Environmental Performance of Transportation - A Comparative Study

Environmental Performance of Transportation A Comparative Study A. M. Fet1, PhD, O. Michelsen2, MSc Norwegian University of Science and Technology

H. Karlsen3, Eng. Ålesund College

SYNOPSIS Information about the environmental performance of transport systems is believed to become increasingly important in the future. Waterborne transport has traditionally been considered the best alternative from an environmental point of view. However, in recent years more attention has been put on the way environmental performance is documented. This paper shows how the environmental performance of different transport chains can be compared against each other by a set of environmental impact categories. By the use of different weighting models, slight variations in performance are observed.

INTRODUCTION This paper presents the results from the Norwegian project “Environmental Performance of Transportation - A Comparative Study”, a co-operation between the Norwegian University of Science and Technology (NTNU), Det Norske Veritas (DNV) and Ålesund College. The project is based upon the pre-project1 “Life cycle Evaluation of ship transportation - Development of methodology and testing”, which demonstrated that Life Cycle Assessment (LCA) is an appropriate method to identify and evaluate environmental impacts during the life cycle of a ship. However, it concludes that to evaluate the environmental performance of different transport chains, both methodological development, improvement of relevant databases and evaluation aspects must be addressed in further research. The goal of the present projects is therefore to establish models and guidelines for the documentation and comparison of environmental performance of different transport chains in a life cycle perspective. This requires: • a simplification of the method of evaluating the environmental performance of transport chains, • a common set of environmental impact categories for the transport sector, and • guiding principles on how to allocate infrastructure activities to the environmental burden of the transport chain.

METHODS The working method has followed the LCA-methodology2: 1. Goal and scope definition; hereunder system description and identification of relevant environmental impact categories for the transport sector. 2. Inventory analysis; hereunder a systematisation of data-collecting algorithms in accordance with the system structure. 3. Impact assessment; hereunder characterisation and valuation according to six different valuation methods. 4. Interpretation and comparison of environmental profiles of the transport chains.

Authors’ Biographies Annik Magerholm Fet is professor in environmental life cycle management at NTNU and is MSc in physics. Ottar Michelsen is a researcher at NTNU and is MSc in biology. Harry Karlsen is a lecturer at Ålesund College and has an engineering degree within marine machinery systems.

System description A transport chain is defined as the combination of different transport systems that enables transportation from A to B. This is illustrated by Fig. 1. Chain

Transport Chain

Railway transport

Road transport

Aviation transport

W aterborne transport

Stations

Terminals

Airports

Harbours

Trains

Heavy duty vehicles

Aircrafts

HSLC vessels

Rails

Roads

System

Subsystem

Fig. 1: Transport chain, transport systems and their sub-systems Impact categories Environmental impact categories are defined slightly different by different organisations. Impact categories proposed by OECD3, ISO4, LCANET5, UN6 and the Norwegian government7 are grouped mainly in two groups; ecological and human impacts, and resource use. It is important that the categories reflect changes over a period of time keyed to the problem, are reliable and reproducible, and are calibrated in the same terms as relevant policy goals or targets. The impact categories in this study are mainly based upon the OECD Core Set of indicators. However, the impact categories identified by the Norwegian government are also taken into account since they turn out to be political questions. Table I lists the considered impact categories and the most important substances that contribute to each category. Table I Relevant impact categories in this study. Impact categories Climate change Acidification Toxic contamination Local Air Pollution (dust) Photo oxidant formation Noise Eutrophication Energy consumption Land use

Substances CO2 , N2O, CH4 SO2, NOX, NH3 Pb, TBT, other anti fouling paint (Cu based) Particles NMVOC Area exposed to more than 55dBA NH3, NOX MJ m2

Case studies The project has studied three transport cases: Case 1: Paper transport Moss – Hamburg Case 2: Passenger transport Bodø – Svolvær Case 3: Frozen fish transport Ålesund – Paris They include transport means such as small air craft, cargo vessel, ferry, trailer and high speed light craft. In addition infrastructure like harbours, roads etc. are included.

Function of transport chain and functional unit A functional unit is a measure of the performance of the functional outputs. Its primary purpose is to provide a reference to which to relate the inputs and outputs. This reference is necessary to ensure comparability of LCA-results, and it is particularly critical when different systems are being assessed to ensure that such comparisons are made on a common basis. When comparing transport chains it is necessary to reflect that the distance travelled differs from alternative to alternative. This can be a significant contributor to the difference in environmental performance. Environmental performance should therefore not be expressed per km. Functional units for transport chains become: For transport of goods: ton per route described between point A and B. For transport of passengers: passengers per route described between point A and B.

INVENTORY PRINCIPLES The principles for describing the transport chain and collecting data for the systems and sub-systems are as follows: 1. Describe the function of the transport chain. 2. Describe the transport chain and its combination of transport systems between point A and B. 3. Describe the transport system; the means and the infrastructure, the typical total route of the transport means, distances, time use, and if the transport means have multiple purposes, e.g. transport of passengers and cargo. 4. Decide level of operational profile; shall emissions be based on average figures or a detailed description of machinery load. 5. Describe the operational parameters for each transport sub-system, capacity and exploited capacity: • For a vehicle: Average parameters are satisfactory. • For an aeroplane: Separate between take off, cruising, landing and idle. • For a ship: Separate between sailing and idle (lying at quay), preferably also manoeuvring. • For a terminal/harbour: Goods treatment and energy consumption, energy requiring activities needed to handle the goods, percentage of passengers or treated goods per year. • For an airport: Ground activities like cleaning, defrosting, fuel service, passenger facilities.

Calculation and allocation principles The amounts of substances contributing to the impact categories, see Table I, are calculated by the following formulas.

Ei is the total emission of substance i [kg per trip]. Exhaust gas emissions can be calculated after (1) or (2): Ei = F ⋅ D ⋅ eai • • •

F is the fuel consumption [kg/km]

D is the distance from point A to point B for each transport means [km] eai is the exhaust emission factor for substance i [kg/kg] Ei = P ⋅ edi ⋅ T

• • •

(1)

(2)

P is engine power (average or a detailed power-pattern) [kW] edi is the exhaust emission factor for the substance i [kg per kWh] T is time [h]

Dust or particulars are calculated by (3)

Ei = D ⋅ ebi ⋅ F • • •

(3)

F is the fuel consumption [kg/km] ebi is the emission factor for substance i [g/kg fuel] D is the distance from point A to point B for each transport means [km]

Leakage of eco - toxic substances from ship antifouling by (4)

E i = T ⋅ eci ⋅ A •

• •

(4)

eci is the leakage factor for eco-toxic substance i [g/(m ·h)] T is time [h] A is area of wet surface [m2 ] 2

The calculation of land area consumption is based on the sum of the area required at any time during the transport. The land area required for the transport of cargo and passengers has to be allocated to the transport chains according to their degree of use of the area, e.g. by time used, number of operations, amount of passengers and cargo or economic turnover. For loading/transfer the required area may be shared with several users at the same time. For the allocation of land use to the transport chain the actual occupied area can be calculated by multiplying the time share of the area normalised with one year and multiplied by the fraction the actual transport means i is part of the total loading / transfer.

Occupied land area a r according to time used is calculated by (5). If other allocation principles are used the area may be calculated after similar looking formulas.

a r = Br ⋅ Lr ⋅ k i ⋅ t r / T y •

B r is the breath of the road/quay [m]

• • • •

L r is the length of the transport route/vessel [m] k i is the fraction of total activity t r denote the time used on the area for given sub-system [h] T y is one year [h]

The total area consumption

(5)

An due to noise is expressed as the area exposed to noise levels above 55 dBA4. It is estimated

by formula (6). A typical allocation principle is to multiply with the fraction reflecting the subsystems share of the total activity.

An = 2 ⋅ Rn ⋅ Dr • •

(6)

Rn is the radius where the noise is above the defined level [m] Dr is the length of the route [m]

For multipurpose-vessels it is necessary to relate the emissions

Ei* =

Ei* to the functional unit: Ei ⋅M C

•

Ei is the total emitted substances [g]

• •

C is the exploited capacity [ton] M is the amount transported reflected by the functional unit [ton per FU]

(7)

APPLICATION OF DATA COLLECTING PRINCIPLES TO THE TRANSPORT OF FROZEN FISH. This paper mainly present Case 3 where two transport chains are evaluated; Chain A from Ålesund harbour with ship to Ijmuiden harbour and trailer to Paris, and Chain B from Ålesund with trailer to Oslo harbour, ferry to Kiel harbour and trailer to Paris, see Fig.2.

Chain B

Chain A

Fig.2: Transport Chain A and B for frozen fish between Ålesund and Paris. 4

This is the limit set by the Norwegian Ministry of Environment for new buildings and roads.

1

The function of the transport chain is to transport 500 tons frozen fish between Ålesund and Paris. The functional unit for the transport chains is 1 ton frozen fish transported from Ålesund to Paris.

2 3

Chain A is a combination of waterborne (67% of distance, 1 ship) and road transport (33 % of distance, 22 trailers). The waterborne transport consists of the harbours in Ålesund, Måløy and Ijmuiden and the RoRo-ship M/V Nordjarl between Ålesund and Ijmuiden. The road transport consists of the terminal in Ijmuiden, the trailer and the road between Ijmuiden and Paris, see Fig. 3. Transport Chain A

Road transport

Waterborne transport

Terminal in Ijmuiden

Harbours in Ålesund, Måløy and Ijmuiden

Termo Trailer HFR/Norfrig

RoRo-ship M/V Nordjarl

Road Ijmuiden - Paris

Fig. 3: Chain A 4 5

The level of detailing is based upon the principles to gather sufficient data on resource use (energy and fuel) and the use of areas to be able to calculate the most important environmental aspects. For the sub-systems in Chain A the following data are collected: • Fuel consumption per km. • Travelling distance, time and velocity. • Load capacity, exploited capacity and fish load. • Average load of machinery and exhaust emission factors for machinery systems. • Wet surface area (for ship). • Area exposed to noise > 55 dBA. • Area and time in harbour /terminal. • Energy use by handling equipment and by freezing units.

Similar principles are used to describe the transport chain B; a combination of road and waterborne systems. The road system consists of a VolvoFH-12 connected to a termo-trailer HFR/Norfrig and the road between Ålesund and Oslo and between Kiel and Paris. The waterborne system consists of the ferry M/V Kronprins Harald (or M/V Prinsesse Ragnhild) and the harbours in Oslo and Kiel. Exhaust emission factors for ships are based on Lloyd’s8, emission factors for trailers are based on Norwegian statistical data9, distance-calculations are based on net-information10, fuel consumption, time-data, capacity etc. are based on personal communications11, and data on harbour areas are based on figures from Sintef12. The leaching rate of TBT is based on figures from IMO13. For a ferry the total emissions and consumption are allocated to one trailer by dividing on the ferry-capacity and multiplying with the share of the capacity that the trailer occupies (8):

Etrailer , i = • • • • •

E ferry ,i ⋅U C

(8)

Etrailer,i is the share of the ferry`s total emissions that are allocated to the frozen fish transport [g]. Eferry,i is the total ferry emissions of substance i as calculated in equation (1) or (2) above [g]. C is the exploited capacity for the ferry (measured in private car units). U is the number of private cars units that the trailer occupies. The emission (Etrailer,i = Ei ) are then related to the functional unit according to equation (7).

INVENTORY RESULTS By using formula (1), (2) and (7) the emissions to air are calculated for every substance within each impact category. The calculations are based on fuel consumption for main machinery systems, auxiliary engines, for vehicle driving and for freezing. From each sub-system in the transport chain the total amount of each substance are summarised and presented in the inventory table, Table II.

TBT-leakage is calculated after formula (4) by using a leakage-rate of 0,0017 g/(m2· h). The result is then multiplied by exploited capacity and divided with capacity and by transported tons fish. This gives the leakage per ton fish transported. For the ferry formula (7) must be used in addition. The occupied harbour area is calculated by using the vessel length, quay with, time in harbour, traffic and goods flows11. The calculations on area occupation due to trailer traffic is based on vehicle length and with, average speed, time on road, number of vehicle per functional unit, see formula (5). The area exposed to noise >55dBA is based on noise measurements.11 and statistical data. For the trailers it is assumed that the 55 dBA limit is exceeded only when driving. Formula (6) is used and allocation is based on share of total activity. Table II: Inventory results per ton frozen fish, Chain A and Chain B. Impact category Climate change

Acidification

Toxic contamination Local Air Pollution (Dust) Photo oxidant formation Noise Eutrophication Energy consumption Land use

Substance CO2 N2O CH4 SO2 NOX NH3 Pb TBT Particles NMVOC Area exposed to more than 55dBA NH3 NOX MJ m2

Transport chain A 84 kg 0,24 g 1,5 g 938 g 1286 g 0,022 g (no data) 0,10 g 24 g 36,6 g 10,4 m2 0,022 g 1286 g 930 MJ 0,23 m2

Transport chain B 138 kg 0,71 g 4,4 g 867 g 1802 g 0,064 g (no data) 0,034 70 g 106 g 94 m2 0,064 g 1802 g 1812 MJ 0,66 m2

ENVIRONMENTAL IMPACT ASSESSMENT Characterisation The total environmental impact within each category j, EP(j), is the sum of each substance multiplied with its characterisation value after formula (8)14. EP(j) = Σ(Qi • EF(j)i) • •

(8)

Qi is the emissions of compound i, EF(j) i is the characterisation factor of compound i related to impact category j.

Normalisation Different valuation methods require different normalisation values. Ideally the impact categories should be normalised according to the total contribution within each category (i.e. climate change should be normalised according to global emissions and noise to local noise level), but as a simplification the different impact categories are normalised against total Norwegian emissions or consumption14. The normalisation-value, MP(j), within each category j is calculated by formula (9). MP(j)=Σ(ENi • EF(j)i) • •

(9)

ENi is the total emission or consumption of compound i in Norway, and EF(j)i is the characterisation factor of compound i related to impact category j.

The value is then used to find the relative contribution for each impact category by dividing the potential contribution EP(j) with the normalisation value for the category MP(j). Table III shows characterisation and normalisation values. Fig. 4 shows the characterised and normalised results of the inventory of transport chain A and B. It shows that Chain A has a lower impact than chain B within each impact category.

Table III: Characterisation and normalisation values. Impact category

Compound

Climate change14

CO2 N2O CH4 SO2 NOX NH3 Pb (to air) TBT Cu particles NMVOC Area >55dBA (m2) NH3 NOX MJ Area (m2)

Acidification14

Toxic contamin. 23

Local air pollution Photo oxidant form. Noise Eutrophication14 Energy consumption Land use

Reg. value (Qi)

Characterisation (EF(j) i) 1 320 25 1,00 0,70 1,88 160 250 2 1 1 1 3,64 1,35 1 1

Contribution (Qi • EF(j)i)

Normalisation

EP(j)

55 598 000 000

EP(j)

237 448 000

EP(j)

8 453 000 344 700 000 24 800 000 36 146 088 884

EP(j)

671 081 500 813 PJ 485 719 000

0,00000001 0,000000009

Realtive contribution

0,000000008 0,000000007 0,000000006

Transport chain A

0,000000005

Transport chain B

0,000000004 0,000000003 0,000000002 0,000000001

ns

us nd

co

La

pt um

at ic ph tro

e

n io

n io

se oi N

En

er

gy

Eu

C

lim

at

e

ch

an

ge Ac id ifi ca To tio xi n c co nt am Ph in ot at o io ox n id an tf or Lo m ca at la io ir n po llu tio n (d us t)

0

Impact category

Fig. 4: Normalised inventory results for fish transport. Chain A is mainly waterborne, Chain B is mainly land transport.

Valuation Valuation techniques are used to compare the relative importance of different environmental impact categories. Weighting methods require ethical and ideological values and the values are difficult to evaluate15, 16. ISO4 define weighting as “the process of converting indicator results by using numerical factors based on value choices”. In this study these methods are used: • The Eco-indicator 9917 • EPS18 • The ExternE Methodology19 • Valuation according to political goals

• •

Valuation according to panel procedures Valuation according to the recommendations in the OECD project on Environmentally Sustainable Transport (EST)20

The Eco-indicator 99 is based on a damage function approach, both EPS and ExternE are monetary methods while EPS uses a wide range of indicator. ExternE only focus on emissions of CO2, NOX and particles. Valuation according to political goals is based on distance-to-target, where the targets are the political goals for environmental improvement in Norway7. The valuation according to panel procedures reflects the relative weight a panel put on the selected impact categories. The valuation according to political goals and panel procedures are based on characterisation and normalisation values in Table III, whereas the methods Eco-indicator 99, the EPS and the ExtrenE are based on characterisation values and weighting factor incorporated in the methods. The valuation according to the OECD-recommendations is also a distance-to-target method where reduction targets in proportion to 1990-level of CO2, NOX, VOCs, particles, noise and land-use/land-change are used.

INTERPRETATION Fig. 5 shows the results of the valuation according to political goals, whereas Fig. 6 shows the results according to the OECD-recommendations. As the figures show, chain A has a better environmental performance than chain B. The other valuation methods show similar results. This should be of no surprise since the environmental performance of chain A is better for all impact categories except toxic contamination (TBT), see Fig. 4. Weighting is in this case actually superfluous to decide which chain that has the best environmental performance. However, there are some differences in which categories and compounds that contribute most to the environmental burdens recognised in the different methods. Whereas acidification (NOX) is the most important category in valuation according to political goals, EST and ExternE, climate change is the single most important category when valuation is done according to panel procedures while fuel consumption when using EPS. If valuation results are used to identify where the most important potentials for improvements are, this is important to notice. The impact categories used in Eco-indicator 99 are rather different from the other methods, so directly comparison is difficult.

Relative environmental performance of transportation of 1 ton frozen fish.

Land use Energy consumption Eutrophication Noise Local air pollution (dust) Photo oxidant formation Toxic contamination Acidification Climate change

Transport chain A

Transport chain B

Fig. 5: Valuation according to political goals..

Relative environmental performance of transportation of 1 ton frozen fish.

Land use Noise Particles VOC NOX CO2

Transport chain A

Transport chain B

Fig. 6:Valuation according to recommendations in the EST project..

Results from Case 1 and Case 2 The interpretation of the results from Case 121 and Case 222 is not as clear as for Case 3 since non of the transport chains in these cases (airplane, high speed vessel, trailer, ship) shows the best environmental performance for all impact categories. This means that valuation is necessary if overall environmental performance for the transport chains is to be compared. However, the results from the valuation turn to be very similar also for these cases, independent of valuation method. In Case 2 (passenger transport) the transport chain mainly based on a small aircraft shows the best performance, followed by a transport chain based on high speed vessel. The transport chain based on private car shows the worst environmental performance. In Case 1 (paper transport) the different valuation methods give different answers to what transport chain that has the best environmental performance. Both the valuation according to political goals and the Eco-indicator 99 indicate that a trailer based transport chain has the best performance, whereas the other methods indicate that a cargo vessel based transport chain has the best performance. The differences between the chains are however small for all valuation methods. The weight on acidification in the different valuation methods is crucial.

DISCUSSIONS AND CONCLUSIONS The paper has demonstrated models for how to compare the environmental performance of transport chains. However, it has not shown how to optimise each chain. This will require more detailed data on machinery systems. It was decided to only study the operational phase since previous studies23 show that cradle to gate data for fuel contribute less than 10% to the impact caused by the combustion of the fuel during the life time of a transport means, and the building of the subsystems contribute less than 1% to the total environmental burdens24. Also the maintenance of the transport systems will give minimal contribution. These conclusions depend however on the system boundaries. The importance of the impact categories is discussed. The impact category toxic contamination (TBT, Pb, defrosting fluid etc.) is difficult to evaluate since local impacts are not included in some of the evaluation models. The use of land-area and the effects of noise exposure were also evaluated. We see from Fig. 4 that land use contribute minimal to the total environmental burden. However, the results show that for chain B which to a great extent is a route through a dens populated areas, noise should not be neglected as an important impact. The results seam to turn out very similar independent of which of the six valuation methods that are used. So far the project concludes that a simple weighting method is recommended, e.g. the six indicators recommended in the OECD EST project, see Fig. 6.

The preliminary results are interesting information for further research, for decision making for transport companies and for governmental bodies. The transport companies may use such information to report the environmental performance of transport chains and to plan their logistics. For governmental bodies the information can be used for taxation. It is important to notice that a means is a part of a transport chain, and therefore taxes should be put on the transport chain instead of the single means. Databases with environmental performance data for transport chains, not only for single transport means, should also be developed. At last the project results are of great value for further research on how to optimise the economic and environmental performance of transport chains, and for the development of eco-efficiency indicators26 for transportation.

ACKNOWLEDGEMENTS The authors want to thank other participants in the project for valuable contributions to the discussion of the methodology and the weight models. Especially to T. Johnsen DNV who has been responsible for performing the study behind Case 1 and 2, to E. Sørgård DNV for his contribution to the structuring of the project, and to A. Brekke and E. Hertwich at the LCA-laboratory at NTNU. They have contributed in the valuation process.

REFERENCES 1. 2. 3. 4. 5. 6.

7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18.

Fet, A. M. (HiÅ), Sørgård, E. (DNV): “Life Cycle Evaluation of Ship Transportation – Development of Methodology and Testing”. Report no. 10/B101/R-98/008/00, Ålesund College, 1998. ISO 14040: “Environmental management – Life Cycle Assessment – Principles and framework”, International Organization for Standardization (1997) OECD: “OECD Core Set of Indicators for Environmental Performance Reviews”, Environmental Monographs N° 83. OECD/GD893)179 (1993) ISO 14042: “Environmental management – Life Cycle Assessment – Life Cycle Impact Assessment”, International Organization for Standardization (2000) Finnveden, G. and Lindfors, L.G., ‘LCANET Theme Report. Life Cycle Impact Assessment and Interpretation’, http://www.leidenuniv.nl/interfac/cml/lcanet/ftheme3.htm (february 1997) The World Resources Institute, The United Nations Environment Programme, The United Nations Development Programme and The World Bank, “World Resources 1998-1999. A Guide to the Global Environment”, Oxford University Press (1998) Norwegian Ministry of the Environment: “Regjeringens miljøvernpolitikk og rikets miljøtilstand”, Stortingsmelding nr. 8 (1999-2000) (”Norwegian white paper”) (October 1999) Lloyd’s Register of Shipping: “Marine exhaust gas emissions research programme: Steady state operation”, London 1990. Bang, J., et al.: Utslipp fra vegtrafikk i Norge. Dokumentasjon av beregningsmetode, data og resultater. SFT rapport 99:04. Statens forurensningstilsyn, Norway (1999). http://www.cwlease.com/cwlint/index2htm Personal communication (ref Technical report: Karlsen, H, “Frozen Fish transport from Ålesund to Paris”, Ålesund College, 2000): • Nor-Cargo, Ålesund (Ytterland E.), • Stavangerske Linjefart AS (Valen H.), • Volvo Norge (Nordvik. D), • Waagan Transport AS (Kleppe, A. J.), • Carrier Transicold of Scandinavia, Denmark, • Global Fish International AS (Eikrem. J.), • M/V NORDJARL, the captain. Aanondsen et al., ”Exponofi-project: Detaljert beskrivelse av aktuelle havner og terminaler”, SINTEF Fiskeri og havbruk AS, May 1999. IMO, “Measures to control potential adverse impacts associated with the use of tributylin compounds in antifouling paints”, MEPC. 46 (30), 1990. http://www.ssb.no/emner/01 Wenzel, H. et al: ‘Miljøvurdering av produkter’, Miljø- og Energiministeriet – Miljøstyrelsen & Dansk Industri (1996) Finnveden, G., ‘A Critical Review of Operational Valuation/Weighting Methods for Life Cycle Impact Assessment’, AFR-Report 253, Swedish Environmental Protection Agency (June 1999) Goedkoop, M. and Spriensma, R. ‘The Eco-indicator 99: A damage oriented method for Life Cycle Impact Assessment (Preliminary Internet version)’, PRé Consultans (October 1999) Steen, B. ‘A systematic approach to environmental priority strategies in product development (EPS). Version 2000 – Models and data of the default method’, CPM report 1999:5. Centre for Environmental Assessment of Products and Materials (CPM), Chalmers University of Technology (1999)

19. ExternE, ‘Externalities of Energy – What is the ExternE Methodology’, http://externe.jrc.es/overview.html (1997) 20. Wiederkehr, P. ‘Environmentally Sustainable Transport – International perspectives. OECD’s EST Project’. In OECD ‘Project on Environmentally Sustainable Transport (EST) – The Economic and Social Implications of Sustainable Transportation – Proceedings from the Ottawa Workshop’, ENV/EPOC/PPC/T(99)3/FINAL/REV1 (1999) 21. Johnsen, T., ”Environmental comparison of alternative transport chains for paper. A case study”. Det Norske Veritas, Technical Report No. 2000-3295, Rev. 0, 2000. 22. Johnsen, T., ”Environmental comparison of alternative transport chains for passengers. A case study”. Det Norske Veritas, Technical Report No. 2000-3296, Rev. 0, 2000. 23. Johnsen, T. (DNV); Fet, A.M. (HiÅ): “Screening Life Cycle Assessment of M/V Color Festival”. Report no. 10/B101/R-98/009/00, Ålesund College, 1998. 24. Fet, A. M., “Systems Engineering Methods and Environmental Life Cycle Perfromance within Ship Industry”, Doktor Ingeniøravhandling 1997:21, Institutt for termisk energi og vannkraft, NTNU, ITEV-rapport 1997:1. 25. Fet, A. M., Michelsen, O., Johnsen, T., Sørgård, E., Environmental performance of transport –A comparative study. In prep. Norwegian University of Science and Technology, Trondheim, Norway. 26. Thorvik, A. ”Corporate Sustainability reporting”, World Business Council for Sustainable Development, Geneva, 2000.