Sustainability 2015, 7, 11438-11461; doi:10.

3390/su70911438
OPEN ACCESS

sustainability
ISSN 2071-1050
www.mdpi.com/journal/sustainability
Article

Application of Environmental Input-Output Analysis for

Corporate and Product Environmental Footprints—Learnings
from Three Cases
Louise Laumann Kjaer 1,*, Niels Karim Høst-Madsen 2, Jannick H. Schmidt 3 and
Tim C. McAloone 1

1
Department of Mechanical Engineering, Technical University of Denmark, 2800 Kgs. Lyngby,
Denmark; E-Mail: [email protected]
2
NIRAS, 3450 Allerød, Denmark; E-Mail: [email protected]
3
2.-0 LCA Consultants, 9000 Aalborg, Denmark; E-Mail: [email protected]

* Author to whom correspondence should be addressed; E-Mail: [email protected];

Tel.: +45-4525-4638.

Academic Editor: Marc A. Rosen

Received: 12 June 2015 / Accepted: 13 August 2015 / Published: 25 August 2015

Abstract: An increasing number of companies are expanding their environmental impact

reduction targets and strategies to include their supply chains or whole product life cycles.
In this paper, we demonstrate and evaluate an approach, where we used a hybrid Environmental
Input-Output (EIO) database as a basis for corporate and product environmental footprint
accounts, including the entire supply chain. We present three cases, where this approach was
applied. Case study 1 describes the creation of total corporate carbon footprint accounts for
three Danish regional healthcare organisations. In case study 2, the approach was used as
basis for an Environmental Profit and Loss account for the healthcare company, Novo
Nordisk A/S. Case study 3 used the approach for life cycle assessment of a tanker ship.
We conclude that EIO-based analyses offer a holistic view of environmental performance,
provide a foundation for decision-making within reasonable time and cost, and for
companies with a large upstream environmental footprint, the analysis supports advancing
their sustainability agenda to include supply chain impacts. However, there are implications
when going from screening to implementing the results, including how to measure and
monitor the effect of the different actions. Thus, future research should include more detailed
models to support decision-making.
Sustainability 2015, 7 11439

Keywords: environmental footprint; carbon footprint; life cycle assessment; hybrid

environmental input-output database; sustainable business development; green supply chain;
environmental profit and loss; natural capital accounting

1. Introduction

Companies with ambitious sustainability agendas are starting to expand their efforts from reducing
directly controlled environmental impacts within their own operation to also including supply chain
embedded impacts [1]. These companies recognize that the indirect impacts from consumption of raw
materials, machinery, equipment and other supporting products and services can have a significant share
of the total environmental impact related to the company’s activities. Looking at greenhouse gases as a
single indicator, carbon footprint studies show that, depending on the sector, embedded emissions
upstream in the value chain can account for up to and over 90% [2,3]. This is especially the case for
service-oriented companies, where environmental loads, such as air emissions, are not produced at the
actual site of activity but are indirect impacts [4]. The growing field of green supply chain management
reports that many companies experience a pressing need to know how other stakeholders in the value
chain influence the environmental performance of the products or services that the company delivers to its
customers [5,6]. This can be supported by another growing field: environmental (especially carbon)
footprinting (see, e.g., Lenzen [7] or Wiedmann [8]), putting quantitative measures into play—on both
corporate and product levels. An environmental footprint can be defined as “a measure of natural resource
usage and environmental impacts for which a company or product is responsible within the life-cycle
of its entire operations” (adjusted from Ewing et al. [9]). In carbon footprinting, only greenhouse gas
emissions are captured.
Life Cycle Assessments (LCA) that are based on Environmental Input-Output (EIO) analyses are
reported to be a promising approach to achieve environmental quantifications, including all supply chain
impacts [3,8,10–12]. An EIO analysis is a “top-down” approach, capable of capturing impacts from the
entire supply chain and is useable as a screening tool to inform estimation of the anticipated life cycle
emissions [2]. However, the approach is also criticised, mainly for being too aggregated, thus causing
the need to hybridize with “bottom-up” data collection, where the generic EIO model is evaluated as
being insufficient [2,13–15].
A number of studies on applied EIO analyses can be found in the literature [2,14,16,17]. In most
of these studies, focus has been on demonstrating EIO analysis as a tool to overcome the hurdles of
time-consuming data collection and incompleteness of traditional LCA methods.
In this paper, we take the research a step further. Based on applied EIO analyses, we evaluate how
the approach can be used by companies in their strategic sustainability work, including supply chain
management and procurement activities. We attempt to investigate how companies might apply
EIO analysis for corporate and product environmental footprints, aiming to answer the following
two research questions:
Sustainability 2015, 7 11440

RQ1: What are the Strengths and Weaknesses of the EIO Approach?
RQ2: How Does Application of the EIO Approach Influence the Companies’ Environmental Agenda?

To answer these questions, we have conducted a literature review on EIO analysis and used three case
studies to demonstrate and evaluate the approach, utilising a hybrid EIO database as the basis for the
calculations. The cases are very diverse in nature but are all characterised as complex systems that
depend on many inputs of products and services. The first two cases analysed environmental footprints
on corporate levels and the third case was a life cycle assessment on the product level (albeit as large
and complex a product as a ship). In total, five companies were involved in the research and they all
provided feedback on the analyses through meetings, workshops and interviews in order to conclude on
the applicability of the approach. We discuss the findings with support of relevant external literature.
The three cases are introduced in the following.

1.1. Case 1: Corporate Carbon Footprint Account of Three Danish Regional Healthcare Organisations

Denmark is administratively divided into five geographical Regions, which share the responsibility
for all treatment provided by the Danish Healthcare System [18], with the primary function being provision
of hospital services. In addition to their core business of healthcare provision, all five organisations are
focused on reducing greenhouse gas emissions from their activities, through climate strategies, action
plans and various initiatives. The total carbon footprint accounts presented in this article were
commissioned by three of the five Danish Regions, independently of each other. For full reports (in
Danish) see [19–22].
The carbon footprint accounts were commissioned with the purpose of expanding the scope of
greenhouse gas reporting from use of energy to also include embedded emissions in the delivering
products and services, in order to show an ambitious climate change agenda and prioritise carbon
reducing initiatives. The purpose was also to add knowledge to the ongoing process of green
procurement, including environmental demands in tenders and purchases. Collaboration with the three
Regions on creating and updating carbon footprint accounts has been ongoing for more than five years.

1.2. Case 2: EIO Analysis Used for Environmental Profit and Loss Accounting of Healthcare
Company Novo Nordisk A/S

An Environmental Profit and Loss Account (EP&L) is an effort to account, in financial terms, for the
resources upon which a company, and its entire value chain, rely, plus the environmental impacts
generated. EIO analyses can be used to quantify the environmental impacts, which are then monetised.
In 2012, sports apparel company PUMA was the first global corporation to publish an EP&L [23].
Inspired by this case, in 2013, The Danish Environmental Agency initiated an EP&L of the Danish
healthcare company Novo Nordisk A/S as a new case example [24,25]. Novo Nordisk is a global
healthcare company, which core business is diabetes care. The company is responsible for the production
of more than 50% of the world’s insulin. The company also has leading positions within haemophilia
care, growth hormone therapy and hormone replacement therapy. Environmental and social
responsibility is included in the company’s articles of association, and the business philosophy is one of
balancing financial, social and environmental considerations in what is called the Triple Bottom Line [26].
Sustainability 2015, 7 11441

For Novo Nordisk, conducting an EP&L was a natural step to enhance knowledge on supply chain
impacts and explore the opportunities to reduce the company’s environmental impact in the upstream
part of the value chain.

1.3. Case 3: Life Cycle Assessment of a Tanker Ship Operated by Shipping Company TORM A/S

TORM A/S operates within the area of clean petroleum product shipping. As in many other shipping
companies, TORM’s environmental concern has to do with adhering to regulations set by the
International Maritime Organisation (IMO). Furthermore, after experiencing a period of rising bunker
fuel prices TORM was incentivised to reduce their ships’ fuel consumption. The economic recession,
which had its onset in 2008–2009, forced many shipping operators to evaluate a series of operational
and technical optimisation solutions, in an effort to cover costs and create a sustainable shipping
environment [27]. In this case study 3, the EIO approach was tested as a method to combine Life Cycle
Costing (LCC) and LCA in a common framework, using financial data as the main basis of both. This
article presents the results of the assessment of one of TORM’s tanker ships. The analysis and elaborated
methodology can be found in Kjær et al. [28].

2. EIO Analysis

The first Input-Output (IO) models were developed for economic calculations, by U.S. economist
Leontief, in 1936 [29]. He constructed a linear model of the U.S. economy relating the production inputs
of goods and services in an economy to the production outputs of other sectors [2,10]. Today, almost all
countries regularly compile IO tables as part of their national accounts. An IO table accounts for all
product flows between sectors in the economy. The IO table can be extended with direct emission and
resource data for each sector, making it possible to calculate the life cycle impacts from cradle to gate
per monetary unit spent for each product output. Utilising the environmentally extended IO model to
gain environmental insights is termed an Environmental Input-Output (EIO) analysis. The EIO analysis
is consistent with how impacts are calculated in a traditional life cycle assessment. The only differences
are that the transactions between the activities are measured in monetary units instead of physical units,
such as kilograms or kWh, and that the data collection in EIO is top-down, whereas a traditional or
process-based LCA typically uses a bottom-up approach. In a process-based LCA, data are collected for
all the processes that have been identified as important to include within the chosen system boundary.
This approach has the disadvantage of truncation errors, since it will always be a subjective evaluation
of which processes to include. Analysts are often forced to simply ignore many facets of life cycle
impacts due to lack of data availability and time restrictions [30], and, especially, uses of services are
typically ignored [4]. Studies show that process-based LCAs can underestimate emissions by as much
as 50% [4,14,31]. Furthermore, doing a comprehensive process-based LCA of all the products and
services a company are requiring, in both direct operations (in the production chain) and indirect
operation (administration, R&D, marketing, etc.), would quickly become a prohibitive task. EIO
analyses offers a top-down approach without any cut-offs and with the possibility of easy inclusion of
uses that are only measured in monetary terms, such as services [4]. This makes the approach suitable
for hotspot analysis and boundary decisions [2,10,32].
Sustainability 2015, 7 11442

Despite its many qualities, there are also a number of drawbacks with the EIO approach, as described
in the literature. The main weakness mentioned is sector aggregation, meaning that sectors may be too
heterogeneous to correctly reflect a particular process or product [11,17,32–34]. Other potential
weaknesses, depending on the EIO model, include: data age [17,33], use of monetary units (price
fluctuations and inhomogeneity) [33], insufficient handling of waste treatment [14,33], a limited number
of environmental indicators [14], potential exclusion of capital investments [14], and no or limited
accounting for regional differences and international trade [11,35].
Taking a hybrid approach, where the detail level of a process-based LCA is combined with the
completeness of EIO models, is suggested by many as the way forward to mitigate uncertainty
in compiling life cycle inventories [2,13,14,32,36–38]. The term hybrid can refer to two things: a hybrid
database and a hybrid analysis. In a hybrid database, IO analysis and mass flow analysis are fully
integrated, creating a mixed unit database with biophysical units for all mass and energy products and
monetary units for all service products and capital goods. The advantage of a hybrid database is the
possibility of using physical data instead of monetary data when available, since physical data is often
less uncertain and also more stable over time than monetary data [39].
In a hybrid analysis, the “top down” EIO data is combined with “bottom up” process-based data in
either a tiered or embedded (also called integrated) approach [34,39]. In the tiered approach, EIO data is
used to supplement the process-based data by, e.g., adding service and capital good expenses to the
inventory. The embedded approach fully integrates the process-based data into the EIO model, resulting
in the same completeness as the original EIO table but at a higher level of detail.
The approach used in the case studies presented in this article was based on a hybrid unit
EIO database (described in Section 3) and includes the use of hybrid analysis for selected processes.
Both tiered and embedded approaches were applied, depending on wishes of the study commissioners.
For the tiered approach, it should be noted that this approach inevitable entailed truncation errors when
comparing to the fully EIO integrated parts of the analysis.
This article contributes to the ongoing research on EIO analyses by demonstrating the use of a hybrid
EIO approach for comprehensive corporate and product environmental footprints and discusses its
usability for companies in advancing their environmental agenda.

3. Applied Method

In applying the EIO approach, we utilised a specific database, modelling approach and data collection
strategy, as described in the following.

3.1. Hybrid Database FORWAST

The analyses done in the three case studies were based on an EIO database called FORWAST.
FORWAST was developed as part of the Research Project of the 6th European Union Framework
Programme [40]. Later improvements of the original FORWAST model are listed in Table 1.
Sustainability 2015, 7 11443

Table 1. Modifications of the original FORWAST model.

References for Documentation
Modification
of the Improvements
Danish agriculture and food industry was further detailed. Hermansen et al. (2010) [41]
More emissions were added: nitrate, phosphate. Hermansen et al. (2010) [41]
More emissions were added: particles. Kjær et al. (2011) [22]
Import to Denmark was divided into import
Kjær et al. (2011) [22]
from EU27 and from rest of the world (RoW).
Land occupation was added as a resource input in the database. Schmidt and Muñoz (2014) [42]
A model for indirect land use changes was embedded in the model. Schmidt and Muñoz (2014) [42]

FORWAST is a so-called hybrid database as it is based on economic data from the national accounts,
as well as process-specific data from life cycle inventories (used for sector disaggregation). The
transactions in the model are in different units: dry matter for physical products, energy units for
electricity/heat/steam, and monetary units for other flows in the economy such as services. The database
distinguishes between different waste fractions and their treatment: recycling, incineration and landfilling.
Waste streams are determined by detailed mass balances. The model is based on data from 2003.
The original FORWAST database included 132 products/sectors. In refining the model, relevant
sectors were disaggregated using statistical data to provide a better match to the specific uses in each of
the case studies. These sectors included “Chemicals” (to separate pharmaceuticals from other chemical
products), “Machinery and equipment”, “Health and social work”, “Land transport and transport via
pipelines”, and “Hotels and restaurants”.

3.2. Hybrid Modelling

For relevant processes we hybridised with process-based data. An example was the modelling of
ambulance transportation used by the Regions (case study 1). The process “Taxi operation”, which was
a disaggregation of the sector “Land transport and transport via pipelines”, was modified to represent
ambulance transportation by using the fuel efficiency of the ambulances in the actual fleet. Another
example from case study 2 (Novo Nordisk) was the modelling of glucose used for insulin production.
The EIO sector most appropriate for glucose was identified and assessed. Detailed glucose production
data were merged into the EIO model to make a new and more precise profile for glucose production. A
full explanation of this particular hybridisation can be found in Høst-Madsen et al. [24].

3.3. Data Collection and Analysis

Since the used database was a hybrid EIO table using both monetary and physical units, upstream
emissions were calculated either per monetary unit (e.g., Danish krone (DKK)) spent on products or
services, or from physical consumption such as energy use, fuel use, and materials use in physical units
(kWh or kg).
In order to ensure full completeness in scope and full economic balance, complete annual expenditure
(referred to in this research as “spend”) accounts for each entity (e.g., hospital, production site, ship) were
provided by the company as the starting point. For the ship, data for five years of operation were gathered.
Sustainability 2015, 7 11444

Each spend was assessed and matched against the best representing EIO sector. For very broad and/or
large spend categories an extra assessment was made, e.g., by gathering invoices from suppliers, making
it possible to disaggregate the spend category to better match the categories in the EIO model.
As the FORWAST database required input in the currency EURO2003, all spend were turned into
2003 prices using inflation indexes and transferred into EUROs by the currency rate in 2003.
The process of gathering and analysing data is presented in Figure 1. First, all spend were summed
up for the categories chosen for the final results. Then, for fuel end energy consumption the financial
input data were replaced with data on actual consumption in physical units for which hybrid processes
were created for each type of energy sources. The same was done for selected focus areas within, e.g.,
transportation and purchased materials and products, where the companies were able to provide data in
physical units (kg) to allow more precise calculations. For materials and products where data on uses in
physical units were not retrieved, dry weight commodity prices for the specific sector as provided by
FORWAST were applied. As an example, for a ship’s use of engine stores, which mostly covered hand
tools, we used the average price that the sector “Transport by ship” paid for “Fabricated metal products”.

Figure 1. Process of data analysis.

Calculations were performed in SimaPro—software developed to assist LCA and carbon footprinting.
SimaPro facilitated analyses with both the EIO tables and other large LCA databases, such as Ecoinvent,
allowing for hybridisation.
For case study 1 (Regions), the Greenhouse Gas Protocol (GHG protocol) was used as framework for
presenting the results [43]. The GHG protocol is limited in the environmental scope, as it only deals with
greenhouse gas emissions. However, it provided a framework, dividing emissions into scopes, which
Sustainability 2015, 7 11445

were also used by the Carbon Disclosure Project in their reporting framework [44]. Scope 1 covers
emissions from sources under the company’s direct control, scope 2 covers indirect emissions from
purchased energy (e.g., electricity) and scope 3 covers all other indirect emissions upstream and
downstream in the company’s value chain, including purchased products and services. The International
Panel of Climate Change (IPCC) global warming potentials in a 100-year time horizon (GWP100) were
applied to convert single emissions, such as CO2, CO, CH4 and N2O, into a single unit: greenhouse gases
measured in CO2 equivalents (CO2e).
For case study 2 (Novo Nordisk), the results were distributed between own operation and supply
chain impacts, which were further divided into tier 1, 2 and 3 suppliers. The following environmental
indicators were obtained: Greenhouse gas emissions: CO2e. Air pollution: SO2, NMVOC, NH3, particles
(<10 µm) and NOx. Water use: Water quantities (m3). The method for monetary valuation is outside the
scope of this study and can be found in the EP&L methodology report [24].
For case study 3 (a TORM tanker ship), CO2e, SO2, particles,and NOx were included in the inventory,
however the analysis was only used to show results for the carbon footprint of the ship.
For capital investments, different scopes were applied for the different cases. For the Regions,
emissions related to capital spend were presented in a separate account to be able to monitor the
operational development independently of capital investment. Capital spend were merely included to
highlight that capital investments also have a carbon footprint. For Novo Nordisk, capital spend were
not included. For the tanker ship, emissions from capital investments (ship building and five year dry
dock repairs) were distributed over the operational lifespan of the ship.

3.4. Follow-up and Evaluation

Different activities where initiated with the companies to get their feedback on the applicability of
the approach and contributions to their environmental agenda, especially in order to answer RQ2.
For the Regions, the carbon footprint accounts ran in parallel with consultancy support on climate
mitigation strategies, involving regular meetings and workshops with key stakeholders. During these
meetings, it was discussed how the carbon footprint accounts would feed into the strategic work and
how the effect of initiatives could be monitored in future updates of the accounts. These activities
provided valuable feedback on the methodology’s usefulness and limitations, as well as knowledge on
opportunities and barriers for influencing the supply chain.
For Novo Nordisk, the case study was followed up by a presentation to the management, including a
discussion on the results and implications of the study. Furthermore, the authors of this paper conducted
a semi-structured interview with the Programme Director, Corporate Sustainability, in order to get more
detailed feedback on the applicability of the analysis. Finally, we consulted an expert review of the
PUMA EP&L [45] as a means to validate the applicability of the approach. In TORM, results were
presented and discussed at meetings with a vice president in the technical division. Since the primary
environmental strategy of the company was related to reducing fuel consumption, the discussions
focussed on which initiatives could contribute to this, including technical retrofits and more efficient
operation. Thus, in contrast to the first two cases, the study did not reveal any reasons to change this
agenda, and, therefore, no further discussions on influencing the indirect impacts in the supply chain
were initiated.
Sustainability 2015, 7 11446

Since, in all cases, connecting the environmental footprints to green supply chain initiatives was still
preliminary, we have supplemented the discussion in this paper with selected core literature on green
supply chain management.

4. Case Study Results

This section presents the outcome of the three case studies showing examples of calculated results
and findings in relation to the two research questions.

4.1. Case Study 1: Total Corporate Carbon Footprint Account of Three Danish Regional
Healthcare Organisations

The carbon footprint accounts for the Regions were made as yearly accounts. Results were divided
into categories and provided for each organisational sector (primarily hospitals) under each Region’s
administration. Results showed that 15%–21% of greenhouse gas emissions were related to the use of
energy (scope 1 and 2 in the GHG Protocol), while the remaining 79%–85% were embedded upstream
the value chain in purchased products and services (scope 3 in the GHG Protocol). Figure 2 shows the
distribution between scope 1 + 2 and 3 for the year 2011 for all three Regions.

Figure 2. Corporate carbon footprint accounts for three Danish Regions 2011, division
between scopes according to the GHG protocol.

In addition to providing the absolute number in CO2 equivalents (CO2e), a number of key performance
indicators (KPIs) were established, in order to monitor the development over the years. KPIs included:
CO2e/employee, CO2e/square meter, CO2e/hospitalised patient days (also called “bed days”),
CO2e/ambulant visits and CO2e/DKK spend (DKK is the Danish currency). Each Region commissioned
the studies independently and, thus, had different KPIs and year ranges. As an example, Table 2 shows
Sustainability 2015, 7 11447

the development of three KPIs for The North Denmark Region (all organisational sectors included) for
the years 2007 to 2012.

Table 2. Key performance indicators (KPIs) for The North Denmark Region 2007–2012.
KPI 2007 2008 2009 2010 2011 2012
t CO2e/employee 17.4 18.2 17.1 16.5 14.7 14.6
Kg CO2e/m2 323.6 347.0 351.9 344.1 302.5 302.3
Kg CO2e/DKK 67.8 63.3 60.9 61.1 56.8 53.2

The KPI CO2e/DKK helped to show the carbon efficiency in relation to financial spend. Since the
total carbon footprint would increase or decrease with the annual budget, monitoring CO2e/DKK took
that variable out of the equation. In Figure 3 we show the carbon footprint in relation to financial spend
for the Capital Region of Denmark in 2011. The figure shows that “patient articles”, including medicine,
was the largest spend category, accounting for 38% of the total spend. However, “Patient articles” only
accounted for 19% of the carbon footprint. This has to do with the fact that medicine is generally relative
expensive, and a lot of the money spent on medicine goes to service-related activities, such as research
and development. An example of the opposite was the category “food” that only accounted for 2% of
the total spent, while it accounted for over 10% of the carbon footprint.

Figure 3. Corporate carbon footprint account for the Capital Region of Denmark 2011.
Carbon footprint and financial spend for each category.

The carbon footprint accounts firstly had the purpose of showing company responsibility and secondly
to serve as an eye-opener to the indirect emissions embedded in the supply chain (scope 3). In relation
to RQ1, the strength of the EIO approach was that it made a first screening of the total impact feasible, since
the analyses were based on annual financial data, which was readily available within the organisations.
Sustainability 2015, 7 11448

The main limitation was seen as the ability to move from screenings based on average data to more
detailed accounts based on data from the actual supply chain in order to initiate and monitor improvements.
For all three Regions the following focus areas were identified as hotspots and recommended to detail
and monitor: “energy”, “food”, “transportation” and selected products within “patient articles” and
“large equipment”. For food, it was recommended to investigate if the amount of meat (especially red
meat) could be reduced and to reduce the amount of food waste. Monitoring of this area would require
collecting data on actual purchased amounts in physical units (kg) divided into product categories for all
participating kitchens and canteens.
Transportation was divided into patient transport and personnel transport, with patient transport
constituting the largest part. Before being able to monitor the development within patient transport, it
would be necessary to implement a system where the transport provider would report information, such
as: distance travelled (km), occupancy rate (for translation into person*km), and fuel consumption (km/L),
since these would be the main variables for reducing the impact from the transport service. As a
supplement to established KPIs related to quality, safety and cost, The North Denmark Region chose to
measure these environmental related KPIs for planned patient transport and implement them in future
updates of the carbon footprint account.
Based on the analysis, it was also recommended to investigate reduction potentials within purchased
equipment (electrical and non-electrical) and patient articles (medical articles, textiles, hygiene products,
etc.). A first step would be to acquire more detailed information from the suppliers in order to ensure
transparency. Relevant data could be: emission data, amount of materials used and production method
including geography, power use and power sources. Based on this, the Regions would have the necessary
information to initiate a dialogue with suppliers on how they could improve the environmental
performance of the products. Thus, in relation to RQ2, the carbon footprint accounts contributed to an
increased focus on purchase and consumption as important focus areas in the Region’s climate strategies
and action plans. However, for all three Regions, we did not experience that reducing upstream emissions
was a highly prioritised political goal from the beginning, and, as in any political organisation, changes
take time. Nevertheless, within each of the three Regions that commissioned the analyses, passionate
employees were driving the agenda forward. One concrete result of the carbon footprint accounts was
the North Denmark Region, which was the first public company disclosing all upstream scope 3
emissions to the Carbon Disclosure Project (CDP).
During the work on climate mitigation strategies for the companies, it became clear that focus areas
were not only decided based on their relative size in CO2e. Other selection parameters such as supplier
relationship and supplier readiness were also important. As an example, it would not be feasible to ask
a transportation provider to use electric vehicles for patient transport, since no transport company would
be able to provide such a fleet. Another selection parameter was the ability to monitor the improvement.
If it was not possible to gather sufficient data to document improvements, the project would not have the
required political backup. Other influencing parameters were current political focus (what was “hot
topics” on the political agenda) and previous initiatives (where they already had commitments). For one
Region, Region Zealand, five decisive parameters were formulated to prioritise climate change actions:
(1) CO2—effect (reduction potential)
(2) Implementation costs
Sustainability 2015, 7 11449

(3) Organisational process

(4) Branding effect
(5) Data availability (to monitor improvements on a detailed level)
Based on these parameters, 11 initiatives were selected for further exploration. The following areas
were aiming at reducing scope 3 emissions: reduction of food waste, food waste as biogas, travel, and
green procurement. Green procurement had two main focuses; (1) focus on requirements towards energy
consuming products and (2) focus on requirements towards non-energy consuming products. Again,
getting from a carbon footprint screening to a robust monitoring system, where improvements can be
traced and documented, was seen as a challenge needed to be overcome for the EIO approach to be fully
implemented with political goals and actions.

4.2. Case Study 2: Novo Nordisk Environmental Profit and Loss Account

Figure 4 illustrates the total monetised environmental impacts of Novo Nordisk in 2011 for the three
environmental KPIs: water use, greenhouse gasses and air pollution. The total monetised environmental
impact amounted to 223 million Euros. The main monetised environmental impact (total 77%) related
to greenhouse gasses, which supported the company’s own assessment of environmental impacts.
As shown on Figure 4, impacts were divided according to two spend types: Direct spend and Indirect
spend. Direct spend related to the production of Novo Nordisk pharmaceutical products and devices, and
included operational energy usage, water usage and the raw materials purchased from the supply chain.
Indirect spend covered all products and services, which were not part of the final consumer product. This
included, i.e., IT equipment, furniture and machinery.

Figure 4. Total monetised environmental impact distributed between the three environmental
KPIs: water use, greenhouse gasses and air pollution.
Sustainability 2015, 7 11450

Results were furthermore distributed between own operation and tier 1, 2 and 3 suppliers to illustrate
where in the value chain the main impacts occurred. Tier 1 covered finished products and services,
tier 2 processed materials and tier 3 raw materials. As shown in Figure 5, the main environmental impacts
occurred in tier 1 and tier 3 (total 75%). Novo Nordisk’s own operations constituted only 13% of the
total environmental impact.

Figure 5. Monetised environmental impacts distributed between tiers, shown for Direct and
Indirect spend.

The environmental impacts relating to Indirect spend amounted to 70% of the total. Although the
majority of impacts lay within Indirect spend, Direct spend was where Novo Nordisk had a greater sphere
of influence, since the majority of their own operation was represented here, as shown in Figure 5. In other
words, the majority of Novo Nordisk’s environmental impacts lay outside its direct control and would
therefore be more challenging to reduce.
In relation to RQ1, using the EIO approach as the basis for the EP&L was discussed with Novo
Nordisk, and the uncertainty of the results was taken up as a discussion point. Novo Nordisk has worked
for many years with conducting LCAs on their products. In these analyses, accuracy in the results was
considered very important, as it was used for reporting to customers. However, in the case of EP&L, the
EIO approach was considered appropriate. The uncertainty involved in both environmental impact
calculations and in the subsequent monetary valuation were not considered crucial, since the analysis was
primarily used as a communication tool to put new subjects on the agenda—the actual numbers were not
to be used for reporting or monitoring. In this way, the EP&L assisted in validating current reduction
strategies and identify whether there were areas which could be included in future strategies. However,
for the methodology to be fully implemented in the strategic work in Novo Nordisk, it would have to
develop from a screening tool into a strategic tool that could support decision-making. This process
Sustainability 2015, 7 11451

would involve moving from accounting to creating budgets and enable scenario development. Having a
budget-EP&L would enable comparison of investments before they were made. It would help quantify
different strategic paths, and help the decision makers in an organisation make more sustainable decisions.
Similar to the carbon footprint accounts for the three Regions described above, the primary goal of
the first EP&L was to show that Novo Nordisk took their environmental responsibility seriously and to
show honesty and transparency regarding the environmental impact of their business. However, in
relation to RQ2, the process of conducting the EP&L, inspired Novo Nordisk to explore opportunities
to reduce the environmental impact of the activities that were not directly related to the production of
their products. In terms of influencing the environmental agenda within the company, the Programme
Director reported a mind-set change in the company, stating that: “the understanding has changed away
from the stance that ‘environment is just something, which happens in production’”.
Being a company with high focus on sustainability, Novo Nordisk’s suppliers were already required
to document that their work was carried out in accordance with the laws and standards for environment,
health and safety, labour and human rights, and business ethics. The EP&L provided additional
knowledge to enhance this work with green supply chain management. Furthermore, the EP&L was an
eye-opener on how much of the impact was related to Indirect spend. While initiatives within core
activities related to Direct spend were already well integrated in the organisation, the field of Indirect
spend would require new initiatives aimed at all suppliers, not only those directly linked to the core
business, such as production of insulin and devices.
The EP&L provided more than just measuring impacts; it also served as a risk mitigation tool. As the
Programme Director put it in the interview: “We need to move away from only thinking about our own
production; environmental regulations keep getting shrewder, which makes the whole value chain
important”. Putting a monetary value on company externalities made the company aware of the risk of
the external costs becoming internalised in the future. If, e.g., a global carbon tax were to be introduced
(even though this was not seen as likely, but other means might trigger similar consequences), it would
result in higher prices on products with a high carbon footprint. It would also influence Novo Nordisk’s
own production facilities and this knowledge could help encourage investments in green technologies
and energy optimisation. Identified risks were moreover related to other environmental impacts, which,
in the future, might influence investment decisions, e.g., where to build a new production site. Thus, even
though the EP&L “only” served to measure impacts for one year, it contributed to inspire and suggest ways
of working with sustainability in a business context in the future.

4.3. Case Study 3: Life Cycle Assessment of a Tanker Ship

The life cycle assessment (LCA) of a tanker ship was based on gathered data for five years of ship
operation, extrapolated to represent the whole life cycle, which was assumed to be 20 years. Impacts
from ship building and dry dock repairs were distributed over the operational lifespan of the ship. Further
assumptions can be found in Kjær et al. [28]. Results showed that the dominant contributor to the carbon
footprint of a ship was the fuel combustion during the use stage. In CO2e, fuel combustion accounted for
80% of the impact. The second greatest contributor was extraction and production of the fuel, which
accounted for 9% of the impact. Thus, only 11% of the carbon footprint was not related to fuel, but
stemmed from the production and maintenance of the ship including the machinery and equipment on-board
Sustainability 2015, 7 11452

the ship, use of crew consumables including food, crew travel and use of services such as inspections,
port services, yard service, etc. The overall result of the Life Cycle Cost (LCC) and LCA, divided into
categories, are shown in Figure 6.

Figure 6. TORM tanker ship. Categories’ relative share of CO2e emissions and financial spend

The results showed that spend categories with high service content such as operational expenses had
low CO2e per USD. Port and transit services were somewhat higher due to the use of towage ships when
in port. Scrapping resulted in net avoided or “negative emissions” to the environment, because the scrap
was assumed to substitute virgin ore in the iron production. All costs were mapped as value added
throughout the life cycle to avoid double counting [28].
In relation to RQ1, the EIO approach had the advantage that the LCC and LCA could be based on the
same initial financial data. In contrast to the two other cases presented in this article, the indirect
emissions constituted only a small part of the overall environmental footprint and, thus, the embedded
uncertainty in the EIO model used to quantify these indirect impacts could be tolerated. Being closely
related to cost savings, fuel efficiency was already high on the agenda in TORM. In relation to RQ2, it
was not considered part of the company strategy to try and influence the indirect impact caused by suppliers,
and in this sense the environmental agenda of the company did not change as a result of the analysis.
However, presenting the costs and environmental burden together showed where there were potential
misalignments between the two and where there were risks of external costs being internalised in the
Sustainability 2015, 7 11453

future. In the case study, fuel accounted for 89% of the CO2e but only 36% of the cost. With the
likelihood of climate change continuing being an increased societal issue in the future, there could be a
risk of fuel prices rising as a consequence of carbon taxes or other market-based measures. Thus, the
analysis was used to discuss how the company could gain from increasing the focus on fuel savings, e.g.,
by investing in optimal maintenance, better quality ships, and on longer-term alternative energy sources.
Strategies for reducing fuel consumption included changes in operational patterns (involving decreased
speed), hull and performance monitoring, and technical retrofits. Since TORM was merely a ship
operator and not ship builder, how the ships were designed or might be retrofitted for being energy-efficient
would require close collaboration with suppliers. It was, however, not considered within the current
scope to include supply chain related (indirect) impacts in reduction strategies. Potential collaboration
would focus on how suppliers could help TORM reduce the fuel consumption of the ships.

5. Discussion

In this paper we discuss strengths and weaknesses of the EIO approach (RQ1) and how the approach
might influence a company’s environmental agenda (RQ2), including green supply chain initiatives.
Two issues of concern were identified during the research, and are discussed in the following:
uncertainty of the results (found as the main weakness in relation to RQ1) and integration with strategic
actions (found as the main challenge in relation to RQ2).

5.1. Uncertainty

Uncertainty of the results depended on two aspects: the EIO model used and the data inputs available
for the study. Model related uncertainties were: data age, geographic reach and sector aggregation. The
applied model was based on data from 2003, and it was therefore assumed that the technology mix was
representative of present consumption. The consumer price index was used to translate all spend to 2003
prices. However, the same general index was used for all types of products and services, which added to
the uncertainty. In regard to geography, a modified version of the EU27 table was used to represent the
Rest of the World (RoW). This especially could have influenced the results in case study 3, TORM, who
have most of their purchases on the global market. Another model related weakness was the limited
number of included elementary flows in the model. As an example, in case 2 (Novo Nordisk), another
model was applied for calculating water use, because this was not included in the FORWAST model.
In regard to sector aggregation, the uncertainty depended on the match between the financial spend
category and the EIO sector. When an EIO sector was chosen, it was assumed that the purchases could
be covered by the average products in that EIO sector. However, if the spend category was very broad
(e.g., “acquisitions”, which could cover many product types), the uncertainty of selecting the right
EIO sector was high. The same applied to the opposite case where a spend category was very specific
(e.g., “coffee machines”) but the EIO sector very broad (e.g., “Machinery and equipment”, which
contained coffee machines but also many other products). Thus, the best match would be when both the
spend category and the EIO category were specific (e.g., the match of the spend category “medicine”
and EIO sector “pharmaceuticals”). The next best match occurred when both the spend category and the
EIO sector were equally broad, hence, it could be assumed that the spend category covered the same
range of product types as covered by the EIO sector. Figure 7 illustrates the four different situations.
Sustainability 2015, 7 11454

Figure 7. Matching spend categories and EIO processes.

In the case studies, the uncertainty was reduced by disaggregating both EIO sectors (through
statistical data and hybridisation with more detailed data) and spend categories (e.g., by gathering data
on supplier invoices). However, this process took time and was, therefore, only done on selected focus
areas with high importance and/or high volume. The fact that the process of hybridisation can be time
consuming and require expert knowledge is also confirmed by Ewing et al. [9]. However, one important
prerequisite for the companies to implement changes was the ability to calculate the effect of potential
initiatives and monitor the effect after implementation. Thus, the need for more detailed data on selected
focus areas quickly arose. Data collection and analysis then became an iterative process with the aim of
reducing uncertainty and improve the analysis framework to better represent the actual circumstances
regarding the system being analysed.

5.2. Integrating Environmental Footprint Analysis with Strategic Actions

In all three cases, the environmental footprint screenings were conducted based on historic data, and
the analyses were by nature descriptive studies with the aim of showing the relationship between direct
and indirect impacts and identify hotspots, taking all upstream impacts into account. For case studies 1
and 2, the purpose of commissioning the study was furthermore to show external stakeholders that the
companies were responsible and had high standards for environmental concern. This is supported in an
expert review of the first EP&L conducted for PUMA, where the initiative was seen as an innovative,
refreshing and honest corporate initiative [45]. As mentioned, one case company used the results in their
compliance work, by reporting to CDP.
However, the question is if and how the analysis influenced or changed the companies’ environmental
agenda and broadened their scope. Case studies 1 and 2 showed that for these organisations and
companies the amount of supply chain hidden impacts was very high compared to what was currently
targeted in their reduction efforts. However, for case study 3, the case was quite the contrary and the
company merely used the study to support their current strategy on reducing direct emissions. Thus, the
current environmental agenda was only challenged in case studies 1 and 2. For these organisations and
Sustainability 2015, 7 11455

companies, getting numbers on the “hidden” impacts inspired them to start implementing actions aimed
at the indirect impacts caused by the supply chain. However, we also experienced that it was a
communicational challenge to explain the results in a positive way in order to inspire actions. Arguments
against included: perception that these impacts were outside the company’s sphere of influence, changes
would be difficult to implement, and that it would be difficult to monitor the effect of the actions. For
case study 1, employees responsible for climate change mitigation had previously focused primarily on
energy consumption and transport. Addressing impacts related to purchased goods and services required
new and different initiatives. It involved new stakeholders within the organisations, including
management, procurement and financial accounting. This indicates that for companies to be able to use
environmental footprints proactively, a coordinated effort between departments and other influencing
initiatives such as supply chain management and CSR strategies are needed. This is supported in the
expert review of the EP&L account for PUMA, where it is stated that it is important to align the analysis
with other efforts and work together as part of a bigger strategy initiative [45].

5.2.1. Impact Reduction Strategies

Using the EIO approach to calculate the impact based on financial data provided the opportunity to
relate the calculated impacts to the financial spend. This relationship would however only be relevant in
cases with options to switch from high carbon intensive to low carbon intensive uses. As an example,
switching from spending money on food products to, e.g., medicine in hospitals would hardly make any
sense. Focus was, therefore, on how consumption within each category could be optimised. Showing the
relationship between carbon footprint and financial spend was however interesting, as it highlighted the
fact that more service-intensive uses had lower impacts per money spend than product-intensive uses.
The case studies presented in this article, as well as other studies, confirm the fact that services are less
carbon intensive per monetary unit spend than physical products [46].
An EIO-based screening will highlight environmental hotspots totally independent of other
influencing factors, such as sphere of influence and political/management attention. However, as we saw
in the case studies, the “hard numbers” were only one of many considerations to take into account when
prioritising initiatives and efforts within the organisation’s sustainability agenda. For example, in case
study 1, even though hospital food was chosen as a focus area, it was not considered feasible to cut down
on the amount of meat for nutrition reasons. Instead, reducing food waste and using food waste for biogas
production were chosen as feasible strategies. When the different feasible actions were identified, the
reduction potential could be calculated and key performance indicators formulated.
Looking at actions aimed at reducing supply chain impact, the companies in case studies 1 and 2
where initially starting to integrate performance measures in their green purchase and supply
management. In the following we look at the literature on green supply chain management and discuss
how the environmental footprint account might be integrated.

5.2.2. Integration with Green Supply Chain Management

According to Seuring and Müller [5], governments/authorities, customers, and other stakeholders,
including NGOs, place a range of “triggers” on the focal companies (buying firms) to increase their
involvement in green/sustainable supply chain management. The authors divide mitigation strategies
Sustainability 2015, 7 11456

into “supplier management for risk and performance”, where standards and management systems, such
as ISO 14001, as well as supplier evaluation schemes, play central roles, and “supply chain management
for sustainable products,” where life cycle management is an essential activity [5]. Both strategies might
be supported by an environmental footprint analysis, putting quantitative measures into play. Supply
chain management in general might involve all suppliers to the company, and thus relates to the
accounting approach of case studies 1 and 2. “Supply chain management for sustainable products,”
evolves around the life cycle of the product or service produced by the company, and relates to the life
cycle assessment approach demonstrated in case study 3.
Risk avoidance is often a main driver for supply chain programs [47], and this was also seen in the
case of Novo Nordisk. Using the EIO approach for the purpose of an EP&L provided an understanding
of where hotspots were located within the supply chain and was, thus, seen as a risk identification tool.
As also stated in the expert review of PUMA’s EP&L, the account “provides an early view of emerging
risks, enabling businesses to respond strategically to protect and enhance shareholder value”.
Kogg and Mont [6] distinguish between three approaches to how companies address sustainability
aspects in the supply chain: de-selecting products or suppliers, who does not live up to a certain standard;
indirect influence through compensation schemes, NGO support, etc.; and direct approaches, such as
setting up specific requirements, motivate, support and enable sustainable development by inter-organisational
collaboration. De-selecting products or suppliers, which do not live up to a certain standard, does not
necessarily reduce the environmental impact, since the suppliers might simply deliver to another
customer instead. Likewise, the actual effect of using compensation schemes and supporting NGOs can
be very difficult to measure. It can thus be argued that a company will only make true measurable
changes through directly influencing the suppliers. In case study 2 (Novo Nordisk), supplier demands
had so far focused on compliance and reporting of procedures. This is supported by Kogg and Mont [6]
who state that companies are more likely to report in terms of procedures and steps they have undertaken
for dealing with suppliers, rather than reporting on the actual outcomes in terms of reductions of
environmental and social impacts. Thus, measuring and monitoring actual reductions of environmental
impact is still not common practise. One famous case of supply chain engagement is Walmart, who
experienced how simply asking suppliers to measure, report and state a reduction target on their
corporate greenhouse gas emissions led to improvements [48].

6. Conclusions

Companies are faced with a pressing need to address the environmental impact of their operation. For
many companies, a large share of the environmental impact is embedded in the upstream supply chain,
requesting the company to act as a responsible customer and implement green supply chain management
initiatives. In this article we have applied the EIO approach for corporate and product environmental
footprint accounts, including the entire supply chain. The first research question was:
RQ1: What Are the Strengths and Weaknesses of the EIO Approach?
We conclude that the greatest strength of the EIO approach was that it made analyses of the total
supply chain impacts feasible and enabled identification of hotspots in both direct and indirect
operations. The possibility of using readily available financial data as starting point made a first
Sustainability 2015, 7 11457

screening fast and relatively easy. For corporate footprint accounting, yearly financial accounts together
with supplier invoices provided the data input. For LCAs of products, Life Cycle Cost (LCC) inventories
can do the same, as also demonstrated by Junnila [12]. Not having to deal with truncation issues and
risks of leaving out potentially important contributors, the approach provided an opportunity to prioritise
all environmental release hotspots equally—independently of where in the value chain they arose.
However, uncertainty related to the modelling approach was identified as an issue in both literature and
the case studies. When addressing important focus areas, EIO models need to be optimised by
disaggregating relevant sectors and hybridising with process-based data in order to gain more
representative results. Future development of EIO models should have the following focuses:
• They should have a global scope, thus being multiregional to account for geographical differences.
• They should include the elementary flows needed to support what is of environmental concern.
• They should contain a wide range of both aggregated and disaggregated product groups/industry
sectors in order to enable the best match with available data for the system being analysed.
In the newly announced Exiobase v2, Input-Output tables are integrated with life cycle assessment
and mass flow analysis, and the database includes trade balanced supply-use tables for 43 countries and
five Rest of the World regions [49]. This provides a strong platform for creating analyses on various
levels ranging from environmental footprint of nations to corporate footprints and environmental
assessment of products [50]. Increasing the level of detail of product groups and sectors will strengthen
the model even further. Hybridising the model might come from integration with more detailed LCA
databases, such as EcoInvent, but might also be facilitated through projects where industries make data
available in order to gain a stronger foundation for the analyses aimed at their specific activities.
Moving from evaluating the approach as a method towards application of the results, the second
research question was:
RQ2: How Does Application of the EIO Approach Influence the Companies’ Environmental Agenda?
The case studies showed that the approach provided a good foundation for identifying where to
prioritise when looking for actions that could improve the environmental profile of the system under
analysis. For the companies and organisations with large upstream environmental footprints, the analyses
supported advancing their sustainability agenda to include supply chain impacts. When combined with
monetary valuation of impacts as in an EP&L, the analysis was also reported to serve as a risk mitigation
tool. The approach seemed less relevant for high-energy consuming products, such as ships, where focus
necessarily must be on reducing the impact within direct operation. Nevertheless, also in this context,
the approach provided an overview and put all contributors into perspective.
However, the magnitude of environmental impact was not the only criteria when the companies
selected focus areas for their sustainability work. Other influencing factors included: sphere of influence,
political focus, supplier relationship, and ability to monitor. More work is needed on integrating
environmental footprints with strategies on green supply chain management and green purchasing, both
in regards to organisational implications and in regards to calculating the actual environmental effect of
different actions. In this regard, there is a need for exploring how (hybrid) EIO models can be used
proactively to support decision-making in the development of more sustainable products and services,
and may be used as a tool to support investment decisions.
Sustainability 2015, 7 11458

Lastly, outside the academic community, development of guidelines for practitioners are needed.
Huang et al. [2] states that one of the reasons why EIO methods are lacking behind is the lack of expertise
available to assist companies to undertake EIO studies.

Acknowledgments

We would like to thank the participating companies and organisations: The Capital Region of
Denmark; The North Denmark Region; Region Zealand; Novo Nordisk; TORM. We also thank Trucost
and the Danish Environmental Agency for the collaboration on the Novo Nordisk EP&L.
This article was made possible thanks to research co-funded by DTU and the TORM Foundation. We
would also like to thank the three anonymous reviewers for their insightful comments.

Author Contributions

Louise Laumann Kjaer and Niels Karim Høst-Madsen worked as consultants in NIRAS for several
years and have developed the approach described in this paper together with Jannick H. Schmidt from
2.-0 LCA consultants. The work on corporate carbon footprint accounting for the three Regions was led
by Louise Laumann Kjaer and Niels Karim Høst-Madsen. Niels Karim Høst-Madsen led the project on
the Novo Nordisk EP&L. Louise Laumann Kjaer conducted the study on the TORM tanker ship.
Jannick H. Schmidt played a central role in developing and refining the FORWAST model and also
served as a reviewer of the conducted analyses. In writing the article, Louise Laumann Kjaer led the
work and conducted the literature review needed to write the paper. Tim C. McAloone supervised the
process and contributed to the writing and proofing of the paper. Jannick H. Schmidt contributed to the
description of the FORWAST model and the validity of the conclusions made. Niels Karim Høst-Madsen
contributed to the description of the methodology and the case study results. All authors have read and
approved the final manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

1. The Carbon Trust. Carbon Footprints in the Supply Chain: The Next Step for Business.
Available online: http://www.carbontrust.com/resources/reports/footprinting/carbon-footprints-in-
the-supply-chain-the-next-step-for-business (accessed on 28 April 2015).
2. Huang, Y.A.; Lenzen, M.; Weber, C.L.; Murray, J.; Matthews, H.S. The role of input-output
analysis for the screening of corporate carbon footprints. Econ. Syst. Res. 2009, 21, 217–242.
3. Larsen, H.N.; Solli, C.; Pettersena, J. Supply chain management—How can we reduce our
energy/climate footprint? Energy Proc. 2012, 20, 354–363.
4. Junnila, S.I. Empirical comparison of process and economic input-output life cycle assessment in
service industries. Environ. Sci. Technol. 2006, 40, 7070–7076.
5. Seuring, S.; Müller, M. From a literature review to a conceptual framework for sustainable supply
chain management. J. Clean. Prod. 2008, 16, 1699–1710.
Sustainability 2015, 7 11459

6. Kogg, B.; Mont, O. Environmental and social responsibility in supply chains: The practise of choice
and inter-organisational management. Ecol. Econ. 2012, 83, 154–163.
7. Lenzen, M. An Outlook into a Possible Future of Footprint Research. J. Ind. Ecol. 2014, 18, 4–6.
8. Wiedmann, T. Editorial: Carbon footprint and input–output analysis—An introduction. Econ. Syst.
Res. 2009, 21, 175–186.
9. Ewing, A.; Thabrew, L.; Perrone, D.; Abkowitz, M.; Hornberger, G. Insights on the use of hybrid
life cycle assessment for environmental footprinting—A Case Study of an Inland Marine Freight
Transportation Company. J. Ind. Ecol. 2011, 15, 937–950.
10. Matthews, H.S.; Small, M.J. Extending the boundaries of life-cycle assessment through
environmental economic input-output models. J. Ind. Ecol. 2001, 4, 7–10.
11. Hendrickson, C.; Horvath, A.; Joshi, S.; Lave, L. Economic Input-Output Models for Environmental
Life-Cycle Assessment. Environ. Sci. Technol. 1998, 32, 184–191.
12. Junnila, S. Life cycle management of energy-consuming products in companies using IO-LCA.
Int. J. Life Cycle Assess. 2008, 13, 432–439.
13. Suh, S.; Lenzen, M.; Treloar, G.J.; Hondo, H.; Horvath, A.; Huppes, G.; Jolliet, O.; Klann, U.;
Krewitt, W.; Moriguchi, Y.; et al. System boundary selection in life-cycle inventories using hybrid
approaches. Environ. Sci. Technol. 2004, 38, 657–664.
14. Majeau-Bettez, G.; Strømman, A.H.; Hertwich, E.G. Evaluation of process- and input-output-based
life cycle inventory data with regard to truncation and aggregation issues. Environ. Sci. Technol.
2011, 45, 10170–10177.
15. Lee, C.-H.; Ma, H.-W. Improving the integrated hybrid LCA in the upstream scope 3 emissions
inventory analysis. Int. J. Life Cycle Assess. 2012, 18, 17–23.
16. Wiedmann, T.; Lenzen, M.; Barrett, J.R. Companies on the Scale—Comparing and Benchmarking
the Sustainability Performance of Businesses. J. Ind. Ecol. 2009, 13, 361–383.
17. Suh, S.; Huppes, G. Missing inventory estimation tool using extended input-output analysis. Int. J.
Life Cycle Assess. 2002, 7, 134–140.
18. Danish Regions. The Regions—in Brief. Available online: http://www.regioner.dk/in+english/
publications+and+policy+papers/the+regions+in+denmark (accessed on 28 April 2015).
19. Kjær, L.L.; Høst-madsen, N.K.; Skygebjerg, L.; Mikkelsen, K.D.; Jørgensen, R. Region Sjælland
Klimaregnskab 2009–2011. Available online: http://www.niras.dk/temaer/klima/referencer/
klimaregnskab-for-region-sjaelland-2009-11.aspx (accessed on 27 May 2015).
20. Mikkelsen, K.D.; Madsen, N.K.H.; Szeler, A. Region Nordjylland Klimaregnskab 2007–2012.
Available online: http://www.rn.dk/Om-Region-Nordjylland/KlimaRegion (accessed on 27 May 2015).
21. Kjær, L.L.; Høst-Madsen, N.K.; Szeler, A.C.; Mikkelsen, K.D.; Jørgensen, R. Region Hovedstaden
Klimaregnskab 2011. Available online: http://www.niras.dk/temaer/klima/klimaregnskab.aspx
(accessed on 17 August 2015)
22. Kjær, L.L.; Mikkelsen, K.D.; Madsen, N.K.H.; Schmidt, J.H. Metoderapport: Virksomheders
klimaregnskab efter klimafodaftryksmetoden. Available online: http://lca-net.com/p/1867 (accessed
on 27 May 2015).
23. PPR/PUMA. PUMA’s Environmental Profit and Loss Account for the year ended 31 December
2010. Available online: http://glasaaward.org/wp-content/uploads/2014/01/EPL080212final.pdf
(accessed on 27 May 2015).
Sustainability 2015, 7 11460

24. Høst-Madsen, N.K.; Damgaard, C.K.; Szeler, A.; Jørgensen, R.; McManamon, D.; Bullock, S.;
Taylor, J.; Sireyjol, A.; Schmidt, J.H. Methodology report for Novo Nordisk’s environmental profit
and loss account. Available online: http://www2.mst.dk/Udgiv/publications/2014/02/
978-87-93178-03-8.pdf (accessed on 27 May 2015).
25. Høst-Madsen, N.K.; Damgaard, C.K.; Szeler, A.; Jørgensen, R.; Mikkelsen, K.D.; McManamon, D.;
Bullock, S.; Taylor, J.; Sireyjol, A.; Schmidt, J.H. Novo Nordisk’ s environmental profit and loss
account. Available online: http://www2.mst.dk/Udgiv/publications/2014/02/978-87-93178-02-1.pdf
(accessed on 27 May 2015).
26. Novo Nordisk. Our Triple Bottom Line. Available online: http://www.novonordisk.com/
sustainability/how-we-manage/the-triple-bottom-line.html (accessed on 27 May 2015).
27. Armstrong, V.N. Vessel optimisation for low carbon shipping. Ocean. Eng. 2013, 73, 195–207.
28. Kjær, L.L.; Pagoropoulos, A.; Hauschild, M.; Birkved, M.; Schmidt, J.H.; McAloone, T.C. From
LCC to LCA Using a Hybrid Input Output Model—A Maritime Case Study. Proc. CIRP 2015, 29,
474–479.
29. Leontief, W. Quantitative input-output relations in the economic system of the United States.
Rev. Econ. Stat. 1936, 18, 105–125.
30. Hendrickson, C.T.; Lave, L.B.; Matthews, H.S. Environmental Life Cycle Assessment of Goods and
Services: An. Input-Output Approach; Resources for the Future: Washington, DC, USA, 2006.
31. Lenzen, M. Errors in Conventional and Input-Output—Based Life-Cycle Inventories. J. Ind. Ecol.
2001, 4, 127–148.
32. Minx, J.C.; Wiedmann, T.; Wood, R.; Peters, G.P.; Lenzen, M.; Owen, A.; Scott, K.; Barrett, J.;
Hubacek, K.; Baiocchi, G.; et al. Input–Output Analysis and Carbon Footprinting: An Overview of
Applications. Econ. Syst. Res. 2009, 21, 187–216.
33. Suh, S.; Nakamura, S. Five years in the area of input-output and hybrid LCA. Int. J. Life Cycle Assess.
2007, 12, 351–352.
34. Finnveden, G.; Hauschild, M.Z.; Ekvall, T.; Guinée, J.; Heijungs, R.; Hellweg, S.; Koehler, A.;
Pennington, D.; Suh, S. Recent developments in Life Cycle Assessment. J. Environ. Manag. 2009,
91, 1–21.
35. Wiedmann, T. A review of recent multi-region input–output models used for consumption-based
emission and resource accounting. Ecol. Econ. 2009, 69, 211–222.
36. Suh, S. Input-output and hybrid life cycle assessment. Int. J. Life Cycle Assess. 2003, 8, 257.
37. Deng, L.; Babbitt, C.W.; Williams, E.D. Economic-balance hybrid LCA extended with uncertainty
analysis: Case study of a laptop computer. J. Clean. Prod. 2011, 19, 1198–1206.
38. Crawford, R.H. Validation of a hybrid life-cycle inventory analysis method. J. Environ. Manag.
2008, 88, 496–506.
39. Weidema, B.P.; Ekvall, T.; Heijungs, R. Guidelines for application of deepened and broadened LCA.
Deliverable D18 of work package 5 of the CALCAS project. Available online: http://lca-net.com/
publications/show/guidelines-applications-deepened-broadened-lca/ (accessed on 27 May 2015).
40. FORWAST. Available online: http://forwast.brgm.fr/ (accessed on 1 June 2015).
41. Hermansen, J.E.; Kristensen, I.S.; Nguyen, T.L.T.; Weidema, B.; Dalgaard, R.; Andersen, M.S.
Pilotprojekt—Samfundsmæssig effekt af miljøforbedrende tiltag i jordbruget. Available online:
http://lca-net.com/p/1312 (accessed on 28 April 2015).
Sustainability 2015, 7 11461

42. Schmidt, J.H.; Muños, I. The carbon footprint of Danish production and consumption—Literature
review and model calculations. Available online: http://www.ens.dk/sites/ens.dk/files/klima-
co2/klimaplan-2012/VidenOmKlima/_dk_carbon_footprint_20140305final.pdf (accessed on 28
April 2015).
43. WBCSD. WRI Corporate value chain (scope 3) accounting and reporting standard—Supplement to
the GHG Protocol Corporate Accounting and Reporting Standard. Available online:
http://www.ghgprotocol.org/standards/scope-3-standard (accessed on 27 May 2015).
44. CDP. Driving sustainable economies. Available online: https://www.cdp.net/en-US/Pages/
HomePage.aspx (accessed on 28 April 2015).
45. PPR. An Expert Review of the Environmental Profit and Loss Account—What the Experts say: The
Way Forward. Available online: http://www.kering.com/sites/default/files/e-pl-review_final-for_
publicationwebsitefinal_final_1.pdf (accessed on 27 May 2015).
46. Suh, S. Are services better for climate change? Environ. Sci. Technol. 2006, 40, 6555–6560.
47. Jira, C.F.; Toffel, M.W. Engaging Supply Chains in Climate Change. Manuf. Serv. Oper. Manag.
2013, 15, 559–577.
48. Plambeck, E.L. Reducing greenhouse gas emissions through operations and supply chain management.
Energy Econ. 2012, 34, S64–S74.
49. Wood, R.; Stadler, K.; Bulavskaya, T.; Lutter, S.; Giljum, S.; de Koning, A.; Kuenen, J.; Schütz, H.;
Acosta-fernández, J.; Usubiaga, A.; et al. Global sustainability accounting—Developing EXIOBASE
for multi-regional footprint analysis. Sustainability 2015, 7, 138–163.
50. Schmidt, J.H. Full Integration of LCA with Other Assessment Tools—New Application Areas and
Harmonized Modelling Approaches. In Proceedings of the SETAC Europe 24th Annual Meeting,
Basel, Switzerland, 11–15 May 2014. Available online: http://lca-net.com/publications/show/full-
integration-lca-assessment-tools-new-application-areas-harmonized-modelling-approaches/ (accessed
on 3 June 2015).

© 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/4.0/).