaluation of product features, selection

SC

from strategic planning alternatives, screening a set of Key Process Indices, making
integrated decisions with multiple outputs from different shareholders. In this paper, the

Objective (design and selection of the most sustainable gasoline blends among I5, E5, and
M5).

Criteria or in our case, metrics such as environment impacts, risk assessment, energy
impacts, ON, mileage loss.

Sub-criteria (eight categories of environmental impacts etc.).

TE
D

M
AN
U

collected data have been organized in different levels (Figure 6) as follows:

Configuration of a hierarchy starts from objective; e.g. gasoline blends, then by metrics
(criteria); e.g. safety risk, environmental impact and energy impact, and finally, by sub-

EP

criteria; e.g. the values of each metric. In mathematical language, let {A1, A2, , An} and {w1,
w2, , wn} be metrics and current weights, respectively, where n denotes the number of

AC
C

metrics. The matrix W represents the ratios of all weights (Alonso, 2006):

14

ACCEPTED MANUSCRIPT
a) Pair wise comparison of elements in each criteria and building matrix: The matrix A=[aij]
exhibits the referees (experts) preference within pairwise comparison of criteria and subcriteria (ai vs. aj for all i,j = 1,2,,n) where aij is positive and reciprocal (i.e. aij > 0 and aij

M
AN
U

SC

RI
PT

= 1/ aji).

The aij elements are the referees estimates of the ratios wi/wj (i.e. aij = wi/wj) for i,j =
1,2,,n. So, the matrix W consists of existing weights of metrics and therefore, it is our

TE
D

objective to find it by eigenvector method.

The comparisons are scored in the range from 1/9 (for least importance) to 9 (for outright
superiority). All other odd and even numbers (1/8, 1/7, , 7, 8) fall in between.

EP

b) Calculation of the sum of each column (SCi).

AC
C

c) Normalization of the matrix by dividing each elements of the column by the sum of the
same column.

d) Averaging of each row (xi) of the normalized matrix. xis represent the Eigenvector or the
vector of the relative weights of metrics in the given level of hierarchy.
e) Determine Eigenvalue, max:

max= SCi xi = SC1 x1 + SC2 x2 +

(9)

15

ACCEPTED MANUSCRIPT
where SCi is the summation product of column i of original matrix and xi is summation
product of row i of matrix A.
f) Consistency analysis: Both matrices A and W are equal when matrix A is absolutely
consistent then, max = n where n is the number of matrix and Aw = max w. However, the

RI
PT

referees (experts) judgments are not perfect in almost all cases. In other words, max n
when matrix A is not absolutely consistent so, the consistency index (CI) and the level of
inconsistency (CR or Consistency Ratio) have to be measured.

M
AN
U

CI = (-n) / (n-1)

SC

g) Calculate the consistency index (CI) using the following equation:

(10)

The term ( n) is called the consistency condition, which is used in calculating the

TE
D

consistency index. CI is similar to standard deviation of error estimation and mean deviation.

EP

h) Verification of the consistency ratio through the following equation:

< 0.1 for all matrices A with n>3 and

(11)

AC
C

CR = CI / RI

< 0.08 for n = 3 to yield satisfactory results

Random index (RI) is the mean deviation of selected comparison values by chance from true
ones (Table 7).
i) Continuing steps 1 to 8 above for sub-criteria.
The results are summarized in Table 5, in which the highest priority belongs to environmental
impacts, then to safety risk index, followed by octane number, and subsequently mileage loss.
Also, the price is not considered in the analysis at this stage because the ranking of economic

16

ACCEPTED MANUSCRIPT
factors is contentious (some always place it first, some always place it last). To make an
unbiased analysis, economic issues are addressed in a follow-up cost-KPI analysis, as will be
discussed later.
MakeItRational Analytical Hierarchy Process Software is a user friend software package to

RI
PT

use AHP methodology. A quick demo has been provided in the following link:
http://makeitrational.com/analytic-hierarchy-process/ahp-software#!prettyPhoto

SC

9. Result

The weights and the ranks of the three major factors are shown in Table 5, yet it is difficult to

M
AN
U

compare the blends based on this table because the contribution of each metric to the values
in Table 4 is unidentified. For a better verdict with respect to the positive and the negative
effects of these factors, the values of similar factors; e.g. PEI, in all blends were compared as
per AHP methodology to develop Table 6, where the breakdown of the weight of each factor
has been elucidated.

TE
D

It is now possible to ascertain the blends in terms of the Key Performance Indicator (here one
may take KPI as a proxy for net benefits or utility) based on the data of Table 6. For instance,
The KPI of isooctane blend (I5) follow:

EP

ON=76, Mileage loss = 0.28, PEI/h = 229.4 and safety risk index = 2.21; thus:

AC
C

KPI = 2% + 5% + 10.6% + 4.2% = 21.8%

(12)

Similarly, the KPI of E5 can be calculated:

ON=75, Mileage loss= -1.83, PEI/h = 229.9 and safety risk index = 2.1; thus:
KPI = 2% + 2% + 4.6% + 16.6% = 25.2%

(13)

As such, the KPI of M5 is:

ON=80, Mileage loss = -2.65, PEI/h = 228.5 and safety risk index = 2.21; thus:

17

ACCEPTED MANUSCRIPT
KPI = 8% + 1% + 39.8% + 4.2% = 53.0%

(14)

The final step is to perform cost-KPI (similar to cost-benefit) analysis using the information
about blend prices in Table 3 as the input cost and the KPIs (as expressions of the utility or
benefit) as shown in Figure 7. This figure shows that the cost-KPI for each blend is a useful
screening tool for decision makers to pick the best blend among several alternatives. In this

RI
PT

case, M5 (35.9, 53) is the best overall and then, E5 (36.2, 25.2), and finally I5 (38.7, 21.8).
Ignoring the new risk index leads to a deceptive ranking of the blends if only the existing
WAR algorithm is taken into account (Figure 8).

SC

In such case, the new ranking starts with M5 (35.9, 64.5) then, I5 (38.7, 23.8) and finally E5
(36.2, 11.7) since KPI for I5 is doubled compared to that of E5. Even though the price of I5 is

M
AN
U

higher than that of E5 by 66 US$/m3, it accounts for a higher ON and lower mileage loss.
Although only three blends have been analyzed M5, E5 and I5 it is clear that the proposed
methodology is effective in choosing the best blend with the metrics that are available
through analysis and cost data. It is also necessary to apply care in the choice of new metric;

10. Conclusions

TE
D

i.e. new risk index, as adding or dropping it will lead to different results.

New Sustainability Indices were presented in this paper to evaluate different blends. The

EP

indices employ three main metrics to evaluate the friendliness to the environment and to measure
the safety risks; however, depending on the nature of products considered, other metrics may be

AC
C

used such as mileage loss and octane number as in the present case study. The three principal
metrics of the New Sustainable Indices are:

An existing methodology for the estimation of the Potential Environmental Impacts of

materials in PEI/h, called WAR algorithm.

A new methodology for the estimation of the impacts of energy consumption by a process on
the environment in kg CO2 (or SO2 or NO2)/h.

A new risk assessment methodology for inherent safer design (ISD) in the number of affected
people/year.

The second and the third new methodologies have been introduced by Ordouei et al. (2014 a,b).

18

ACCEPTED MANUSCRIPT
It is strongly recommended to employ all of the principal metrics of the new sustainability indices
as they can resolve the current environmental and safety issues at early stage of a chemical
process design and retrofit (conceptual design) as well as new chemical product design.

The new methodology of sustainability measurement is derived by integrating of all aforementioned indices into Analytical Hierarchy Process (AHP) methodology in order to

RI
PT

calculate the Key Performance Indicators (KPI), which has a technical concept in this paper.
As a case study, the blends of a typical fossil fuel (gasoline molecules) and three chemicals
(methanol, ethanol and isooctane) were modeled by Aspen HYSYS and the factors affecting
the quality of 5% blends were studied for the specific M5-E5-I5 set. The most important

M
AN
U

environment, safety risk and economic factors (costs).

SC

metrics are octane number, mileage loss, the impacts of the energy and material on the

Since the energy consumption was almost equal for all blends, the corresponding methodology
was ignored for simplicity. All factors except cost were weighted according to the Analytical

Hierarchy Process methodology in order to rank the blends in terms of Key Performance
Indicators (KPI), followed by a cost-KPI analysis.

The cost-KPI analysis indicated that methanol blend (M5) was the best substitution for pure

TE
D

gasoline. It is even more economical and environmentally friendlier than the equivalent
ethanol blend E5. Further study on the application of risk index metrics showed that M5 is
also inherently safest blend. The risk index has a significant consequence in ranking the

EP

blends.

The new methodology for sustainability measurement has several advantages as follow:

AC
C

1. Simple and user friend and based on the reliable databanks prepared by US EPA.
2. Quantitative and mathematical model for risk and environmental impacts evaluation.
3. The indices require relatively readily available data, such as the composition and the
flow rate of process streams.

4. It is applicable to both the early stages of product design where there is a lack of
process data and existing product for better comparison.
5. The methodology is not limited to chemicals but can be extended to fossil fuels,
biofuels, syngas etc.
6. They can be used as a strong screening tool for chemical engineers, product designers,
decision makers and regulators.

19

ACCEPTED MANUSCRIPT
7. They can be used as commercial tool for publicity to convince the regulators and
potential clients for their commitment to sustainable products manufacturing.
8. Since the most convenient step for source/hazards reduction is the conceptual design,
the new sustainability indices help to reduce capital costs by a decrease in hazards
generation and sustainable products.

RI
PT

9. It provides detailed information in terms of the environmentally friendliness of

products.

10. The new methodology provides a technical concept called Key Process Index (KPI)
enabling the product designer to perform cost-benefit analysis for ranking products,

SC

which demonstrates that a sustainable product is coined to an economic design.

The new sustainability methodology has some potential areas of improvement such as

M
AN
U

screening tool for investors in a new product or even a new chemical process by
incorporating all indices into a software package. Then, the investors wont need to have
technical knowledge such as estimation of the impacts of materials and energy on
environment, risks to the process safety and even knowledge about the process itself. All they
need is to enter the data of the composition and the conditions of the streams to the software.

AC
C

EP

product.

TE
D

These data are available for several proposals through a public tender for the new process or

20

ACCEPTED MANUSCRIPT
References
Albahri T. A. Structural group contribution method for predicting the octane number of pure
hydrocarbons and their mixtures. Fuel Chemistry Division Preprints. 2002; 47 (2), 531-532.
Alexandrovna S. J. and Tuyen D. C. Development of a detailed model for calculating the

RI
PT

octane numbers of gasoline blends. Strategic Technology IFOST 2010. 2010; 430-432.
Alonso J. A. Consistency in the analytic hierarchy process: a new approach. International
Journal of Uncertainty, Fuzziness and Knowledge-Based Systems. 2006; 14 (4); 445-459.

planning. Trans IChemE. 2007; 85 (B6), 533-540.

SC

Al-Sharrah G. K., Edwards D., Hakinson G. A new safety risk index for use in petrochemical

Amimi M. and Bienstock C. C. Corporate sustainability: an integrative definition and

Production. 2014; (76), 12-19.

M
AN
U

framework to evaluate corporate practice and guide academic research. Journal of Cleaner

Aspen Technology Inc., Aspen HYSYS 2006 (Version 2006: 20.0.0.6728), A chemical
process simulator. Aspen Technology Inc. 2006.

TE
D

Batterham R.J. 2006. Sustainability-The next chapter. Chemical Engineering Science 61;
4188-4193
Business Analytic Center. Isooctane (CAS 26635-64-3) market research report 2014. Market
(May 18, 2014)

EP

Publishers (Market Report Database). ID: I96EB7C2124EN.

http://pdf.marketpublishers.com/trade_cas/isooctane_26635-64-3_market_research_report.pdf

AC
C

Cabezas H., Bare J. C., and Mallick S. K. Pollution prevention with chemical process
simulators: the generalized waste reduction (WAR) algorithm. Computers in Chemical
Engineering. 1997; 21, S305-S310.
Cabezas H., Bare J. C., and Mallick S. K. Pollution prevention with chemical process
simulators: the generalized waste reduction (WAR) algorithm-full version. Computers in
Chemical Engineering. 1999; 23, 623-634.
Cabezas H., Young D. M., Mallick S. K., and Bare J. C. CHEMCAD 6.2.0.3348. Chemstation
Inc. 2009.

21

ACCEPTED MANUSCRIPT
Carew A.L. and Mitchell C.A. Teaching sustainability as a contested concept: capitalizing on
variation in engineering educators conceptions of environmental, social and economic
sustainability, Journal of Cleaner Production. 2008; (16), 105-115.

Cobb C., Schuster D., Beloff B., Tanzil D. Benchmarking Sustainability. Chemical

RI
PT

Engineering Progress. 2007; 103, 38-42.

Computer Support Group Network (Csgnetwork). An online calculator for the determination
of the minimum octane rating (Knock Index). 2012:

(Sept. 29, 2014)

Dow Chemical Company (July 23, 2013):

SC

http://www.csgnetwork.com/octaneratecalc.html

M
AN
U

http://www.sustainability-indices.com/sustainability-assessment/corporate-sustainability.jsp
Energy Information Administration (EIA). Petroleum and other liquids. 2014.
http://www.eia.gov/oog/info/twip/twip.asp

(May 18, 2014)

Energy Information Administration (EIA). US Refinery isooctane production capacity. 2015.

TE
D

(May 18, 2015)

http://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=8_NA_8PIO_NUS_BSD&f=A

EP

Financial Post. 2012.

(Sept. 29, 2014)

FTSE4Good Environmental Leaders Europe 40 Index (2015):

(May 19, 2015)

AC
C

http://opinion.financialpost.com/2012/01/31/the-cost-of-ethanol/

http://www.ftse.com/products/home

ICIS, trusted market intelligence for the global chemical, energy and fertilizer industries.
2014.
http://www.icis.com/chemicals/channel-info-chemicals-a-z/#top

(May 18, 2014)

Lin T.B. and Chou T. C., Pd migration. 1. A possible reason for the deactivation of pyrolysis
gasoline partial hydrogenation catalysts. Industrial and Engineering Chemistry Research.
1995; 34 (1), 128-134.

22

ACCEPTED MANUSCRIPT
Manley J. B., Anastas P. T. and Cue Jr. B. W. Frontiers in Green Chemistry: meeting the
grand challenges for sustainability in R&D and manufacturing. Journal of Cleaner
Production. 2008; (16), 743-750.
Minnesota Department of Commerce. Consumer guide: gasoline octane facts and myths.

RI
PT

2004.
http://mn.gov/commerce/weights-and-measures/images/OctaneFacts.pdf

(Sept. 29, 2014)

NREL (National Renewable Energy Library, a national laboratory of the US department of

energy, office of energy efficiency and renewable energy). The impact of ethanol blending on

SC

US gasoline prices. Subcontract report NREL/SR-670-44517. November 2008.

M
AN
U

Ordouei M. H., Elkamel A., and Al-Sharrah G. A simple new index for tracking the energy
efficiency of process operations. Environmental Engineering & Management. 2014a; (In Press).
Ordouei M. H., Al-Sharrah G. and Elkamel A. A new simple index for safety risk assessment at
early stage of a chemical process design. Chemical Engineering Science. 2014b; 119; 218-229.

Rao P.K. Relation between knock and physical properties explored. Hydrocarbon

TE
D

Processing. 2007; pp. 89-97.

Reid R.D. and Sanders N.R. Operations Management, 5th Edition. Chapter 3: Product Design
and Process Selection. New York. Wiley Plus. 2012.

EP

Saaty T.H. Decision making with the analytical hierarchy process. International Journal of
Services Sciences. 2008; 1 (1), 83-98.
Speight J.G. The Chemistry and Technology of Petroleum. 4th ed. CRC Press: Taylor and

AC
C

Francis Group. 2007.

US Energy Efficiency and Renewable Energy, Department of Energy. Alternative fuels data
center. 2014.

http://www.afdc.energy.gov/fuels/blends.html

(Nov. 29, 2014)

US Governmental Accountability Office (GAO). Gasoline Markets, special gasoline blends

reduce emissions and improves air quality but complicate supply and contribute to higher
process. June 2005. GAO-05-421.

23

ACCEPTED MANUSCRIPT
http://www.wiley.com/college/reid/0471320110/pdf/i_ch03.pdf

(Nov. 29, 2014)

Van Lante H. and van Til J. I. Articulation of sustainability in the emerging field of
nanocoatings. Journal of Cleaner Production. 2008; (16), 967-976.
Young D. M., Cabezas H. Designing sustainable processes with simulation: the waste

RI
PT

reduction (WAR) algorithm, Computer and Chemical Engineering. 1999; 23, 1477-1491.
Young D., Scharp R., Cabezas H. The waste reduction (WAR) algorithm: environmental
impacts, energy consumption, and engineering economics. Waste Management. 2000; 20,

AC
C

EP

TE
D

M
AN
U

SC

605-615.

24

ACCEPTED MANUSCRIPT
TABLES
Table 1.Comparison of the heating values of chemicals used in this case study.
Heating Value of
Heating Value of 5%
Chemical
Mileage Loss, %
Chemicals, kJ/kg
Chemicals in Blends, kJ/kg
4.23 x 104

Ethanol

2.68 x 104

4.14 x 104

Methanol

1.99 x 104

4.11 x 104

Pure Gasoline

Reference

4.22 x 104

0.28

RI
PT

4.47 x 104

-1.83

-2.65
0

M
AN
U

SC

Isooctane

Table 2. Price comparisons of pure gasoline and other chemicals used in this case study.
Price, US$/m3

Reference

Gasoline

970

(EIA, 2014)

Isooctane
Ethanol

TE
D

Chemical

(Business Analytic Center, 2014)

691

(ICIS, 2014)

534

(ICIS, 2014)

AC
C

EP

Methanol

1,825

Table 3. Comparison of the prices of blends used in this case study.

Gasoline Blend

Price, US$/ m3

Isooctane

1,012

Ethanol

956

Methanol

948

ACCEPTED MANUSCRIPT

M
AN
U

SC

Energy Impacts, PEI/Year

RI
PT

Table 4. The summary of the blends metrics.

Metrics
Blends, 5%
Isooctane
EtOH
MeOH
Octane No.
76
75
80
Mileage Loss, %
0.28
-1.83
-2.65
3
Price, US$/ m
1,012
956
948
Environmental Impacts,
229.4
229.9
228.5
PEI/h
Safety Risk Index,
2.21
2.10
2.21
No. of Affected
People/Year

Table 5. The result of weighting and ranking of factors.

Analytical Hierarchy Process
Octane No.
Mileage Loss

EP

TE
D

Environmental Impacts
Safety Risk Index

Weight

Rank

12%

8%

55%
25%

1
2

Table 6. The total impacts of each metric for cost-KPI analysis.

Mileage Loss, 8%
-2.56
1%
-1.83
2%
0.28
5%

AC
C

Octane No., 12%

75
2%
76
2%
80
8%

PEI/h, 55%
228.5
39.8%
229.4
10.6%
229.9
4.6%

Safety Risk, 25%

2.21
4.2
2.10
16.6
2.21
4.2

Table 7: Random Index (RI) used in Analytical Hierarchy Process (AHP).

n' a
RI
a

1-2
3
4
0.0
0.58
0.9
n = Number of metrics

5
1.12

6
1.24

7
1.32

8
1.41

9
1.45

10
1.49

ACCEPTED MANUSCRIPT
FIGURES

83

84

82

80
78

75

76

74
73

74

RI
PT

80

76

Methanol, 5 kg/h

74

Ethanol, 5 kg/h

72.5

Isooctane, 5 kg/h

72

SC

Research Octane No., RON

82

70

M
AN
U

68
66
85

90

95

Gasoline, kg/h

TE
D

Figure 1. Comparing the octane number of gasoline and MeOH, EtOH, and isooctane blends.

956
948

Price, US$/m3

AC
C

EP

1012
1020
1010
1000
990
980
970
960
950
940
930
920
910

Isooctane

Ethanol

Methanol

Figure 2.Comparison of the price of gasoline blends with 5 wt% of isooctane, ethanol and
methanol (i.e. I5, E5 and M5).

SC

RI
PT

ACCEPTED MANUSCRIPT

M
AN
U

Figure 3. Modeling of gasoline and methanol blends with HYSYS 2006.

229.5
229
228.5

228.5

AC
C

228

229.4

PEI/h

EP

230

TE
D

229.9

227.5

Methanol Blend

Ethanol Blend

Isooctane Blend

Figure 4. The comparison of impacts of gasoline blends with 5 wt% of isooctane, ethanol
and methanol (i.e. I5, E5 and M5) on the environment.

ACCEPTED MANUSCRIPT
2.24

2.21

2.21

2.10

2.12

Safety Risk Index,

No. Affected People/Year

2.06

2.00
Isooctane
Blend

SC

MeOH Blend EtOH Blend

RI
PT

Axis Title

2.18

AC
C

EP

TE
D

M
AN
U

Figure 5. Safety risk index for gasoline blends with 5 wt% of isooctane, ethanol and
methanol (i.e. I5, E5 and M5).

Figure 6. Employing AHP methodology for selection of sustainable gasoline blend.

ACCEPTED MANUSCRIPT
60

1020

1012

980

956
30

948
21.8

960

25.2

20

940

10

920

900

Isooctane

EtOH

RI
PT

40

Price, US$/m3

Key Process Index

SC

1000

Price, US$/m3

50

M
AN
U

Key Process Index

53

MeOH

1012

60

EP

40
30

1020

1000

980

956

AC
C

Key Process Index

50

64.5

948

960

23.8

Price, US$/m3

70

TE
D

Figure 7. Cost-Benefit (KPI) analysis of three gasoline blends with 5 wt% of isooctane,
ethanol and methanol (i.e. I5, E5 and M5).

Price, US$/m3
Key Process Index

940

20
10

920

11.7

900

Isooctane

EtOH

MeOH

Figure 8. Cost-KPI analysis of three gasoline blends with 5 wt% of isooctane, ethanol and
methanol (i.e. I5, E5 and M5) ignoring risk to safety.

ACCEPTED MANUSCRIPT
Highlights:

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

A new sustainability quantification methodology containing three indices was

developed
The methodology ascertains sustainable processes without compromising economy
The indices can be used at early stage of product formulation in lack of detailed data
They calculate risk, CO2 emission, environmental & energy impacts of process arrays
They were successfully examined by biofuel-gasoline blends to find sustainable blend