Optimization of recovery processes for multiple economic and environmental criteria

This study focuses on process optimization for multiple economic and environmental criteria, or otherwise termed as sustainability criteria.. Applications that considered only environmen

LEE SU-QI ELAIE

ATIOAL UIVERSITY OF SIGAPORE

OPTIMIZATIO OF RECOVERY PROCESSES FOR MULTIPLE

ECOOMIC AD EVIROMETAL CRITERIA

LEE SU-QI ELAIE

A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF EGIEERIG

DEPARTMET OF CHEMICAL AD BIOMOLECULAR

EGIEERIG

ATIOAL UIVERSITY OF SIGAPORE

2009

This Masters thesis would never have been possible without the encouragement given by the people around me It gives me great pleasure to be able to express thanks for their unconditional support

Firstly, I would like to extend my gratitude to my supervisor, Prof G P Rangaiah, from the Department of Chemical and Biomolecular Engineering in the National University of Singapore (NUS) I am awed with his drive for research, and his constant advice and guidance have brought me to where I am today He has given me many opportunities to pursue several works and to prepare manuscripts, pushing me to

my maximum potential I am indebted to him more than he knows

My lab mates – Masuduzzaman, Suraj, Lakshmi, Haibo and Sharliza – have also made my journey in NUS meaningful The technical knowledge exchanged between one another enhanced our research capabilities Also, the conversations and jokes shared have made the office livelier and the journey of pursuing graduate studies endurable Special mention goes to my friends – Stella, Ming Juan and Phyllis With them, I could always relive the undergraduate moments which were the most eventful moments of my life

Last, but not least, I would like to thank my parents and my boyfriend, Jia Le When I am feeling down, their patience and understanding would lift my spirits up miraculously Their unwavering love and concern have enabled me to complete my thesis

Elaine Lee January 2009

4.2.2 Biotechnology, Pharmaceutical and Chemicals 64

5.2.3.2 Case B: Optimization for Many Objectives 94 5.2.3.3 Case C: Optimization for Several Objectives 98

5.2.4.2 Case E: Optimization for Many Objectives 105 5.2.4.3 Case F: Optimization for Several Objectives 108

5.3.2.1 Case G: Bi-objective Optimization 112 5.3.2.2 Case H: Optimization for Many Objectives 116

6.2.2 Ranking of Solutions for Solvents Recovery 129

A.1 Excel, Visual Basic for Applications and HYSYS Interface 152

B.1 Net Flow Method in Visual Basic for Applications 159

Summary

“Sustainable development is development that meets the needs of the present, without compromising the ability of future generations to meet their own needs” as given by World Commission on Environment and Development There are three spheres of sustainability – economic development, environmental stewardship and societal equity This is often touted as the “triple bottom line” Of the three, only the first two are quantifiable based on process design and operating variables

While economic criteria such as profit before taxes, payback period and net present worth are well established, environmental objectives are still novel and there is no general consensus on the method for calculating the environmental index Environmental indices can be measured via environmental metrics or environmental impact indices For the former, it mainly comprises ratios that indicate the efficiency of the plant in terms of production or energy For the latter, many contributing factors have been identified for environmental impacts: impact on humans, ecosystem – terrestrial and aquatic, global warming, ozone depletion, photochemical oxidation, acid rain and eutrophication Several aggregation methods for the environmental indicator have also been discussed

This study focuses on process optimization for multiple economic and environmental criteria, or otherwise termed as sustainability criteria The different process applications that have been studied for both economic and environmental criteria are reviewed Applications that considered only environmental criteria are also included

as it is of interest to identify the different methods that have been used to quantify the environmental performance of a process Many of the previous studies that employed environmental indices for optimization, used aggregated environmental index as the

Hence, feasibility and usefulness of process optimization for more than two economic and environmental objectives are studied in this work

Two recovery processes have been selected for the optimization using sustainability criteria They are: a VOC (volatile organic component) recovery system and a solvent recovery system These processes are optimized for both economic and environmental objectives using the elitist non-dominated sorting genetic algorithm For the environmental objectives, the contributing factors to the environmental impacts are optimized individually or grouped into a few indices where appropriate The Pareto-optimal solutions are obtained to elucidate the trade-offs present, and the decision maker would be better equipped in choosing one of them for implementation Thereafter, net flow method is then used to identify the preferred Pareto-optimal solution based on the preferences declared by the decision maker The preferences provided by the decision maker should be more objective since s/he is aware of the quantitative trade-offs present

in the objective functions Insights gained from considering a number of environmental objectives for process optimization are highlighted Conclusions and recommendations for further research are provided at the end of the thesis

omenclature

AEP = Annual Equivalent Profit ($/yr)

AHI = Atmospheric Hazard Index

AHP = Analytic Hierarchy Process

AP = Acidification Potential

ATMP = Atmospheric Potential

ATP = Aquatic Toxicity Potential

BOD = Biochemical Oxygen Demand

C[i,j] = Global Concordance Index

CAHi = Chemical Atmospheric Hazard for Chemical i

CCP = Cumulative Cash Position ($)

CCR = Cumulative Cash Ratio

CFCs = Chlorofluorocarbons

ck[i,j] = Individual Concordance Index for Criterion k

COM = Cost of Manufacturing ($/yr)

CPI = Chemical Process Industries

CST95 = Critical Surface Time 95

CTAM = Critical Air Mass

CTWM = Critical Water Mass

CUi = Cooling Utility for Condenser of Column i (i = 1, 2, 3) or Cooler i (i

= feed, sol, prod) (°C)

D = Depreciation ($/yr)

DC = Direct Costs ($)

DDB = Double-declining Balance Method

Dk[i,j] = Discordance Index for Criterion k

DPBP = Discounted Payback Period (yr)

EDIP = Environmental Design of Industrial Products

EII = Environmental Impact Index

EP = Eutrophication Potential

EPA = Ecotoxicity Potential to Air

EPS = Environmental Priority Stages

EPW = Ecotoxicity Potential to Water

Fabs = Absorbent Flow Rate (kmol/hr)

FCI = Fixed Capital Investment ($)

Fi,j = Flow Rate of Stream i of Column j (where i = dist or btm; j = 1,2,3)

(kg/hr)

or btm; j = 1,2,3) (kg/hr)

FSi = Feed Stage for Column i (i = dist, prod, 1, 2, 3)

GD i = Green Degree of Chemical Compound i

GWP = Global Warming Potential

HCPW = Human Carcinogenic Toxicity Potential to Water

HEN = Heat Exchanger Network

Hj i = Impact Hazard Value for each chemical component i

HNCPW = Human Non-carcinogenic Toxicity Potential to Water

HTP = Human Toxicity Potential

HTPE = Human Toxicity due to Dermal Exposure

HTPI = Human Toxicity Potential by Ingestion

HUi = Heating Utility for Column i (i = 1, 2, 3)

IE = Industrial Ecology

IFj,k = Importance Factor for Impact Category j and Area k

IINH = Inhalation Toxicity Index

I j = Magnitude of deterioration on Impact Category j

IMCSD = Inter-Ministerial Committee on Sustainable Development

IMPACT = Impact Assessment of Chemical Toxics

IPC = Process Composite Index

IPCC = Intergovernmental Protection on Climate Change

Jk = Value of Objective Function k

LC50 = Lethal concentration that would cause death in 50% of Pimephales

promelas (mg/kg)

LCA = Life Cycle Analysis/Assessment

LD50 = Lethal-dose that produces death in 50% of rats by oral ingestion

(mg-min/m3)

M = Mass Flow Rate of a Stream (kg/hr)

MACRS = Modified Accelerated Cost Recovery System

MEN = Mass Exchanger Network

Mi = Mass Flow Rate of Component i (kg/hr)

MOO = Multi-Objective Optimization

NPV = Net Present Value ($)

NPW = Net Present Worth ($)

NSGA = Non-dominated Sorting Genetic Algorithm

ODP = Ozone Depletion Potential

PAT = Profit after Taxes ($/yr)

PBP = Payback Period (yr)

PBT = Profit before Taxes ($/yr)

PCOP = Photochemical Oxidation Potential

PEI = Potential Environmental Index

Pk = Preference Threshold for Objective k

PVR = Present Value Ratio

Qr = Energy Consumption (kW)

rd = Discount Rate (yr-1)

RHKi = Heavy Key Reovery for Column i (i = 1, 2, 3)

RLKi = Light Key Recovery for Column i (i = 1, 2, 3)

ROI = Return on Investment

rt = Tax Rate (yr-1)

SCENE = Simultaneous Comparison of Environmental and

Non-environmental Process Criteria SGA = Scaled Gradient Analysis

SL = Straight Line Method

SMD = Solid Mass Disposal

SOO = Single Objective Optimization

SOYD = Sum of the Years Digits Method

SPI = Sustainable Process Index

Stagei = Stages for Column i (i = abs, dist, prod, 1, 2, 3)

T abs = Absorbent Temperature (°C)

TAC = Total Annual Cost ($/yr)

TAPPS = Total Annualized Profit per Service Unit ($/yr-unit)

TCI = Total Capital Investment ($)

tD = Project Lifetime (yr)

TRACI = Tool for the Reduction and Assessment of Chemical and other

environmental Impacts TTP = Terrestrial Toxicity Potential

T VOC = VOC Temperature (oC)

TWA-TLV = Time weighted average of the threshold limit values

VBA = Visual Basic for Applications

V k = Veto Threshold for Objective k

VOC = Volatile Organic Compound

WC = Working Capital ($)

WCHi,j = Weighted Category Hazard for Impact Category j for Chemical i

wj = Weighting Factors for Impact Category j

Wk = Weights for Objective k

WMO = World Meteorological Organization

Yji = Normalized Impact Factor

Greek symbols:

σi = Final Ranking Score

σ[i,j] = Outranking Matrix

List of Tables

Page

Table 2.1: Different definitions for Return on Investment (ROI) 19

Table 3.2: GWPs and ODPs of some substances; refer to the respective

references in the foot-note for the extended list

44

Table 3.3: PCOPs of some organic substances (Heijungs et al., 1992) 47

Table 3.5: Eutrophication potential of compounds (Heijungs et al., 1992) 50

Table 4.1: Economic and Environmental Criteria – Petroleum Refinery and

Petrochemicals

60

Table 4.2: Economic and Environmental Criteria – Biotechnology,

Pharmaceutical and Chemicals

66

Table 4.3: Economic and Environmental Criteria – Downstream Processing 71

Table 4.4: Economic and Environmental Criteria – Energy Systems and Heat

Integration

76

Table 4.6: Environmental Criteria – Biotechnology, Pharmaceutical and

Chemicals

83

Table 5.1: Objectives, Decision Variables and Constraints for VOC Process

Table 5.3: Comparison of Two Selected Pareto-optimal Solutions 108

Table 5.4: Compositions of Components in Spent Wash Solution 110

Table 5.6: Comparison of Two Selected Pareto-optimal Solutions 118

Table 6.1: NFM Parameters for Ranking VOC Recovery Application 129

Table 6.2: NFM Parameters for Ranking Solvent Recovery Application 131

List of Figures

Page

Figure 5.3: Selected Results for Operation Optimization of VOC Recovery

for PBT and PEI

95

Figure 5.4: Selected Results for Operation Optimization of VOC Recovery

for Eight Objectives

97

Figure 5.5: Selected Results for Operation Optimization of VOC Recovery

for Five Objectives

Figure 5.8: Optimal Objective Values for Design Optimization of VOC

Recovery for Five Objectives

109

Figure 5.9: Sequences 1 and 2 from Chakraborty and Linninger (2002) 111

Figure 5.10: Selected Results for Design Optimization of Solvent Recovery

for NPW and PEI

115

Figure 5.11: Selected Results for Design Optimization of Solvent Recovery

for Ten Objectives

120

Figure 6.1: (a) Individual concordance index, and (b) discordance index

calculations used in NFM algorithm to determine ranking scores for the Pareto domain solutions

127

Figure 6.2: Ranking of Pareto-optimal Solutions by Net Flow Method for

VOC Recovery Design Optimization for Several Objectives

130

Figure A.1: Excel-VBA-HYSYS Setup for MOO of Processes 152

Figure A.2: Flowchart for NSGA-II implemented in VBA; (*) indicate steps

which require VBA and HYSY interface

158

Chapter 1: Introduction

Chapter 1 Introduction

1.1 Optimization of Chemical Processes

Optimization refers to finding one or more feasible solutions which correspond to the maximum and/or minimum of one or more objectives The need to find such optimal solutions in a problem comes mostly from the purpose of designing and operating a plant for minimum fixed capital cost and/or operating cost, for maximum reliability, and others As a result, optimization is one of the major quantitative tools in industrial decision making In general, it has become an integral part of our society, without which many activities would not be as efficient as they are now A plethora of problems in the design, construction, operation, and analysis of chemical plants (as well as many other industrial processes) can be resolved by optimization As society evolves with ever changing economic and environmental landscape, there is still room to further optimize the current industrial operations

There are generally two types of optimization problems, namely, the single objective optimization (SOO) and the multi-objective optimization (MOO) The first type (SOO) considers only one objective in the optimization procedure In employing this method of optimization, the decision maker would need to choose the objective that is of greatest relevance to the problem at hand Since practical applications require several objectives to be considered simultaneously, there is growing interest in the optimization

of more than one objective – commonly known as the MOO Bhaskar et al (2000) presented the background of MOO, different methods and their applications in the

chemical engineering are up to the year 2000 Their review shows that there were around

30 journal publications on MOO of various processes before 2000; on the other hand, about 80 applications of MOO in chemical engineering have been published since 2000, according to Masuduzzaman and Rangaiah (2008) These two reviews provide a comprehensive summary of chemical engineering applications, and interested readers may refer to them for more detailed information

There are many techniques that can be used to solve MOO problems These

techniques can be classified into five different classes: (1) no preference methods; (2) a posteriori methods using scalarization approach; (3) a posteriori methods using multi-

objective approach; (4) a priori methods; and (5) interactive methods (Miettinen, 1999, and Rangaiah, 2008) A summary of the methods used and the applications studied in the field of chemical engineering from 2000 to mid-2007 is provided by Masuduzzaman and Rangaiah (2008) The popular methods used by academia in the field of chemical

engineering are a posteriori methods using scalarization approach (i.e weighting and constraint method) as well as a posteriori methods using multi-objective approach (e.g

ε-genetic and evolutionary algorithms)

As the objectives may be partially or totally conflicting, the solution of a MOO problem will not be unique and there will be many optimal solutions, which are known as Pareto-optimal solutions Each one of them, when compared to another, is better in at least one objective value and is worse in at least one other objective value Thus, Pareto-optimal solutions elucidate the trade-offs present among the objectives, and equip the decision maker in choosing one of them for implementation based on other information, his/her experience and preferences

Chapter 1: Introduction

For any operation or design case of a chemical process, only a single optimal solution would be required for implementation Hence, with the plethora of solutions obtained for MOO problems, a choice for one solution has to be made In order to choose

a single solution, ranking methods such as rough set or net flow methods (Thibault, 2008) could be used These methods require information on the Pareto-optimal solutions and inputs from the decision maker before the solutions can be ranked The availability of Pareto-optimal solutions also provides a quantitative foundation in reducing the biasness

of the decision maker when his/her inputs are required for the ranking methods (Deb, 2001)

1.2 Objectives in Process Optimization

As Lord Kelvin, a physicist, once said: “When you can measure what you are speaking about and express it in numbers, you know something about it” The formulation of objective functions is one of the crucial steps in the application of optimization to a particular problem Hence, one must be able to translate a verbal statement or concept of the desired objective into mathematical terms The choice in the number and type of objectives is dependent on the purpose of the study There are many objectives available and they are briefly discussed below

Very often, as chemical industries are profit-driven, the objective functions are economics-related They can be material metrics or profitability measures Material metrics are ratios that measure the efficiency of the chemical process – e.g amount of product per unit of feed, amount of waste emitted per unit of product On the other hand, profitability measures are economic objectives commonly used by the industries to

measure the performance of the chemical processes They include revenues, manufacturing costs, profits, net present worth, payback period, etc Chemical processes can also be optimized for objectives that are not related economics These objectives can

be product quality, energy efficiency, environmental impacts, sustainability, process safety, operation time, robustness, etc

1.3 Motivation and Scope of Work

The World Commission on Environment and Development (1987) defined sustainability

as “the development that meets the needs of the present, without compromising the ability of future generations to meet their own needs” This is often quoted in almost every article advocating sustainability (e.g Azapagic and Perdan, 2000; Sikdar, 2003a and 2003b; Heinzle et al., 2006; Darton, 2006) Increasing attention and emphasis are being placed on sustainability For example, an Inter-Ministerial Committee on Sustainable Development (IMCSD) has been established in Singapore in February 2008 (available at http://app.mewr.gov.sg/web/Contents/ContentsSSS.aspx?ContId=1034) IMCSD seeks to create a national framework and strategy for Singapore’s sustainable development in view of the rising domestic and global challenges The journey towards sustainability is not possible if it only depended on environmentalists and/or the government; it requires the awareness of every individual and their efforts to realize this laudable objective

There are three spheres of sustainability: economic development, environmental stewardship and societal equity (e.g., Azapagic and Perdan, 2000; Sikdar, 2003a and 2003b; Heinzle et al., 2006) This is often touted as the “triple bottom line” Economic

Chapter 1: Introduction

indicators measure the profitability of a chemical process Usually, these are the first criteria for companies; if the process is not economically feasible, the project would be aborted Environmental indicators are concerned with efficient use of raw materials and energy in the process as well as the environmental impacts caused by emissions The latter include human toxicology, ecotoxicity, global warming, ozone depletion, acidification, eutrophication and photochemical oxidation (Young and Cabezas, 1999; Jolliet et al., 2003; Bare et al., 2006) Societal indicators measure different aspects of working conditions and regulations; these indicators are driven by the government and

company’s policies, but not by engineers per se Hence, a chemical engineer can and

should consider both economic and environmental performance of the process in order to make it sustainable As Steve Johnson, the administrator of US Environmental Protection Agency, once said: “We have the responsibility to sustain – if not enhance – our natural environment and our nation’s economy for future generations” (http://www.epa.gov/Sustainability/, accessed on 1st Aug 2008)

Identifying methods to quantify economic and environmental performance of chemical processes is essential Economic objectives (e.g payback period and net present worth) have always been the most important criteria to the industries, and are well established On the other hand, environmental indices are still in the development stage and there exist a number of methods to calculate them There are almost 80 journal papers since the year 1995 discussing the calculation of one or more environmental indices and then using them to quantify the performance of chemical processes

There are several differences among the environmental indices employed A number of investigators chose to consider selected individual environmental components

(e.g critical air mass, CTAM based on LC50 by Stefanis et al., 1996; critical water mass,

CTWM by Steffens et al., 1999; inhalation toxicity index, IINH using EFRAT

methodology by Chen et al., 2002; human toxicity potential by ingestion, HTPI, aquatic toxicity potential, ATP and acidification potential, AP by Kim and Smith, 2005; weighted sum of benzene and carbon dioxide production by Janjira et al., 2007) Other researchers chose to work with aggregated indicators which are more popular Examples of aggregated indices are the Potential Environmental Index (PEI) using the Waste Reduction (WAR) algorithm (Young et al., 2000; Chen and Feng, 2005), Eco-Indicator

99 (Hugo et al., 2004; Dominguez-Ramos et al., 2007), Atmospheric Hazard Index, AHI (Gunasekera and Edwards, 2003) and Green Degree (Zhang et al., 2008)

Besides using aggregated or single environmental indicator, more than one environmental indicator can be employed since there are many contributing components For example, Azapagic and Clift (1999) optimized the formation of boron products by minimizing global warming potential (GWP) and photochemical oxidation potential (PCOP) simultaneously via bi-objective optimization They also employed MOO for GWP, ozone depletion potential (ODP), production rate and costs using the ε-constraint method Steffens et al (1999) used annual costs as the economic objectives with two environmental indicators (CTWM and SPI) for penicillin production Kim and Smith (2005) optimized the recovery of acetic acid from aqueous waste mixtures using four objectives – profit, HTPI, ATP and AP Results in Azapagic and Clift, 1999) and other works reviewed in Chapter 4, clearly show that minimization of one environmental component (e.g GWP) does not necessarily minimize another environmental component (e.g PCOP)

Chapter 1: Introduction

In the studies described above, MOO has been employed for two conflicting objectives – economic and environmental type Economic objectives are well-established and the choice of a profitability measure may be easy; even then, one may like to consider more than one profitability measure On the other hand, the choice of environmental performance indicator requires more care Most of the reported works have employed aggregated indicators, providing a final environmental performance index There are, however, many contributing factors for the environmental performance index For example, the toxicity impact on humans (HTP), ecotoxicity (ETP), GWP, ODP, etc Several studies have illustrated that minimization of an environmental performance index does not guarantee minimization of each contributing factor In addition, there are two main issues about the use of aggregated indicators First of all, the method of normalizing impact categories, and whether it brings the impact factors on the same platform, is debatable Secondly, unless the decision maker has an in-depth understanding of the process and the impacts it imposes, the weights given by the decision maker are highly subjective Since it is desirable to minimize the contributing components of the environmental performance index, they should be optimized as individual objectives together with the economic performance index One of the several MOO algorithms can be employed for this purpose

In this study, feasibility and usefulness of considering several economic and environmental objectives are investigated For this, two case studies are chosen: a VOC (volatile organic component) recovery system (Chen et al., 2003) and a solvent recovery system (Chakraborty and Linninger, 2002) Seven groups were identified for environmental impacts – HTP, ETP, GWP, ODP, PCOP, AP, and eutrophication (EP)

The last 5 environmental impact components can be lumped together as impact on the atmosphere, where necessary In addition, economic aspects should be considered in the optimization Potential economic criteria are profit before taxes (PBT), payback period (PBP) and net present worth (NPW) Some or all these objectives will be optimized simultaneously using the elitist non-dominates sorting genetic algorithm, NSGA-II (Deb

et al., 2002) implemented in Excel

Net flow method can be used as a tool in identifying the preferred Pareto-optimal solution (Thibault, 2008); this requires the decision maker’s preferences While the choice of a solution may be subjective, the generation of Pareto-optimal solutions provides a quantitative foundation in reducing the biasness of the decision maker (Deb, 2001) Pareto-optimal solutions and the preferred Pareto-optimal solution for the two case studies will be presented and discussed Insights gained from considering several objectives instead of just two will be highlighted

1.4 Organization of Thesis

Chapter 2 gives an overview on the concept of sustainability and explains the three contributing factors – economic, environmental and societal Thereafter, an overview of the different environmental objectives available as well as different aggregation methods

is presented in Chapter 3 Following on, Chapter 4 provides a comprehensive literature review on the applications studied for both economic and environmental objectives In Chapter 5, two applications are chosen for multi-objective optimization for bi-, many and/or several objectives The two applications are VOC and solvent recovery process The results obtained are presented and the decision variables that have an influence on

Chapter 1: Introduction

the objective functions are discussed MOO would provide decision makers with a myriad of solution; however, for an operation or a design case, a single point would have

to be identified to determine the conditions at which the application is to be designed for

or operated at Hence, in Chapter 6, net flow method would be employed to rank the solutions after the decision maker has provided the necessary parameters The preferred Pareto-optimal solution would be the one chosen for design or operation, wherever applicable Finally, Chapter 7 summarizes the conclusions of this study and recommendations for future studies

Chapter 2 Sustainability

As Lord Kelvin, a physicist, once said: “When you can measure what you are speaking about and express it in numbers, you know something about it” Hence, to have

a better understanding of the sustainability concept, it is preferable to quantify it Based

on the definition of sustainability, numerous indicators have been formulated These include ecological footprint, maximum sustainable yield (MSY), net national product (NNP), emergy, exergy, environmental sustainability index (ESI) and index of sustainable economic welfare (ISEW) (Bartelmus, 2001; Mayer et al., 2004; Cabezas, 2007) Such indicators are applicable to the sustainability of a nation, region or globe; however, they are not useful for businesses or specifically for chemical processes Hence, another set/type of indicators has to be formulated

Chapter 2: Sustainability

There are three spheres of sustainability: economic development, environmental stewardship and societal equity (Azapagic and Perdan, 2000; Sikdar, 2003a and 2003b; Heinzle et al., 2006) This is often touted as the “triple bottom line” Economic affluence

is desirable especially for the poverty stricken; however, economic growth should proceed in a controlled manner that is able to balance the need for social development and equity Moreover, quality of the environment should not be compromised Our environment is important as it supplies our needs while absorbing the wastes generated Figure 1 illustrates how these three spheres overlap to achieve sustainable development (indicated by the red/shaded region) Operating in the “sustainable” zone does not imply that the world has to maintain a particular way of life that is deemed optimal; rather, this state would change with demands as well as any limitations involved over time (Darton, 2007)

Figure 2.1: Three spheres of sustainability

A couple of studies have suggested a fourth sphere, on top of the three shown in Figure 2.1 The first suggestion of the fourth sphere is resource efficiency (Afgan et al.,

2000 and 2007; Darton, 2006), which is defined as the proportion of the resources whose benefits have been realized by the end users For example, in the context of chemical engineering, it would measure the proportion of the raw materials that has been converted

to the desired product (e.g the conversion of ethyl benzene to styrene would have some undesired products formed simultaneously) Azapagic and Perdan (2000) had classified resource use (or efficiency) in the “environment” sphere of sustainability As resource efficiency indirectly indicates the amount of waste that would be produced, it could be used as a measure of the environmental impact a process has Hence, in accordance with Azapagic and Perdan (2000), resource efficiency would be treated as an environmental indicator Interestingly, Spangenberg (2002) suggested that the fourth sphere should be institutional This includes citizens’ turnout at election polls, access to basic necessities, etc Evidently, the institutional sphere is only applicable when analyzing a country or region

Companies interested in issuing sustainability reports could refer to the Global Reporting Initiative (GRI) guidelines which provide a comprehensive list of economic, environmental and social data to be recorded (Weiss and Funnell-Milnar, 2007) This list can be obtained from www.globalreporting.org (accessed in January 2008) Almost 1000 organizations have indicated their usage of the GRI Reporting Framework The companies from the chemical industry that have employed the GRI Reporting Framework are Du Pont, Dow Chemical, BP, ExxonMobil and Shell These reports can be used to benchmark organizational performance with respect to laws, norms, codes, performance standards and voluntary initiatives, to demonstrate organizational commitment to sustainable development, and to compare organizational performance over time On the

Chapter 2: Sustainability

other hand, specifically catered to chemical industries, IChemE presented a set of 50 sustainability metrics involving all three spheres (refer to www.icheme.org/sustainability, accessed in January 2008) Other industries can also employ these metrics with some modifications Both these sustainability reporting guidelines are very similar: from the amount of profits (economic) to the amount of greenhouse gases produced (environmental) to the rate of employee turnover (societal)

Very often, chemical engineers are faced with the design or operation of chemical processes Thus, this chapter seeks to present a review of the various sustainability indicators available in the literature that could be applied to chemical processes These indicators can be classified into the three spheres as shown in Figure 2.1

2.2 Sustainability Indicators

As mentioned above, sustainability incorporates economic, environmental and societal effects Thus, it would be more appropriate if indicators for each aspect are considered separately However, it should be noted that the indicators do not solely affect only one aspect; rather, they affect more than one aspect due to interactions in the society For example, material intensity would be classified as an environmental indicator since it refers to efficient use of resources and the use of resources has a negative impact

on the environment Conversely, processing the resource would add value to it and would have a positive impact on the economy, and the society would benefit from the product formed while future generations would be deprived of the resource (Martins et al., 2007) Hence, although primary effect of an indicator is on the sphere in which it is classified, it could have secondary effect on the remaining two spheres

2.2.1 Economic Indicators

In chemical engineering projects, profitability measures are always used in assessing the feasibility of a project Moreover, as companies are profit-driven, businesses would usually review the economic benefits of a project or process unit before investing in the project/process There is an established set of economic measures/indicators which is often presented in textbooks and used in the industry (e.g., see Turton et al., 2003; Edgar

et al., 2001) Before providing the profitability measures, some concepts of cost estimation have to be introduced first as they will be required for the measures

2.2.1.1 Cost Estimation

There are two sets of costs that the engineer has to consider when designing a new plant

or implementing a new fixture in the existing plant They are (1) capital cost or investment and (2) manufacturing cost

2.2.1.1.1 Capital Investment

Total capital investment (TCI) is the investment required to purchase new equipment, build and install them such that it is ready for production It is the sum of two components: (1) fixed capital investment (FCI) and (2) working capital (WC) FCI is the cost associated with building the plant A small portion of FCI is for the land and is not depreciable FCI can be calculated using capital cost (investment) programs such as CAPCOST© (Turton et al., 2003), Aspen Icarus Process Evaluator (Aspen Technology Inc., 2007) and Cost Analysis and Project Economic Evaluation by SuperPro and EnviroPro Designer (http://www.intelligen.com/costanalysis.shtml; Seider et al., 2004)

Chapter 2: Sustainability

WC is the amount of capital required to start up the plant and finance the first few months

of operation before revenues from the process start This usually includes raw material inventories, salaries, cash and accounts receivable which can be recovered at the end of the project lifetime WC is usually estimated as 15-20% of FCI (Seider et al., 2004; Turton et al., 2003)

Depreciation is the reduction in the value of an asset The total capital amount that can be depreciated is FCI minus the land cost (L) and the salvage value (S) Salvage value is a small proportion of the initial FCI and often it is assumed to be zero Land cost

is also a small fraction of the total and thus its contribution to FCI is negligible There are various methods to calculate the depreciation amount for each year; for example, straight line method (SL), double-declining balance method (DDB), sum of the years digits method (SOYD) and modified accelerated cost recovery system (MACRS) For illustration purposes, depreciation (D) is calculated using the SL method over a project

lifetime of tD years when the salvage value is recovered

D D

FCIS-L-FCI

Direct costs (DC) are costs that vary with the production rate (e.g cost of raw materials) Fixed costs (FC), on the other hand, are costs that do not vary with the production rate

(e.g depreciation) General expenses (GE) are related to the management and administration activities, and not directly to the production process

2.2.1.2 Profitability Criteria

Several profitability criteria are available and the choice of criteria to be used for evaluation of a project is dependent on the engineer/company In analyzing the profitability of a process that is already online, capital costs would have been pre-determined and already incurred, and so the following would most likely be used in the economic analysis

Profit before taxes (PBT) is the difference between the revenue earned from the production process and the total annual cost incurred Revenue (R) is the sales of the product while TAC is the sum of COM and D

Profit after taxes (PAT) is the amount of the profit that the businesses retains after

accounting for the income tax Let rt denote the tax rate for the region Hence PAT is

calculated via equation 2.4

PAT = PBT (1 - rt) = (R – COM – D) (1 - rt) (2.4)

As depreciation is an amount to account for the initial investment, it is not a physical outflow of cash for the business for that particular accounting year Hence, the term cash flow (CF) is used for the physical cash exchanges of the company Thus, CF is the addition of D to PAT

CF = PAT + D = (R – COM – D)(1 - rt) + D = (R – COM)(1 - rt) + rtD

(2.5)

Chapter 2: Sustainability

2.2.1.2.1 Time Criterion

Payback period (PBP) is the number of years for the annual income ($/yr) to recover the initial investment ($) without considering the time value of money There are a number of definitions for PBP One common definition of PBP has the CF as the yearly income and the FCI as the initial investment (e.g., Pintarič and Kravanja, 2006; Heinzle et al., 2006; Turton et al., 2003) Heinzle et al (2006) had an additional definition for PBP, which is the ratio of TCI to PAT

CF

FCIPBP = or

a factor called the discount rate (rd) which represents the minimum acceptable rate of

return that the company is willing to accept for any new investment The future yearly

incomes are brought to the present value using rd As one would notice from the

expression of DPBP in equation 2.7, it requires iterative methods to obtain its value as DPBP appears on both sides of the equation

DPBP d d

1

1ln

1CF

1FCIlnDPBP

of production (i.e bearing dissimilar fixed capital investments), it is preferable to employ the cumulative cash ratio (CCR) which is the ratio of all positive to negative cash flows

CCP = - TCI + CF
tD + WC + S + L ≈ - FCI + CF
tD (2.8)

TCI

WCCF

TCI

LSWCCF

≈+++

⋅

Similar to CCP is the net present value (NPV), also known as net present worth (NPW), which takes into account the time value of money Thus all the positive and negative cash flows are brought to its present value before taking the positive cash flows net of the negative cash flows In addition, the present value ratio (PVR) is the ratio of the present value of positive to negative cash flows (i.e taking into account the time value of money)

D D

d d

d d d

d d

d

1

WC1

11

CFTCI1

LSWC1

11

CFTCI

t t

t t

r r

r

r r

r r

r

+

++

−++

−

≈+

++++

−+

11

CFTCI

1

LSWC1

11

CFPVR

D D

D D

D D

d d

d d d

d d

−+

−+

=

t t

t t

t t

r r

r

r r

r r r

(2.11)

2.2.1.2.3 Interest Rate Criterion

Return on investment (ROI) is the non-discounted rate at which annual earnings is made from the initial investment Several definitions for this available in the literature are presented in Table 2.1 in conjunction with equation 2.12 As illustrated above, profit before and after taxes differ by a fixed tax rate which cannot be altered by the engineer;

Chapter 2: Sustainability

also, TCI is usually approximated as a percentage (115 – 120%) of FCI and is thus not variable with process conditions

InvestmentInitial

EarningsAnnual

Table 2.1: Different definitions for Return on Investment (ROI)

The discounted cash flow rate of return (DCFROR) or more commonly known as the internal rate of return (IRR) is the discount rate at which NPW is zero This means that the expression for NPW in equation 2.10 is set to zero and the resulting equation is

solved for the value of rd

2.2.2 Environmental Indicators

Azapagic and Perdan (2000) had compiled a list of environmental indicators based on the life cycle approach These indicators are classified into three categories: (1) environmental impacts, (2) environmental efficiency, and (3) voluntary actions As the aim of this chapter is to present sustainability indicators that are applicable to a chemical process in the context of its design and operation, only the first two categories will be discussed in the subsequent sections The last category on voluntary actions (e.g management systems or assessment of suppliers) is relevant mainly to larger entities such

as businesses or industries, and thus will not be discussed here

2.2.2.1 Environmental Impact

Environmental impact indicators comprise both local toxicology effects and global atmospheric effects The former includes toxic effects on human beings, terrestrial and aquatic organisms caused by the chemical compounds involved in the process On the other hand, the latter consists of atmosphere-related issues that involves the degradation

in the quality of air, water and soil that surrounds us Several environmental assessment tools have employed toxicity and/or atmospheric indicators Examples of these tools are the waste reduction algorithm, WAR (Mallick et al., 1996; Cabezas et al., 1999; Young et al., 1999 and 2000); the critical surface-time 95, CST95 (Song et al., 2002); the tool for the reduction and assessment of chemical and other environmental impacts, TRACI (Bare

et al., 2003 and 2006); and IMPACT 2002+ (Joillet et al., 2003) Only a brief overview of indicators for toxicology and global atmospheric effects will be presented here; more details are presented in the next chapter

2.2.2.1.1 Local Toxicology Effects

First of all, toxicity is the ability to inflict harm on to a living organism as well as to indicate the adverse effects caused by a chemical It can be measured as the amount of toxic equivalents that is being emitted by the process (Heijungs et al., 1992; Schwarz et al., 2002; IChemE1) Toxicology effects can be measured for three categories of organisms: (1) human beings, (2) terrestrial life forms, and (3) aquatic organisms Toxicity effects on both human beings and terrestrial life forms can be triggered through the ingestion of toxic chemicals The common factor used to measure toxicity via

1

The Sustainability Metrics: Sustainable Development Progress Metrics, recommended for use in the

Chapter 2: Sustainability

ingestion is the lethal-dose that produces death in 50 percent of rats by oral ingestion (LD50) (Young et al., 1999; Joillet et al., 2003) In addition, human beings are the ones who are directly exposed to the chemicals due to working in the chemical plant or as direct or indirect users of the products As such, human beings can come into contact with the toxic chemicals via inhalation and exposure This category of toxicity can thus

be measured using the time weighted average of the threshold limit values (TWA-TLV) (Young et al., 1999) Another measure of toxicity on humans generally, is based on the rate of usage and the MSDS’s R-phrases of the chemical compounds (Vincent et al., 2005; Martins et al., 2007) Finally, for aquatic organisms, the toxicity effects can be

measured by the lethal concentration that would cause death in 50 percent of Pimephales promelas, a representative fish species, (LC50) (Young et al., 1999)

2.2.2.1.2 Global Atmospheric Effects

Besides toxicity, the atmospheric effects of a chemical process are important This is because it is not socially responsible to mar the environment which will have undesirable effects on the current and future generations For example, the release of carbon dioxide seemed harmless initially, but emitting more than what the ecosystem could absorb had proven to be disastrous with impacts such as greenhouse effects Global atmospheric effects are listed below (Heijungs et al., 1992; Young et al., 1999; Azapagic and Perdan, 2000; Song et al., 2002; Joillet et al., 2003; Bare et al, 2003 and 2006) Each atmospheric effect would be further elaborated in the next chapter

• Global warming potential

• Ozone depletion potential

• Photochemical oxidation potential

be replenished in short periods of time; examples of nonrenewable resources are minerals, metals, fuel oil, natural gas and coal

Other than material and energy efficiencies, Azapagic and Perdan (2000) included three more indicators in this category of environmental efficiency: (1) material recyclability, (2) product durability, and (3) service intensity Material recyclability is the potential of the product material to be recovered for other applications This would reduce the amount of fresh raw materials required and the amount of waste disposed Product durability indicates that the product can be used over a relatively long period

Chapter 2: Sustainability

without being depleted or consumed With higher durability, it would lead to a reduction

in the volume throughput for the process, resulting in the decrease in consumption of raw materials Lastly, service intensity would be associated with the reuse, remanufacture or recycle of products Instead of selling their products to customers, companies would lease out their products As such, these products would be returned to the companies and not be disposed off after the customers have obtained utility satisfaction This allows the companies to lease the returned products to other customers (reuse), to restore and refurbish the return products (remanufacture), or to recover the materials for other purposes (recycle)

2.2.3 Societal Indicators

This set of indicators focuses on corporate social responsibility (CSR) by relating human well-being to the activities of business Various societal indicators have been provided by Afgan et al (2000 and 2007), Azapagic and Perdan (2000), Sikdar (2003), Krajnc and Glavič (2005) and Hienzle et al (2006) Some of these societal indicators are number of jobs/employees, accident frequency, income distribution and amount of capital invested

by the company in social and community projects As noted, this set of indicators pertains

to working environments and social rights which are largely dictated by the company and/or government policies, and not by the process design/operation Hence, the social indicators are not relevant in the design and/or operation optimization of chemical processes, which, however, should minimize the potential for accidents