Benchmarks and measures for better fuel efficiency. How AIS data can be used in operational performance analysis. Master of Science Thesis in Nordic Master in Maritime Management JOHAN WIGFORSS Department of Marine and Shipping Technology CHALMERS UNIVERSITY OF TECHNOLOGY Goteborg, Sweden, 2012 Report No. NM-12/29 REPORT NO. NM-12/29 Benchmarks and measures for better fuel efficiency. How AIS data can be used in operational performance analysis. JOHAN WIGFORSS Department of Shipping and Marine Technology CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden, [2012] Benchmarks and measures for better fuel efficiency. How AIS data can be used in operational performance analysis. NAME A. FAMILYNAME © JOHAN WIGFORSS, [2012]. Report no NM-12/29 Department of Shipping and Marine Technology Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone + 46 (0)31-772 1000 Printed by Chalmers Göteborg, Sweden [2012] i Benchmarks and measures for better fuel efficiency. How AIS data can be used in operational performance analysis. JOHAN WIGFORSS Department of Shipping and Marine Technology Chalmers University of Technology ABSTRACT Shipping will face an escalating competition in the future, as more stringent environmental regulations will lead to significant higher fuel costs. Today, the cost of fuel stands for approximate 35-70% of total operational cost. Fuel efficiency measures are vital in order to stay competitive in the future. The issue with the study is to examine how AIS data can be used to compare ships against each other with appropriate benchmarks in order to identify measures for better fuel efficiency. A case study of 44 general cargo ships was carried out with AIS data from 2010-2011. These were two sister groups of 7 700 dwt and 12 700 dwt, with 22 ships in each group. Each group of sister ships were selected from their design and configuration in order to eliminate any design configuration differences in the operational analysis. Disturbance in AIS data was corrected and only voyages with coherent data without time gaps were used in analysis. Ships in study show on a significant potential of improvement in terms of fuel efficiency. Short periods at high speed increase the average fuel consumption in total. All ships were operated at a significant higher average speed than the best economic speed, i.e. lowest cost per nautical mile. There were also tendencies of differences between the operators, where some operators tend to run their ships at a more fuel-efficient way than others. Capacity utilization analysis indicated a spare of 10-20% before hitting the optimum span, which show that fuel efficiency can be improved by increasing the output of the ships i.e. more cargo. However, the most important fuel efficiency measure is speed reduction, i.e. slow steam. The theoretical no anchoring strategy calculations confirm that there are great possibilities to minimize anchoring time in favour of speed reduction. Keywords: shipping, ship operation, fuel consumption, fuel efficiency, AIS data analysis, slow steaming, operational performance. ii ACKNOWLEDGEMENT This master thesis was accomplished under the supervision of Department of Shipping and Marine technology at Chalmers University of Technology with idea of study from the IHS Fairplay in Gothenburg. First of all I would like to thank Torbjörn Rydbergh and Niklas Bengtsson at IHS Fairplay Gothenburg. Without your input and supervision, this study would not have been possible. I would also like to thank my supervisor at Chalmers, Anna Eliasson, for interesting thoughts and encouragement in the writing process of the thesis. Finally I would like to thank Karin, for your patience and support during long hours in front of the computer sorting AIS data. Gothenburg, May 2012 Johan Wigforss TABLE OF CONTENTS iii TABLE OF CONTENTS 1 INTRODUCTION .............................................................................................................. 1 1.1 EMISSIONS FROM SHIPPING ............................................................................................. 1 1.1.1 Regulations ............................................................................................................. 1 1.1.2 Sulphur ................................................................................................................... 2 1.1.3 New regulations effect on shipping ........................................................................ 2 1.1.4 New regulations effects on industries ..................................................................... 3 1.2 MAXIMISE EFFICIENCY FROM ASSETS ............................................................................. 4 1.3 PURPOSE ......................................................................................................................... 4 1.3.1 Research questions ................................................................................................. 4 1.4 DELIMITATION ............................................................................................................... 4 1.5 OUTLINE OF THE REPORT ................................................................................................ 5 2 THEORY ............................................................................................................................. 6 2.1 COST STRUCTURE ........................................................................................................... 6 2.2 SECTORS OF SHIPPING ..................................................................................................... 6 2.2.1 Liner shipping ........................................................................................................ 6 2.2.2 Tramp shipping ...................................................................................................... 7 2.2.3 Industrial shipping ................................................................................................. 7 2.2.4 Charter agreements ................................................................................................ 7 2.3 ENERGY EFFICIENCY ...................................................................................................... 7 2.4 FUEL CONSUMPTION ....................................................................................................... 8 2.5 ENERGY EFFICIENCY MEASURES ..................................................................................... 9 2.5.1 Optimisations by the master. .................................................................................. 9 2.5.2 Weather routing ...................................................................................................... 9 2.5.3 Just in time ........................................................................................................... 10 2.5.4 Trim optimisation ................................................................................................. 11 2.6 CHOICE OF SPEED ......................................................................................................... 11 2.6.1 Slow steaming ....................................................................................................... 11 2.6.2 Economic speed .................................................................................................... 11 3 METHODOLOGY ........................................................................................................... 15 3.1 QUANTITATIVE AND QUALITATIVE RESEARCH .............................................................. 15 3.2 CASE STUDIES .............................................................................................................. 15 3.3 DATA COLLECTION ....................................................................................................... 15 3.4 RELIABILITY, VALIDITY, AND OBJECTIVITY .................................................................. 16 3.5 RESEARCH APPROACH .................................................................................................. 16 4 CASE STUDY ................................................................................................................... 18 4.1 AUTOMATIC IDENTIFICATION SYSTEM (AIS) ................................................................ 18 4.1.1 AISLive ................................................................................................................. 19 4.2 SHIPS IN STUDY ............................................................................................................ 20 4.2.1 Fuel consumption ................................................................................................. 20 4.2.2 Hydrostatic data ................................................................................................... 21 4.3 AIS DATA ..................................................................................................................... 21 4.3.1 Confidence and uncertainty of data ..................................................................... 23 5 RESULT ............................................................................................................................ 25 5.1 OPERATIONAL PROFILES ............................................................................................... 25 5.2 FUEL CONSUMPTION ..................................................................................................... 28 iv 5.3 TRANSPORTATION WORK .............................................................................................. 29 6 ANALYSE ......................................................................................................................... 30 6.1 SPEED ........................................................................................................................... 30 6.1.1 Potential savings of speed reduction .................................................................... 32 6.1.2 Operator differences ............................................................................................ 32 6.2 PORT ............................................................................................................................ 34 6.3 ANCHORING ................................................................................................................. 36 6.4 FUEL EFFICIENCY ......................................................................................................... 37 6.4.1 Capacity utilization .............................................................................................. 38 6.5 THEORETICAL SAVINGS ................................................................................................ 39 7 CONCLUSION ................................................................................................................. 41 7.1 AIS DATA ..................................................................................................................... 41 7.2 FUEL EFFICIENCY BENCHMARKS AND MEASURES ......................................................... 41 8 FUTURE STUDY ............................................................................................................. 43 9 REFERENCES ................................................................................................................. 44 APPENDIX 1 - ECONOMIC SPEED .................................................................................. 46 APPENDIX 2 – SHIPS IN STUDY ....................................................................................... 49 LIST OF FIGURES Figure 1.1; Existing, and possible future emission control areas. (DNV, 2012) ....................... 2 Figure 1.2; HFO and MGO price 2010-2011 in Rotterdam. ...................................................... 3 Figure 3.3; Despatch. Loading/Unloading is finished before laytime ends. ............................ 10 Figure 3.4; Economic speed. Thick line show cost per nautical mile with use of normal bunker fuel (HFO 1%). Dotted lines show cost per nautical mile with low sulphur fuel (MGO 0,1%). .................................................................................................................... 12 Figure 3.5; Shared stakeholder benefit of a speed reduction from service speed. ................... 14 Figure 4.1; AIS system in practice. .......................................................................................... 18 Figure 4.2; Fuel consumption in ton/day at different speeds. .................................................. 21 Figure 4.3; Example of ship’s visualization of track on map. .................................................. 22 Figure 4.5; Definition of voyage in study. ............................................................................... 23 Figure 5.1; 7 700 dwt, fuel consumption per hour. .................................................................. 28 Figure 5.2; 12 700 dwt, fuel consumption per hour. ................................................................ 28 Figure 5.3; 7 700 dwt, transportation work per hour................................................................ 29 Figure 6.1; Speed distribution of the 7 700 dwt ships. ............................................................. 30 Figure 6.2; Speed distribution of the 12 700 ships. .................................................................. 30 Figure 6.3; 7 700 dwt, average speed without and with cargo on-board. ................................ 31 Figure 6.4; 12 700 dwt, average speed without and with cargo on-board. .............................. 31 Figure 6.5; 7 700 dwt, fuel consumption vs. average speed. ................................................... 32 Figure 6.6; 12 700 dwt, fuel consumption vs. average speed. ................................................. 33 Figure 6.8; 7 700 dwt, Time in port per voyage. ...................................................................... 34 Figure 6.9; 12 700 dwt, time in port per voyage. ..................................................................... 34 Figure 6.10; 7 700 dwt, hours in port with loading operation .................................................. 35 Figure 6.11; 7 700 dwt, hours in port with unloading operation. ............................................. 35 Figure 6.12; 7 700 dwt, loading/unloading rate. (Tonne per hour). ......................................... 36 Figure 6.13; 7 700 dwt, anchoring time before loading/unloading of cargo per voyage. ........ 37 TABLE OF CONTENTS v Figure 6.14; 12 700 dwt, anchoring time before loading/unloading of cargo per voyage. ...... 37 Figure 6.15; 7 700 dwt, capacity utilization. ............................................................................ 38 Figure 6.16; 7 700 dwt, theoretical fuel consumption savings. (Tonne/hour). ........................ 39 Figure 6.17; 12 700 dwt, theoretical fuel consumption savings. (Tonne/hour) ....................... 40 LIST OF TABLES Table 1.1; Present, and upcoming sulphur regulations from the IMO. (IMO, 2012)................. 2 Table 1.2; Increase in container freight rates as a consequence of the low sulphur fuel regulations. (Kalli, Karvonen, & Makkonen, 2009) .......................................................... 3 Table 3.1; Daily charter rate and fuel price. ............................................................................. 12 Table 3.2; Best economical speed at different fuel prices. ....................................................... 13 Table 3.3; Costs associated with ship operation. ..................................................................... 13 Table 4.1; Example of AIS data. .............................................................................................. 20 Table 4.2; Ships data. ............................................................................................................... 20 Table 4.3; Fuel consumption at service speed and at berth/anchor. ......................................... 20 Table 4.4; Tonne of deadweight to change draught. (Bodewes Shipyards bv, 2012) .............. 21 Table 4.5; Extract from AIS data with distance and speed manually calculated. .................... 23 Table 4.6; Confidence and uncertainty of AIS data. ................................................................ 24 Table 5.1; List of operators. ..................................................................................................... 25 Table 5.2: Operational profiles. ................................................................................................ 26 Table 5.3; 7 700 dwt, Min, max, and average percentage of time spent in each operational mode. ................................................................................................................................ 27 Table 5.4; 12 700 dwt, Min, max, and average percentage of time spent in each operational mode. ................................................................................................................................ 27 Table 6.1; Potential savings in cost per day. ............................................................................ 32 Table 6.2; 7 700 dwt ships energy intensity. ............................................................................ 38 vi GLOSSARY AIS = Automatic Identification System Bunker = Ship fuel Demurrage = Compensation to the ship owner if cargo handling in port takes longer time than agreed upon. Despatch = Compensation to the charter if cargo handling in port are finished in in advance of what is agreed upon. Dwt = Deadweight tonnage. A measure of how much a ship can safely carry. It is the sum of cargo, fuel, ballast water, fresh water, stores, and crew. ECA = Emission Control Area. Environmental sensitive area designated by the IMO. Economic speed = The speed that gives the lowest cost (shipping cost) per nautical mile. General cargo ship = Ship that can carry packed items, including containers. HFO = Heavy Fuel Oil. IMO = International Maritime Organisation Liner ship = Ship engaged in systematic liner trade. Runs according to a fixed schedule, just like a buss service. MARPOL = International Convention for the Prevention of Pollution from ships by the IMO. MGO = Marine Gas oil. RoRo ship = Roll-on/Roll-off. Ships designed to carry wheeled cargo. SECA = Sulphur Emission Control Area. Environmentally sensitive area, with special rules of sulphur content in bunker fuel, designated by the IMO. Slow steam = Speed reduction of cruising speed. Tramp ship = Ship engaged in tramp trade, i.e. operated without a fixed schedule. Just like a taxi service. Transportation work = A unit of freight measured in tonne-km. The output of moving one tonne of cargo one kilometre. Voyage = A journey from one port to another port. In study, a voyage is defined as the time a ship arrives a port till the time a ship arrives next port. METHODOLOGY 1 1 INTRODUCTION The shipping industry will face an escalating competition from other modes of transport in the future, as shipping companies have to comply with new, more stringent environmental regulations, which will lead to increase of sea transportation cost, due to significant higher fuel price. Efficiency measures are vital for the shipping company and for the industry, in order stay competitive in the future. 1.1 EMISSIONS FROM SHIPPING Emissions from shipping have become a hot topic during the last years. Emission of sulphur dioxide (SO2) from shipping is now exceeding the emissions from emission sources on land, including traffic. Nitrogen oxide (NOx) emissions from shipping are also likely to exceed emissions from land sources in only a few years. (Transportgruppen, 2012) Emissions to air from shipping affect environment and human health in different ways. SO2 and NOx are harmful to the natural environment as they cause acid rain. NOx and volatile organic compound (VOC) are also helping ozone to be created close to ground, which could be harmful to human health and vegetation. Emissions of NOx also contribute to eutrophication, which could harm the sensitive balance in the land and marine ecosystem. (Transportgruppen, 2012) Small particles, that are harmful for human health, are created when SO2 and NOx oxides in the atmosphere to sulphur and nitrogen particles that binds to dust and sot. Studies show on increased unhealthy and shortened lifetime of the population near the coastlines around Europe, where shipping is a large source of the environmental emissions. Carbon dioxide (CO2) and other greenhouse gas emissions from ships contribute to global changes in the climate. (Transportgruppen, 2012) It is possible to reduce emissions from shipping by technical measures on-board the ship. However, reducing the fuel consumption is the most important measure in order to minimize environmental impact, as emissions are in direct relation to fuel consumption. 1.1.1 REGULATIONS Shipping is an international business. Regulations to protect the environment must be implemented on highest international level, due to the nature of the business, i.e. ships travel the globe in and out of national waters. Prevention of pollution of marine environment through ship operations and accidents are covered in the MARPOL convention. The aim with the convention is to prevent and minimise both accidental pollution and pollution from routine ship operation. The convention consists of six annexes, where annex VI cover the prevention of air pollution from ships. It includes limits on sulphur oxide, nitrogen oxide, particulars from exhaust, and emissions of ozone depleting substances. (IMO, 2012) 2 1.1.2 SULPHUR The global limit of sulphur content in maritime fuel is today 3.5%. This limit will be reduced over the next years, and by 2020 the limit of sulphur content is 0.5% globally. However, some areas are especially sensitive and have more stringent regulations, i.e. emission control areas (ECA), see Figure 1.1. Figure 1.1; Existing, and possible future emission control areas. (DNV, 2012) One example of an ECA is the Baltic Sea, which is a sulphur emission control area (SECA), with more stringent sulphur regulations. The limit of sulphur content in bunker fuel in the SECAs is today 1%. By 2015, the limit will be reduced to 0.1%, see Table 1. (IMO, 2012) Year SECA Globally Present (2012) 1.0% 3.5% 2015 0.1% 2020/2025 0.5% Table 1.1; Present, and upcoming sulphur regulations from the IMO. (IMO, 2012) The date of global reduction of sulphur limit is not yet set. In 2018, IMO will analyse the global supply and demand of maritime fuel with low sulphur content. New global regulations will come into force in 2020 if the supply meets the demand. However, if there is a shortage in supply, the regulation will come into force by 2025. (IMO, 2012) 1.1.3 NEW REGULATIONS EFFECT ON SHIPPING A study made by Sjöfartsverket (2009) indicate that the cost of fuel for ships operating in the Baltic Sea will increase with approximately 70%, as a result from the implementation of the new regulations from the IMO, and the use of a fuel type with lower sulphur content, i.e. change from HFO to MGO (Figure 1.2). METHODOLOGY 3 Figure 1.2; HFO and MGO price 2010-2011 in Rotterdam. Research made by Kalli, Karvonen, & Makkonen (2009) indicate that the increase of fuel cost will increase freight cost of a container on the Baltic Sea with approximately 44-55%, see table 2, as the cost of fuel accounts for between 35% and 70% of a ship’s total operational cost (Appendix 1), depending on the type of ship and service. Sulphur content in fuel 1% 0.5% 0.1% Container freight rate 4-13% 8-18% 44-55% Table 1.2; Increase in container freight rates as a consequence of the low sulphur fuel regulations. (Kalli, Karvonen, & Makkonen, 2009) Shipping companies are required to recover the increase in cost to maintain their level of service, meaning the price of shipping on sea have to be increased. However, recovery of fuel cost from cargo customers are challenging when vessel capacity utilization is not 100%, and trade is not evenly balanced. Cargo owners might seek new ways of transportations if the price increase outweighs the advantages of sea transport. They might accept a small change in freight cost, however, shipping in the Baltic Sea face tough competition from other modes of transport, such as rail and truck. 1.1.4 NEW REGULATIONS EFFECTS ON INDUSTRIES The expected change in price of freight cause of new stringent environmental regulations will affect the industries in proximity of the emission control areas. Each industry will be affected differently, as there is a difference in import/export and need of sea transport. Estimations for the Finish industries show, that especially forest-, metal-, and chemistry industry will face significant increase in cost, in many cases with as much as 14-30%. (Kalli, Karvonen, & Makkonen, 2009) Similar calculations have been made for the Swedish industries, which show on similar result as in Finland. (Sjöfartsverket, 2009) The industries will face a tough challenge, as they need to increase their price in order to recover the increase in shipping cost. They are competing on a global market and an increase of price might not be possible, which will lead to smaller marginal. A possible scenario could be movement of industrial production out from the ECA, e.g. Baltic Sea and closer to the market, which will be devastating for the economies around the ECA areas. 0 200 400 600 800 1000 1200 2010 - Q1 2010 - Q2 2010 - Q3 2010 - Q4 2011 - Q1 2011 - Q2 2011 - Q3 2011 - Q4 U SD / to n Year - Quater HFO 1% MGO 0,1% 4 1.2 MAXIMISE EFFICIENCY FROM ASSETS The reduction of fuel consumption trough optimisation of the fleet, use of alternative fuels, or through new technology is a high priority in order to stay competitive in the market even with new regulations. Optimisation in efficiency of existing ships and fleet could help in cutting cost of the operation. Different areas of operation could be improved, for example, voyage planning, weather routing, just in time, ship handling, fleet management, fuel type, etc. The yield of individual measures may be small, but the collective effect across the entire fleet can be significant. In order to identify and setting targets of improvement within the shipping company and in comparison with others in terms of fuel efficiency, benchmarking is utterly important. The benchmarks must be identifiable and easy to access in order to simplify the analysis and realisation of applicable measures. The process of constant improvement of fuel efficiency is vital in order to stay competitive in a market with high competition and increasing costs as a result of higher bunker prices. 1.3 PURPOSE The purpose with the study is to propose benchmarks and measures for better operational efficiency from analysis of AIS data, with main focus on fuel consumption to achieve ship and fleet efficiency. 1.3.1 RESEARCH QUESTIONS The issue with the study is to examine how AIS data can be used to compare ships against each other with appropriate benchmarks in order to identify measures for better fuel efficiency. The study will answer two main research questions: • Is it possible to benchmark ship fuel efficiency from AIS data? • What benchmarks are useful in comparison with other ships? These questions are further divided into sub-questions; Reliability of AIS data? Method of AIS data analysis? What are appropriate benchmarks in terms of fuel efficiency in different operational modes (i.e. sea, port, and anchor)? What measures in fuel efficiency improvement could be found from the benchmarks? 1.4 DELIMITATION The study will limit the segment of study to two sister groups of small general cargo ships of 7 700 dwt and 12 700 dwt, with 22 ships in each group. AIS data from 2010 and 2011 is used in the compilation of operational profiles, where ships with less than 100 days of coherent data has been disregarded in further analysis. Ship sailed distance and fuel consumption are theoretical calculations based on AIS data, result could therefore differ from on-board ship-log recordings. The grouping of ships in tramp/liner traffic is based on analysis of the operational pattern with no confirmation from METHODOLOGY 5 the operator; reality could show on a different type of traffic or a mix of tramp/liner traffic. Further is the same engine configuration, resistance coefficient, and cargo capacity assumed for each group of sister ships. 1.5 OUTLINE OF THE REPORT • Introduction The study is introduced by giving a background to the problem. • Theory The chapter describes the cost structure of shipping and different measures that can be applied in shipping operation with main focus on efficient energy consumption. • Methodology The chapter describes how the study was carried out, and discusses quantitative and qualitative research, case studies, data collection, and reliability – validity – objectivity. • Case study This chapter describes input data to case study and the compilation of operational profiles for 44 general cargo ships from AIS data. • Result The results of the case study are presented in this chapter. Operational profiles for all ships in study are presented. • Analyse This chapter presents an in-depth analysis and discussion of the operational profiles. Benchmarks of the ships in study are set for each operational mode and measures in terms of fuel efficiency improvement are discussed. • Conclusion Findings from result and analyse are summarized. • Future study Suggestions of future areas of study are given in this chapter. 6 2 THEORY The chapter describes the cost structure of shipping and different measures that can be applied in shipping operation with main focus on efficient energy consumption. 2.1 COST STRUCTURE The cost of shipping is the main key in the decision process of shipping operations. Shipping cost is built up of voyage cost and operating cost. Voyage cost is a variable cost that comes with a particular voyage. Operating cost is a cost that originates from a ship operation. (Stopford, 2005) • Shipping cost (SC): o Voyage cost (VC)  Fuel cost (FC)  Port dues and service charges (PS)  Canal dues (CD) o Operating cost (OC)  Manning cost (M)  Insurance cost (IN)  Repair and maintenance cost (RM)  Store and lubricant cost (SL)  Administration cost (AD) SC = VC (FC+PS+CD) + OC (M+IN+RM+SL+AD) (Stopford, 2005) Voyage cost is made up of fuel cost, port dues and service charges, canal dues. Operating cost is made up of manning cost, insurance cost, repair and maintenance cost, store and lubricant cost, administration cost. Some costs of manning, administration, and store and lubricant can be shared within a shipping company in order to achieve economies of scale. (Venus Lun & Browne , 2009) 2.2 SECTORS OF SHIPPING Shipping is a highly international business; companies are privately owned and offer a service of transportation within, or between regions. The shipping industry is mainly divided in three different sectors, liner, tramp, and industrial, divided by the type of service and characteristics of the transported cargo. (Stopford, 2005) 2.2.1 LINER SHIPPING Liner shipping offers a regular service between ports, operated just like a bus service according to a fixed schedule. Cargoes are accepted under a bill-of-lading contract issued by the ship operator to the cargo owner. The cargoes are most often smaller quantities with a high value per tonne that does not by itself fill an entire shipload. THEORY 7 The combination of many small consignments and a regular service put a lot of demand on the administration, which leads to high overhead cost in liner shipping compared to other sectors of shipping. Liner operators are therefore vulnerable to price cutting strategies by other companies operating at the same route. Competition in liner service has generally been regulated by conferences, i.e. agreements between the shipping companies. These are agreed upon in order to stabilize conditions of competition and to set fright rates for all members in the conference. (Stopford, 2005) 2.2.2 TRAMP SHIPPING Tramp shipping, also commonly referred to as bulk shipping is characterised by shipping of cargo with a low value per ton, often a whole shipload from a single shipper. Tramp service is unlike the liner service not running on a fixed schedule. The cargoes are referred to as spot cargoes where a contract arranged between the ship-owner and shipper, either for a single voyage, i.e. voyage charter or for a period of time, i.e. time charter. (Stopford, 2005) 2.2.3 INDUSTRIAL SHIPPING When the shipper, i.e. cargo owner is confident of the amount of cargo he need to transport in the future, he might take the role as shipping operator himself. Industrial shipping is most often carried out by large cooperation’s that transport own goods or raw material essential to their manufacture and distribution supply chain. Industrial shipping has however decreased during the recent years in favour of tramp shipping. Companies have shifted towards a stronger focus on their core business, rather than also being a shipping operator. This has led to increase of tramp market, where operators now have more cargo with a constant flow to choose from. (Stopford, 2005) 2.2.4 CHARTER AGREEMENTS Voyage charter In a voyage charter, the ship-owner agrees to transport a specific cargo between two ports. The charterer pays the ship-owner per tonne or a lump-sum. The ship-owner pay all cost involved in the transport, excluding stevedoring in port. (Stopford, 2005) Time charter In a time charter contract, the ship-owner hands over the operational control to the charterer during a specified time period. The ship-owner still pays the operational cost of the ship, however, the charterer pays the voyage specific costs. (Stopford, 2005) Bareboat charter In a bareboat charter, the ship full operational control is handed over to the charterer. The owner is usually an investment company who has no knowledge about ship operations, i.e. the ship is only an investment. The charterer pays both the operating and voyage specific costs. (Stopford, 2005) 2.3 ENERGY EFFICIENCY Fuel cost is the single most important variable cost. It accounts for approx. 35-70% of the total cost of running a ship (Appendix 1). Building fuel-efficient ship has become more and 8 more important in the shipping industry as oil prices have increased significant over the last decades. By improving the energy efficiency of a ship, more work can be done with the same energy consumption. Improvement in efficiency can be achieved in both the design and in operation of the ship. Some of the improvements can be retrofitted to existing ships, however the most important factors that are determine the energy efficiency of a ship are closely linked to the specification of the ship and more easily changed in the design and building process of a ship. This means that design improvements in efficiency take some time before they will be implemented and have any affect in the efficiency status of a shipping company, as most ships have a service life of approx. 30 years before they are phased out. The lifetime of ship can change over its lifetime, so can the intended market for the ship, which is important to take into account already in the design process. Larger ships are usually more energy efficient per tonne-km than smaller ships. However, smaller ships can usually achieve a higher utilization, which may result in a better energy efficiency. (IMO, 2009) It is important to use the right ships in the shipping network. With larger ships, energy efficiency can improve, however when looking on the whole chain, door-to-door, energy efficiency can be improved if smaller ships, i.e. feeder ships, connect with the larger ship in the spread of distribution, as illustrated in Figure 3.1. This because larger ships become more energy inefficient if they have to sail with low capacity utilization. (IMO, 2009) 2.4 FUEL CONSUMPTION The hull resistance through water and type of engine used to propel the ship forward mainly determine the fuel consumption of a ship. A formula for estimating the power consumption is showed below. (IHS Fairplay, 2012) Pactive = 20% × PAux.Gen + 85% × PMain × ( Vcalc 94% × Vservice )2,5 Pactive = Total power. (kW) PAux.Gen = Auxiliary generator power. (kW) PMain = Main engine power. (kW) Vcalc = Calculated average speed. (Knots) Vservice = Ship service speed. (Knots) The Pactive consumption formula is used when the ship is under way sailing, i.e. a speed ≥ 0,2 knots. If the ship has a calculated speed < 0,2 knots, i.e. at anchor or at berth in port, a Pinactive formula is used. Pinactive = PAux.Eng × ActivityShareVesselType Main port Figure 3.1; Smaller feeder ships connect with large ship in order to achieve energy efficiency in the whole shipping network. THEORY 9 Pinactive = Total power. (kW) PAux.Eng = Auxiliary engine power. (kW) ActivityShareVesselType = % Of Auxiliary engine use in berth or at anchor. The activity share of the auxiliary engine of a general cargo ship is set to 25%, both at berth and at anchor. The auxiliary power used while at sea, i.e. speed ≥ 0,2 knots, is set to 0. The auxiliary power needed while at sea is calculated as an additional load on the main engine. Each ship has a value of fuel consumption in grams per kWh of energy generated by its engine. This value is multiplied with Pactive to produce total fuel consumption in grams. (IHS Fairplay, 2012) 2.5 ENERGY EFFICIENCY MEASURES Energy efficiency could be improved by implementing different measures in the operation of a ship. Measures could be carried out both on-board the ship by the crew and by onshore personnel at the shipping company. 2.5.1 OPTIMISATIONS BY THE MASTER. The master of a ship is in a position where he can optimise the voyage in a way to run efficient, within the limitation from constrains that are set in contractual agreements and scheduling. Except from technical side with ballast and trim optimisation he can also adjust the ship route according to the weather and currents i.e. weather routing and just in time, where he take tide, queues and arrival window into consideration. (IMO, 2009) The efficiency potential of voyage optimisation measures is very hard to access from a general basis. Each ship and route has its own characteristics of operation, it is therefore important to look into the individual operational procedures in order to define areas of improvement and potential of increased fuel efficiency. (IMO, 2009) 2.5.2 WEATHER ROUTING In order to optimise the voyage in terms of fuel consumption, safety, comfort, and minimum time under way, the weather routing systems suggest an optimum track for the intended voyage based on the weather forecast, condition of sea, and the design and specifications of the ship. The master can use the pilot chart atlases, the sailing directions, and historical weather data tables to make a preliminary weather routing. A weather routing agency can also assist the master by suggesting an optimal route in compliance with the weather forecast. The agency can then monitor the ship and suggest changes in the route as voyage progress. (Bowditch, 202) Weather routing is preferable especially if the passage is: • Long passage, about 1 500 nm or more. • There is more than one choice of route. The waters are navigational unrestricted. • Weather (wind, waves, current) is an important factor in the choice of route. Studies show that efficient weather routing can save approximately 2-4% in fuel consumption. (MARINTEK, 2000) However, weather routing is today a common practice in ship voyage planning, significant increase in fuel consumption savings is therefor hard to achieve. 10 2.5.3 JUST IN TIME Shipping has a tradition in operating at a high speed during the sea leg of the voyage, with waiting outside port as a consequence. Arrival just in time when berth is ready has a huge potential in fuel cost savings for the shipping company. However, contractual agreements sometimes favour insufficient operations. The laytime start as soon as the ship owner sent notice of readiness. The laytime specifies time needed for loading/unloading. The ship owner is entitled to compensation from the charterer i.e. demurrage, if the loading/unloading takes more time than specified in the laytime clause, as illustrated in Figure 3.2. The owner should compensate the charterer, normally half of the demurrage rate, i.e. despatch, if the loading/unloading operation is completed in advance of the time specified in the laytime clause, as illustrated in Figure 3.3. Figure 3.3; Despatch. Loading/Unloading is finished before laytime ends. The ship-owner wants to send notice of readiness as soon possible within the time frame specified in the contractual agreement, as the demurrage rate could affect his financial result of the contract in a positive way. Owner’s motivation of just in time delivery of the ship when berth is ready might be reduced cause of the demurrage possibilities, which could affect the choice of speed and fuel consumption in a negative way as he might choose to operate in a higher speed than what would be necessary for a just in time delivery. Studies show that just in time arrivals could save approximately 1-5% in fuel consumption. A higher saving may be achieved if the contractual agreement not favours the operator to operate the ship at a higher speed. (IMO, 2009) Laytime Sea voyage Loading/Unloading Notice of readiness Despatch Time Sea Laytime Sea voyage Loading/Unloading Notice of readiness Demurrage Time Sea Figure 3.2; Demurrage. Loading/Unloading exceeds the laytime. THEORY 11 2.5.4 TRIM OPTIMISATION The energy savings by trim optimisation, i.e. optimal position in the water, is very much depending on the type of ship and nature of operation. An optimal position can be translated into fuel consumption savings, as the optimal trim will reduce the resistance through water. Lockley and Jarabo-Martin (2011) indicate in their report that the potential savings of trim and ballast optimisation could reduce the fuel consumption with approximate 4% compared with ship operation at level trim. However, the study of IMO (2009) indicate significant lower savings, 0-1%, as trim optimisation already is a common practice. 2.6 CHOICE OF SPEED The optimal speed from an economical point of view is defined by MARINTEK, 2000: “The speed that maximizes the difference in between income and expenses (per time unit) of the ship”. However, the optimal speed changes from which view of the different actors in the sea transportation. As Ronen (1982) point out in his research, the owner of the cargo sees the transportation cost in relation to value of the cargo. A cargo with high value could be too expensive to slow steam as the savings in fuel cost might be diminished by the extra cost of an increased lead-time. The ship owner has to weigh his income and cost against contractual agreements, which could vary from time to time. When the supply of sea transport exceeds the demand, a speed reduction may be the best choice for the ship owner. If the demand of sea transport is higher than the supply, a minimum time strategy is normally chosen. (MARINTEK, 2000) 2.6.1 SLOW STEAMING Slow steam became a common practice in shipping during the financial crisis in 2008-2009. The demand of shipping fell rapidly at the same time new capacity was delivered due to previous orders made during the financial boom in the years leading up to the financial crisis. Shipping companies started to use slow steaming as a way to reduce cost and to be able to utilize the fleet in a wider extent than the demand. The practice was supposed to fade out when the economy started to grow again and the demand of shipping rose. However, increase in fuel price and more stringent environmental regulations have led to slow steaming as a normal practice in order to adapt to new market conditions, which is showed in the research of Cariou (2011), where he points out that the concept of slow steaming has reduced fuel consumption with approximate 11% in major container trades worldwide during 2008-2010. With slow steaming you run the ship with approx. 80% of the main engine full power. Which reduce fuel consumption with approx. 40% (Appendix 1). Timing of fuel injection, adjusting exhaust valves, and exchanging other mechanical components in the engine is vital in order to make sure maintenance cost does not overrun fuel cost savings. 2.6.2 ECONOMIC SPEED Running the ship at slower speed means significant lower fuel consumption with lower cost as a result. The economic speed of a ship is achieved at the speed that result in the best possible 12 financial result for the shipping company. Several factors are considered when determining the economic speed of a ship. • Price of bunker. • Relation of fuel consumption and speed. • Daily operating cost. • Operating profit. • Future employment of the ship. • The state of the freight market. • Design speed of the ship. • Technical ability for the engine to operate at a lower speed. • Weather conditions. An easy way to determine the economic speed of a certain ship is to calculate the total voyage costs in relation to speed, as illustrated in Figure 3.4. (Dykstra, 2005) Where the operating cost for each ship category is set as an fixed cost per day, i.e. daily charter rate, and the bunker price a variable, which is the same for all ships (Table 3.1). 7 700 dwt 12 700 dwt Daily charter rate 4066 USD 5403 USD Fuel: HFO 1% sulphur 570 USD / tonne MGO 0.1% sulphur 794 USD / ton Table 3.1; Daily charter rate and fuel price. The economic speed calculations are based on the average charter rates (VHSS, 2012) and fuel prices in Rotterdam (IHS Fairplay, 2012) during the period of study, i.e. 2010-2011. There was a significant steady increase in fuel price over the period, with an increase of 51% of the MGO and 51% of HFO. Charter rates were distributed between 2634 USD/day and 5151 USD/day, respectively 3338 USD/day and 7348 USD/day for the larger ships. The peak period of charter rates occurred in the first quarter of 2011. Figure 3.4; Economic speed. Thick line show cost per nautical mile with use of normal bunker fuel (HFO 1%). Dotted lines show cost per nautical mile with low sulphur fuel (MGO 0,1%). 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Co st / nm (U SD ) Speed (Knots) 7700 dwt 12700 dwt THEORY 13 Where the curve of cost per nautical mile reaches its minimum in Figure 3.4 determine the most economic speed, i.e. lowest cost per nautical mile, for the two ship categories examined in this report. (Dykstra, 2005) 7 700 dwt 12 700 dwt HFO 1% 8.9 knots 9.7 knots MGO 0,1% 7.8 knots 8.7 knots Table 3.2; Best economical speed at different fuel prices. As Figure 3.4 and Table 3.2 show, the best economical speed decreases when the fuel price increases. However, when the charter hire increase, the optimal economical speed increase as the economical speed is a function of the relation between fuel cost and total cost. With higher speed, the fuel cost percentage of the total cost increase. At a speed of 10 knots, the fuel consumption stands for approximately 50% of the total cost. The economic speed calculation does not take into consideration of other variables than charter hire, bunker price, and fuel consumption. Type of trade, state of market, weather, and technical aspects could also affect the economic speed. The model is also only looking at the economic benefit of the charterer. The ship owner’s benefit of a reduction in speed stretches beyond the economical speed of the charterer. The economical speed seen from both the ship owner and the charterer’s point of view in a time charter agreement can be expressed as the stakeholder shared benefit. In this economic model the economic benefit is shared equal between the two main stakeholders, i.e. ship owner and charterer. (Klanac, Nikolic, Kovac, & McGregor, 2010) BSO(V) = BCH(V) BSO(V) = NSS(V) × (ACR - CO) – (ACR - CO) BCH(V) = NSS(V) × (-ACR - CFO(V)) – (-ACR – C′FO) Where: NSS(V) = Number of ship necessary to transport equal amount of cargo. ( VService Vcalc ) ACR = Annual charter rate. ( 365 × t × Daily charter rate ), (t = commercial use of ship per annum). CFO(V) = Fuel cost. C′FO = Fuel cost at service speed. CO = Ship-owner operating cost per ship. 7 700 dwt 12 700 dwt Daily charter rate (VHSS, 2012) 4066 USD 5403 USD Operating cost/year 1200000 USD 1500000 USD t 90% Fuel price/tonne HFO 1% 570 USD Fuel price/tonne MGO 0.1% 794 USD Table 3.3; Costs associated with ship operation. 14 A shared stakeholder benefit from a speed reduction from full service speed is illustrated in Figure 3.5. Benefit can be achieved by between 5.1-14 knots for the 7 700 dwt ships, and between 5.4-15 knots for the 12 700 dwt ships. The maximum shared benefit of approx. 363000 USD annually is achieved at 8.6 knots for the 7 7000 dwt ships, and approx. 512000 USD at 9.2 knots of speed for the 12 700 dwt ships. These numbers are only achieved given prerequisites in Table 3.3. Figure 3.5; Shared stakeholder benefit of a speed reduction from service speed. An even greater benefit is possible if the ship is running on a low sulphur fuel, i.e. higher fuel cost. The optimal speed for the two ship sizes are approximately 1 knots lower with the use of MGO 0.1% sulphur fuel compared with running on HFO 1% sulphur fuel, as the dotted line in Figure 3.5 illustrate. -500000 0 500000 1000000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sh ar ed b en ef it (U SD ) Speed (Knots) 7700 dwt 12700 dwt METHODOLOGY 15 3 METHODOLOGY The following chapter describes how the study was carried out, and discusses quantitative and qualitative research, case studies, data collection, and reliability – validity – objectivity. Methodology is a fundamentally approach where framework and different principles are being set up in order to show how the work should proceed. (Höst, Runesson, & Regnell, 2006) 3.1 QUANTITATIVE AND QUALITATIVE RESEARCH In quantitative research, measurements are made in the data collection process and processed further with statistical methods. The focus in qualitative research is the soft data, e.g. qualitative interviews and interpreted data. (Patel & Davidsson, 2003) The method of study is chosen in order to gather information needed to carry out the research. (Bell, 1995) A research can consist of a mix of both quantitative and qualitative. The formulation of the problem decides witch research approach to be used. (Patel & Davidsson, 2003) 3.2 CASE STUDIES Bell (1995) means that case studies are especially suited for single researchers, as case studies give the researcher the opportunity to carry out in depth analysis of a problem in a limited time. Further describes Silverman (2005) what a case study is: ”The basic idea is that one case (or perhaps a small number of cases) will be studied in detail, using whatever methods seem appropriate. While there may be a variety of specific purposes and research questions, the general objective is to develop as full an understanding of that case as possible” Case studies are often used in the study of processes and changes, where it is common that different kind of information is gathered in order to give an as detail picture as possible of the case. Most common is the use of interviews, surveys, and observations in the case study to gather information. (Patel & Davidsson, 2003) 3.3 DATA COLLECTION There are two types of data, primary- and secondary data. The difference between them lies in which purpose data is gathered. Primary data is gathered in the purpose of research and need, while secondary data is gathered in the need of another purpose. (Eriksson & Paul, 2001) Data collection can be carried out with several different techniques depending on the problem. Interviews, literature, observations, and surveys are examples of methods. The method used, is chosen depending on the purpose with the information. Literature: 16 Any kind of materials such as books, brochures, and scientific papers are seen as literature. This material is seen as secondary data, and is used as a theoretical- and analytical frame for the research. (Bell, 1995) Data collected by others: Processed material, available statics, index data, and archival data are four different kinds of data collected by others. Höst (2006) means that this data has to be carefully used as it was collected in other purposes than what the study refers to. It is therefore important to have a critical approach to this kind of material to maintain high validity and reliability. 3.4 RELIABILITY, VALIDITY, AND OBJECTIVITY It is important that the observations made in the study can be repeated in order to attain a high level of reliability. The reliability is depending on the credibility of the measurement instruments. It question if another researcher get the same results with the same measurement instruments. The method should therefore be independent of the researcher in order to achieve high reliability. (Eriksson & Paul, 2001) Further mean Eriksson & Paul (2001) that the validity is the chosen measure instrument ability to measure what it is intended to measure, and that god validity is attained when the researcher measure what he is intended to measure. With a low reliability, validity gets low. Good reliability is necessary, but not enough to secure high validity. It is therefore possible to have high reliability with low validity. The credibility of a study is also depending on the researcher objectivity. Objectivity means the degree of which different values affect the result. The objectivity can be increased if the research clearly describe the study and give the readers an opportunity to create their own view of the result. (Björklund & Paulsson, 2003) 3.5 RESEARCH APPROACH The aim with this study is to propose measures and benchmarks in operation in order to achieve energy efficient shipping. A case study of 44 general cargo ships was carried out. These were two sister groups of 7 700 dwt and 12 700 dwt, with 22 ships in each group. Each group of sister ships were selected from their design and configuration in order to eliminate any design configuration differences that could affect the result of the operational analyse. Both primary and secondary data have been used for the study. The primary data for the case study and analyse was extracted from the IHS Fairplay database, AISLive and Sea-Web. Further were secondary data from literature gathered in order to give the reader a better understanding of the importance of energy efficiency in shipping and to help in the analyse of the case study. Others collected statistical data, such as bunker prices at major ports in year 2010-2011. The reliability of the study can be considered as high, since the selection of ships were carried out carefully in order to minimize design differences in fuel consumption. The period of 24 months (year 2010, and 2011), also minimize the risk of seasonal differences in analysis. However, the author had no insight in the operator/shipping company strategy of the decision behind the choice of speed and other operational characteristics, several strategy factors could affect the result. METHODOLOGY 17 Since the author has no relation to studied ships/companies, the objectivity of the study has been maintained at a high level. 18 4 CASE STUDY This chapter describes input data to case study and the compilation of operational profiles for 44 general cargo ships from AIS data. 4.1 AUTOMATIC IDENTIFICATION SYSTEM (AIS) The AIS system is a ship tracking system that is built up from transponders on ships. A ship carries an electronic device, which transmits and receive single to and from other ships within a certain range. The electronic device consists of a GPS receiver, a computer, and a radio. The GPS send information about the ship position to the computer, which process the data together with other data from the ship and the send this information to other Ships and shore stations equipped with AIS equipment, as illustrated in Figure 4.1. Figure 4.1; AIS system in practice. The regulation 19 of SOLAS Chapter V from the IMO set the requirements of the navigational equipment to be carried on-board ships. In 2000, IMO adopted a new requirement, which became effective in 31 December 2004, that all ships over 300 gross tonnage engaged on international trade, cargo ships over 500 gross tonnage engaged in domestic trade, and all passenger ships should carry AIS equipment on-board. (IMO, 2012) The purpose with AIS is to improve the safety and efficiency on sea. The AIS system makes it easier to identify other ships and leaves additional information to the users, which increase the awareness in different situations. The quality of decision-making could be improved for both for the shore-based surveillance activities as well as for the on-board personnel with the use of AIS information together with other navigational systems e.g. radar. The AIS transmits three types of information from the ship: Static information, which is entered into the system in the installation process and is only updated if vessel change name or if the particulars are changed due to reconstruction. Dynamic information is updated with 2-10sek intervals from the ship sensors connected to the AIS. Voyage specific information, is manually entered by the crew by every voyage or change in operation. ID, position, heading, speed, t Ship send: ID, position, heading, speed, etc. Ship receives: Other ships, ports, warnings, etc. AIS station CASE STUDY 19 Static information • MMSI (Maritime Mobile Service Identity) • Name of ship • IMO number • Length and breadth • Type of Ship • Position of the AIS antenna Dynamic information • Position of the ship (Latitude and Longitude) and GPS accuracy. • Time (UTC) • Course over ground (COG) • Speed over ground (SOG) • Heading • Rate of turn Voyage specific information • Navigational status (On way using engine, At anchor, Not under command, Restricted ability to manoeuvre, Moored, Restricted by draught, On ground, Engaged in fishing, Under sail) • Draught • Dangerous goods • Destination and estimated time of arrival (ETA) • Intended route (way-points) • Short safety related message (Sjöfartsverket, 2004) Not all vessels on sea are equipped with AIS equipment; small leisure boats, fishing boats, and shore-based station could lack AIS equipment. The crew of the ship can also turn of the equipment, and the equipment could be inaccurate calibrated. It is therefor important to remember that the AIS system is only a supplement to other navigational information and might not show the whole picture in a situation. The AIS information is only as good as the accuracy in the broadcasted information. (Sjöfartsverket, 2004) 4.1.1 AISLIVE AISLive is a global AIS network set up by IHS Fairplay to track ship movements in real-time trough an online application. Positions of ships of over 54 000 ships are updated every third minute. The AISLive network of land and satellite antennas covers over 2 500 ports and 100 countries. IHS Fairplay has stored the AIS data once every hour since 2004. This historical data could for example be used to analyse a ship movement pattern and time in particular regions. Example of AIS data from the IHS Fairplay database: 20 Table 4.1; Example of AIS data. 4.2 SHIPS IN STUDY 44 general cargo ships were selected for the case study (Appendix 2). These were two sister groups, one group of 22 ships with a dwt of approx. 7 700 and a second group of 22 ships with a dwt of approx. 12 700. Within each group of sisters, every ship has the same characteristics i.e. same length, breadth and shape of hull as showed in Table 4.2. This allow comparison of fuel consumption within each sister group, as the resistance and drag coefficients are the same or very close to the same. 7 700 dwt 12 700 dwt Length 118,55 meters 138,5 meters Breadth 15,2 meters 18,1 meters Draught 6,3 meters 8 meters TEU 14 300 550 Mcr 3840 kW 5400 kW Aux. generator 918 kW 2200 kW Aux. engine 550 kW 1530 kW Service speed 14 knots 15 knots Fuel consumption 177 gram / kWh 175 gram / kWh Fuel consumption Aux. engine 210 gram / kWh 210 gram / kWh Table 4.2; Ships data. 4.2.1 FUEL CONSUMPTION The fuel consumption of the two groups of ships is shown in Table 4.3 at service speed, at berth, and at anchor. Further is fuel consumption illustrated as a function of speed in Figure 4.2. 7 700 dwt 12 700 dwt Fuel consumption at Vservice 17 ton/day 24,4 ton/day Fuel consumption at berth/anchor 0,693 ton/day 1,978 ton/day Table 4.3; Fuel consumption at service speed and at berth/anchor. CASE STUDY 21 Figure 4.2; Fuel consumption in ton/day at different speeds. 4.2.2 HYDROSTATIC DATA Hydrostatic data of the two ship sizes help to determine the amount of cargo on-board at the time of a given AIS recording with draught data. The deadweight is a measure of how much a ship can carry. It sums up the weight of cargo, fuel, fresh water, ballast water, provisions, passengers, and crew. 7 700 dwt 12 700 dwt Draught tonne / cm Deadweight Draught tonne /cm Deadweight 3,0 14,36 0 No data available 3,5 14,53 773 4,0 14,74 1456 4,5 14,96 2199 5,0 15,23 2955 5,5 15,56 3727 6,0 15,88 4516 6,5 16,17 5316 7,0 16,39 6132 Table 4.4; Tonne of deadweight to change draught. (Bodewes Shipyards bv, 2012) 4.3 AIS DATA AIS data were extracted from the IHS Fairplay database for all ships during the period 2010- 01-01 to 2011-12-31. The tracking of each ship was plotted onto a map in order to determine area of operation and in order to visualize the correctness of the AIS data. 0 5 10 15 20 25 1 2 3 4 5 6 7 8 9 101112131415 To n/ da y Knots 7700 dwt 12700 dwt 22 Figure 4.3; Example of ship’s visualization of track on map. The AIS data included some faulty recordings of data due to disturbance in transmission. The visualization of the tracks showed on faulty recording where coordinates have been misplaced in the data. These were manually corrected in the data and track was controlled in visualization. AIS recordings with duration from previous recording of less than 10 minutes were deleted, as they generally were displaced coordinates. The AIS data included SOG (speed over ground), however this data is automatically transferred from the Ship GPS to the ship AIS equipment at the time when AIS data is transferred. It does not show on the actual SOG since the last update. Actual average speed since pervious update was manually calculated based on the distance between coordinates and duration from previous recording in order to be able to correctly calculate the fuel consumption of the ship. The AIS data were also missing two weeks, week 15 in 2010 and week 4 in 2011. Duration and distance after each of these weeks were set to 0 in order to minimize the risk of faulty data in analysis. The spherical law of cosines is used to calculate the distance between two coordinates, as planet earth is spherical. Distance = ACOS(SIN(lat1)*SIN(lat2)+COS(lat1)*COS(lat2)*COS(lon2-lon1))*6371 However, Microsoft access, which was used in this study for data analysis do not recognize arc cosines. The formula was therefor modified for the use in Microsoft access: Visualization Find faulty recordings Correct Figure 4.4; Working process of AIS data correction. CASE STUDY 23 Distance = Atn(Sqr(1-(Sin([latt1])*Sin([latt2])+Cos([latt1])*Cos([latt2])*Cos([long2]- [long1]))^2)/(Sin([latt1])*Sin([latt2])+Cos([latt1])*Cos([latt2])*Cos([long2]- [long1])))*6371/1.852) Table 4.5; Extract from AIS data with distance and speed manually calculated. The calculated distance and average speed was added to the data as illustrated in Table 4.5. The ship operational status at each AIS recording was corrected with updated data given from IHS Fairplay, where operational status was divided into three different statuses; “Under way using engine”, “Moored”, “Anchored”. Ports were added to the AIS data by comparison of the zone id with IHS Fairplay database of ports. In the cases where operational status was set “Moored” and no zone id was given, port name was set to “Unknown-“name of region”. Further was transportation work for each AIS recording calculated. Transportation work is a measure of how much cargo the ship have transported, measured in tonne-kilometre. The transportation work was calculated with the use of hydrostatic data, see Table 4.4, where the corresponding deadweight of AIS recording draught was multiplied with distance travelled since previous AIS recording. A list of voyages for all ships was created. Each voyages started when operational status changed to “Moored” from either “Under way using engine” or “Anchored”, as illustrated in Figure 4.6. Only voyages with coherent AIS data were used in analysis. Voyages including AIS recordings with duration from previous recordings of more than 48 hours were deleted. Figure 4.5; Definition of voyage in study. The list of voyages summarizes for each voyage, the time, distance, freight transport, and fuel consumption at sea, at anchor, and in port. The draught was also assed from the AIS data, which indicated if the voyage included a loading or unloading procedure and time spent in different draught modes. 4.3.1 CONFIDENCE AND UNCERTAINTY OF DATA Input Source Confidence Comment Ships particulars data Fairplay database High Ships machinery data Fairplay database, ship-owner websites Moderate /High Some minor individual differences might be found. Sea Port Sea Anchor Port Voyage 24 All ships are however assumed to have the same configuration in the calculations. Ship service speed Fairplay database, ship-owner websites Moderate Same as above Distance, Speed AISLive Moderate Distance is calculated as a straight line between AIS recordings. Accuracy could be affected if there is a landmass between the AIS recordings, i.e. shortest way. Randomly check of distances show on a low inaccuracy during the two-year period. Ship status AISLive, analysis Moderate / High The ship status, i.e. “Under way”, “Moored”, “Anchored” where corrected with updated data from IHS Fairplay. Draught data AISLive Low/Moderate The crew of the ship updates draught manually. Ships in study showed on good regularity of updates. However, the recordings sometimes indicated a delay in draught update after a port call. Table 4.6; Confidence and uncertainty of AIS data. RESULT 25 5 RESULT The results of the case study are presented in this chapter. Operational profiles for all ships in study are presented. 5.1 OPERATIONAL PROFILES Operators are listed in Table 5.1. Each operator was given a corresponding number in order to simplify further analysis. In total, there are 18 different operators, 10 operators for the 7 700 dwt ships and 8 operators for the 12 700 dwt ships. 7 700 dwt 12 700 dwt Nr Operator Nr Operator 1 Feederlines BV 11 BBC Chartering & Logistic GmbH 2 Universal Africa Lines NV 12 OXL NV 3 Flinter Shipping BV 13 Clipper Projects A/S 4 Harren & Partner Ship Mgmt 14 SE Shipping Lines Pte Ltd 5 Hermann Buss GmbH & Cie KG 15 Marlow Navigation Co Ltd 6 Navesco SA 16 BD-Shisnavo GmbH & Co 7 Onego Shipping & Chartering BV 17 Jutha Phakakrong Shipping 8 Scan-Trans Chartering KS 18 Nordana Lina A/S 9 Strahlmann E Reederei eK 10 Transatlantic Rederi AB Table 5.1; List of operators. Operational profiles compiled from AIS data 2010-2011 of the ships in study are listed in Table 5.2, where each ship has been given an individual number in order to simply further analysis. The column, Time days, show total time of ship in analysis, the reason this figure is significant lower than 365*2=730 can be derived from the cleaning of AIS recordings. All voyages where there was a gap somewhere in the AIS recordings with over 48 hours were deleted, in order to increase the level of correctness of analysis. 7 700 dwt Nr NAME Time days % sea % moored % anchored Avg Speed Fuel tonn e/h Avg Voy. Sea (days) Op. Nr. Liner /Tramp 1 VARNADIEP 184 57% 32% 12% 10.47 0.39 3.7 1 Liner 2 VRIESENDIEP 172 53% 38% 10% 10.39 0.40 3.7 1 Liner 3 VOSSDIEP 219 69% 24% 6% 10.12 0.36 4.7 1 Liner 4 VEERSEDIEP 120 51% 31% 17% 9.60 0.35 3.5 1 Liner 5 VELSERDIEP 129 56% 37% 7% 10.87 0.43 3.9 1 Liner 6 VIKINGDIEP 262 58% 32% 11% 10.81 0.43 4.4 1 Liner 7 VECHTDIEP 154 46% 39% 15% 9.95 0.36 3.5 1 Liner 26 8 VLIEDIEP 213 45% 45% 9% 9.80 0.34 3.8 1 Liner 9 NORDLAND 413 62% 37% 1% 11.59 0.51 2.4 1 Liner 10 WISAFOREST 448 63% 36% 1% 11.55 0.50 2.5 1 Liner 11 VASADIEP 210 55% 38% 6% 9.76 0.35 3.2 1 Liner 12 UAL CYPRUS 226 65% 25% 10% 11.36 0.47 4.6 2 Liner 13 FLINTERLAND 307 59% 24% 17% 10.50 0.42 6.8 3 Tramp 14 PAZ COLOMBIA 101 51% 34% 16% 10.03 0.36 3.4 4 Tramp 15 HERMANN SCAN 186 66% 19% 15% 11.12 0.43 6.0 5 Tramp 16 PENSILVANIA 135 49% 28% 23% 9.61 0.36 4.1 6 Tramp 17 VLISTDIEP 136 48% 44% 8% 11.41 0.47 2.7 7 Tramp 18 HARTWIG SCAN 232 65% 22% 13% 10.67 0.43 6.4 8 Tramp 19 HANSEN SCAN 268 59% 22% 19% 10.63 0.39 6.2 8 Tramp 20 LIFTER 257 60% 29% 11% 11.20 0.47 4.3 9 Tramp 21 TRANSCAPRICORN 238 53% 34% 12% 10.21 0.40 2.6 10 Tramp 22 TRANSANDROMEDA 273 55% 35% 10% 11.01 0.47 2.7 10 Tramp 12 700 dwt Nr NAME Time days % sea % moored % anchored Avg Speed tonne/ h Avg Voy. Sea (days) Op. Nr. 23 BBC GEORGIA 216 57% 31% 11% 11.21 0.59 3.3 11 Liner 24 BBC VERMONT 101 50% 35% 15% 10.52 0.50 3.0 11 Liner 25 BBC ALASKA 236 56% 27% 17% 11.94 0.64 4.4 11 Liner 26 BBC MARYLAND 262 61% 25% 14% 12.07 0.69 5.5 11 Liner 27 BBC FLORIDA 316 64% 25% 11% 12.21 0.68 5.0 11 Liner 28 BBC DELAWARE 293 70% 25% 5% 11.35 0.55 5.2 11 Liner 29 BBC ZARATE 245 59% 31% 10% 11.75 0.61 3.3 11 Liner 30 BBC MAINE 350 59% 31% 10% 11.75 0.63 3.8 11 Liner 31 BBC MONTANA 281 63% 26% 12% 11.00 0.53 4.5 11 Liner 32 BRATTINGSBORG 120 51% 37% 12% 11.44 0.64 3.2 12 Liner 33 CLIPPER ANGELA 186 51% 31% 18% 12.65 0.68 5.9 13 Tramp 34 SE PACIFICA 372 64% 25% 11% 10.97 0.53 6.0 14 Tramp 35 SE PELAGICA 341 54% 36% 11% 11.36 0.57 5.5 14 Tramp 36 SE PANTHEA 470 71% 19% 9% 11.89 0.60 7.0 14 Tramp 37 SE POTENTIA 246 57% 30% 13% 11.46 0.59 4.4 14 Tramp 38 ROSARIO 84 33% 51% 8% 11.01 0.53 2.2 15 39 MARSELISBORG 69 51% 18% 4% 12.70 0.73 4.9 16 40 FREDENSBORG 97 51% 29% 38% 12.58 0.76 2.9 17 41 ELSBORG 46 59% 40% 2% 12.01 0.62 5.4 18 42 ELLENSBORG 35 79% 16% 5% 10.98 0.50 7.0 18 43 JANNES H No data available 44 AGGERSBORG No data available Table 5.2: Operational profiles. RESULT 27 7 ships in the 12 700 dwt group, i.e. number 38-44 are not further analysed in study. Their total time in analysis fell below 100 days. 100 days was set as a limit in order to minimise special conditions of a certain voyage. The type of traffic ship has been occupied in during the period of study is showed in last column of Table 5.2. This parameter has been set from analysis of the ship operational pattern; both from visual track analysis and from analysis of port call regularity. Table 5.3 and Table 5.4 shows minimum, maximum, and average time spent as percentage of total voyage time, i.e. sea+port+anchor. 7 700 dwt Sea Port Anchor Min 45% 19% 1% Max 69% 45% 23% Average 57% 32% 11% Liner Tramp Sea Port Anchor Sea Port Anchor Average 57% 35% 9% 57% 29% 14% Table 5.3; 7 700 dwt, Min, max, and average percentage of time spent in each operational mode. In the group of 7 700 dwt ships, the average time in port was 32%. The difference between the ship that spent the longest time in port (45%) and the ship that spent the least time in port (19%) was 26%. The average time at anchor was 11%, with a difference of 22% between the ship that spent the most time at anchor (23%) and the ship that spent the least time at anchor (1%). The average time spent at anchor is higher for the tramp ships (14%) compared with the liner ships (9%), while the liner ships spent an average 5% more time in port than the tramp ships. 12 700 dwt Sea Port Anchor Min 50% 19% 5% Max 71% 37% 18% Average 59% 29% 12% Liner Tramp Sea Port Anchor Sea Port Anchor Average 59% 29% 12% 60% 28% 12% Table 5.4; 12 700 dwt, Min, max, and average percentage of time spent in each operational mode. In the group of 12 700 dwt ships, the average time in port was 29%. The difference between the ship that spent the longest time in port (37%) and the ship that spent the least time in port (19%) was 18%. The average time at anchor was 12%, with a difference of 13% between the ship that spent the most time at anchor (18%) and the ship that spent the least time at anchor (5%). Only small or none difference can be found in the distribution between the operational modes when comparing the 12 700 dwt liner ships against the tramp ships. 28 5.2 FUEL CONSUMPTION Fuel consumption is calculated from each individual AIS recording and summarized for all voyages of each ship. The results are showed in Figure 5.1 and Figure 5.2. The differences in fuel consumption derive from the choice of speed during the sea leg of the voyage. These are theoretical calculations of fuel consumption; actual fuel consumption could show on a different result. However, the fuel consumption formula used in study (4.2.1 Fuel consumption) has showed on a very reliable result in comparison with actual fuel consumption measured on board ships. (IHS Fairplay, 2012) Figure 5.1; 7 700 dwt, fuel consumption per hour. The 7 700 dwt ship with lowest fuel consumption was ship 8 (Vliediep) with 0.34 tonne/h, 33% less than the ship with the highest fuel consumption, ship 9 (Nordland). The average fuel consumption of the 7 700 dwt ships was 41 tonne/h. Figure 5.2; 12 700 dwt, fuel consumption per hour. Ship 24 (BBC Vermont), had the lowest fuel consumption per hour of the 12 700 dwt ships with 0.50 tonne /h, 27.5 % less than the ship with highest fuel consumption per hour, ship 26 (BBC Maryland). The average was 0.6 tonne/h. No general difference in either group between liner and tramp ships could be found. The group of 7 700 dwt show a correlation between length of voyage and fuel consumption per hour, where an average of shorter voyages give a higher fuel consumption. However, the group of 12 700 dwt ships show on the contrary relation between length of voyage and fuel consumption. 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60 1 2 3 4 5 6 7 8 9 10 11 12 to nn e (f ue l) / h ou r Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships 0,30 0,35 0,40 0,45 0,50 0,55 0,60 0,65 0,70 23 24 25 26 27 28 29 30 31 32 to nn e (f ue l) / h ou r Liner ships 33 34 35 36 37 Tramp ships RESULT 29 5.3 TRANSPORTATION WORK Transportation work shows how many tonne-kilometre each ship has carried out. As each ship has an unequal total time of analyse in the study, a measure of tonne-km per hour is used in order to compare the ships against each other. Figure 5.3 shows all 7 700 dwt ships transportation work per hour. Figure 5.3; 7 700 dwt, transportation work per hour. The average of the 7 700 group where 97 008 tonne-km/h. The best performing ship was ship 10 (Nordland), which had done approx. 39% more transportation work than the lowest performing ship, 16 (Pensilvana). There is also a significant difference between the liner and the tramp ships. Liner ships have an average of 102 988 tonne-km/h, 13% more than the average of the ships occupied in tramp traffic. Hydrostatic data was not available for the study for the 12 700 dwt ship. Transportation work calculations have therefor not been carried out. 50000 60000 70000 80000 90000 100000 110000 120000 1 2 3 4 5 6 7 8 9 10 11 12 to nn e- km / h Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships 30 6 ANALYSE This chapter presents an in-depth analysis and discussion of the operational profiles. Benchmarks of the ships in study are set for each operational mode and measures in terms of fuel efficiency improvement are discussed. 6.1 SPEED Choice of speed is the most important parameter in the terms of energy efficiency. The distribution of speed is showed in Figure 6.1, and Figure 6.2. The lines in the diagrams correspond to the percentage of time each ship has spent in each speed. Figure 6.1; Speed distribution of the 7 700 dwt ships. Figure 6.2; Speed distribution of the 12 700 ships. The fuel consumption of each ship is in direct relation to their distribution of speed. As showed in Figure 6.1, the ship with the best performance of fuel consumption per hour (Table 5.2), ship 8 (Vliediep) is also the ship with the lowest peak of speed in the speed distribution diagram. And the ship with the highest fuel consumption per hour, ship 10 (Wisaforest), is the 0% 10% 20% 30% 40% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 % o f t ot al ti m e at s ea . 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Average 0% 10% 20% 30% 40% 50% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 % o f t ot al ti m e at s ea . Speed (knots) 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Average ANALYSE 31 ship with the highest peak of speed and most time of all ships at a speed at or close to the maximum of 14 knots. Minimizing the time spent in higher speed can reduce fuel consumption. The average speed of the ship with or without cargo could be different depending on the type of traffic ship is occupied in. Ships in systematic liner traffic run according to a schedule, pre set from the speed of the ship, and the chose of speed normally remains the same from voyage to voyage. The ships in tramp traffic normally adjust their speed after the contractual agreements, where the laytime and the possibility of extra earning through demurrage could tempt the choice of a higher speed, despite higher fuel consumption. Figure 6.3 and Figure 6.4 show the average speed of the sea leg of the voyage before loading of cargo compared with before unloading of cargo. Figure 6.3; 7 700 dwt, average speed without and with cargo on-board. The average speed is 5% (0.48 knots) higher before unloading than before loading of cargo in Figure 6.3. However the average speed for the tramp ships is 8% higher with cargo on-board, compared with the difference of only 2% for the liner ships. Figure 6.4; 12 700 dwt, average speed without and with cargo on-board. The average speed in Figure 6.4 is 4% (0.47 knots) higher before unloading than before loading of cargo for the 12 700 dwt ships. The difference in speed with cargo compared to without cargo is 4% for the liner ships and 5% for the tramp ships. 7 8 9 10 11 12 13 1 2 3 4 5 6 7 8 9 10 11 12 Av er ag e sp ee d (K no ts ) Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships Without cargo With cargo 9 10 11 12 13 23 24 25 26 27 28 29 30 31 32 Av er ag e sp ee d (K no ts ) Liner ships 33 34 35 36 37 Tramp ships Without cargo With cargo 32 6.1.1 POTENTIAL SAVINGS OF SPEED REDUCTION The ships average speed is significant higher than the best economic speed (see chapter 3.6.2). The economic speed gives the lowest cost per nautical mile. Contractual agreements and individual conditions of a trade could change the most favourable speed. Neither does the economical model take into consideration the potential of extra earning i.e. more voyages by running the ship at a higher speed. 7 700 dwt 12 700 dwt HFO 1% sulphur Calculated avg. speed 10.58 knots 9 374 USD/day 11.57 knots 13 168 USD/day Economic speed 8.6 knots 7 476 USD/day 9.7 knots 10 769 USD/day Potential saving 20 % 18% MGO 0,1% sulphur Calculated avg. speed 10.58 knots 10 918 USD/day 11.57 knots 16 225 USD/day Economic speed 7.8 knots 7 675 USD/day 8.7 knots 11 657 USD/day Potential saving 30% 28% Table 6.1; Potential savings in cost per day. There is a substantial potential in cost reduction by running the ship at a slower speed, as presented in Table 6.1, where cost per day is calculated from the cost per nautical mile, see Appendix 1 - Economic speed. The potential is most likely to increase from 20% to 30% as more stringent environmental regulations will lead to higher bunker prices, i.e. change to 0,1% sulphur fuel. 6.1.2 OPERATOR DIFFERENCES A higher speed means higher fuel consumption. However, when observing over time, as in this study, a higher average speed does not necessarily mean higher fuel consumption, as illustrated in Figure 6.5 and Figure 6.6. As previous analysis showed, the distribution of speed over time is what determines the fuel consumption. Figure 6.5; 7 700 dwt, fuel consumption vs. average speed. The further to the right in the diagram of Figure 6.5, the higher average speed during the period of study. The higher up in the diagram, the higher fuel consumption. The reason why a ship could achieve lower fuel consumption at a higher average speed can be derived from the fuel consumption diagram (Figure 4.2), which shows an exponential increase in fuel 0,30 0,40 0,50 9,00 10,00 11,00 12,00 Fu el c on su m pt io n (t on ne / h ) Avarage speed at sea (Knots) Feederlines Transatlantic Navesco SA Harren & Partner Ship Mgmt Flinter Shipping BV Scan-Trans Chartering KS Hermann Buss GmbH & Cie KG Onego Shipping & Chartering BV Strahlmann E Reederei eK Universal Africa Lines NV ANALYSE 33 consumption at higher speed. A few voyages or part time of a voyage at high speed increases the fuel consumption significant. There are a few differences between the operators of the 7 700 dwt ships. The two Transatlantic operated ships (21 and 22) stand out significant from the rest. Their fuel consumption per hour is approximately 5% higher then the average fuel consumption at the same speed. A comparison of the Transatlantic ships with the best performing ships, operated by Scan- Trans chartering (ship 19) and Hermann Buss (ship 15), shows on an additional potential of 5% in reduction in fuel consumption per hour. Figure 6.6; 12 700 dwt, fuel consumption vs. average speed. OXL NV (ship 32) had approx. 7% higher fuel consumption per hour than the average of the 12 700 dwt ships. The Clipper Projects operated ship (ship 33) was the best performing ship with approx. 6% lower fuel consumption than the average. A correlation between length of voyage and fuel consumption performance was found. In both groups, the ships with the best performance, i.e. lowest fuel consumption compared to average speed, were also the ships with a higher percentage of longer voyages. Reversely, the ships with higher percentage of shorter voyages were also the ships with the highest fuel consumption. From analysis, the optimal in terms of fuel consumption is to keep the speed as low as possible and constant (Figure 6.7). As an example; if the ship sailing on a route that consist of several sea legs, it will be better to operate the ship at the same speed on all sea legs of the route, than operating the ship at a reduced speed at one leg and increased speed on another leg, even though the total time and average speed would be the same. 0,45 0,55 0,65 0,75 10,00 10,50 11,00 11,50 12,00 12,50 13,00 Fu el c on su m pt io n (t on ne / h ) Avarage speed at sea (Knots) BBC Chartering & Logistic GmbH SE Shipping Lines Pte Ltd OXL NV Clipper Projects A/S Knots % o f t im e Figure 6.7; Choosing a slower speed will reduce fuel consumption. 34 However, the profitability of a ship is also a function of its potential earning. A constant speed for a ship in tramp traffic could reduce the possibilities to carry additional spot cargoes and reduce the final result, even though the reduced fuel consumption. 6.2 PORT There were big spreads in distribution of the result when comparing the time in port per voyage among the ships. However, this difference could be a result of a single port call, significantly longer then the majority of the port calls made by the ship. This was confirmed when a comparison with the time in port, excluding the calls with a significant longer port time, as illustrated in Figure 6.8 and Figure 6.9. Figure 6.8; 7 700 dwt, Time in port per voyage. The ship with the least port time of the 7 700 dwt ships was ship 9 (Nordland), with an average of 31 hours per voyage. Figure 6.9; 12 700 dwt, time in port per voyage. The ship with the shortest time in port per voyage among the 12 700 dwt ships was ship 24 (BBC Vermont), with 38 hours per voyage. The group of 12 700 had an average of 8.5% longer time in port than the smaller ships. However, the larger ships have approximate 65% more cargo capacity than the smaller ships, 12 700 dwt compared with 7 700 dwt, an average not more than 8.5% longer time in port show that the time in port not only depend on the ship cargo capacity and cargo handling. 20 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 11 12 Ti m e in p or t / vo ya ge Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships Time in port / voyage Time in port / voyage, excl. long 20 25 30 35 40 45 50 55 60 23 24 25 26 27 28 29 30 31 32 Tm e in p or t / vo ya ge Liner ships 33 34 35 36 37 Tramp ships Time in port / voyage Time in port / voyage, excl. long ANALYSE 35 The relationship between time in port and amount of cargo loaded/unloading was examined in Figure 6.10 and Figure 6.11. The analysis shows that there might be a correlation. However, the result of analysis is too widely spread in distribution to draw any conclusions of the analysis. Figure 6.10; 7 700 dwt, hours in port with loading operation Figure 6.11; 7 700 dwt, hours in port with unloading operation. The loading/unloading rate, tonne/h, indicates the efficiency of the ship cargo handling. Figure 6.12 indicate a faster handling in unloading operations than in loading operations. However, as previously showed in analysis, the spread of the result is too randomly distributed to draw any conclusions about loading/unloading rate. 0 20 40 60 80 100 120 ho ur s in p or t l oa di ng / vo ya ge cargo loaded (tonne) 0 20 40 60 80 100 120 140 160 180 200 ho ur s in p or t d is ch ar ge / vo ya ge cargo unloaded (tonne) 36 Figure 6.12; 7 700 dwt, loading/unloading rate. (Tonne per hour). Several factors affect the ship time in port: • Time of arrival (within/outside working hours) • Use of port cranes/own cranes • Technical performance • Space availability • Manning • Individual port capacity/efficiency (Christiansen , Fagerholt, Nygreen, & Ronen, 2007) Time in port is to compile from AIS data. However, a complete analysis of what determines the time used in port is not possible with data available in study. Every port has its own characteristics and every type of cargo requires different handling. The general cargo ships in study are equipped with own gear, i.e. cranes, and AIS data do not include any information about the usage of on-board equipment. The time in port could also include bunkering operations, with or without cargo operation, which could mislead interpretation of draught data and loading/unloading calculations. To summarise, it is important to recognise that is not possible to examine and give the whole picture the ship port performance without precise data from ship and from port. 6.3 ANCHORING With a minimum anchoring strategy, a slower speed could be chosen with the equal amount transportation work, which reduce cost and increase profitability of the voyage. Ships in systematic liner traffic normally have very little anchoring time due to a precise schedule and good port relations. Anchoring is only carried out due to circumstances beyond the control of the operator, e.g. bad weather, strike, etc., or if the ship is taken out of service. Ships in tramp traffic normally spend more time at anchor than the liners. Tramp ships could anchor while waiting for new orders and cargo. They are also more often at anchor outside the loading/unloading port waiting for berth/cargo to be ready. The contractual agreements with the charterer sometimes favour inefficient operation, where the ship operator runs the ship faster than necessary due to the possibilities of demurrage compensation. The Figure 6.13 and Figure 6.14 show the difference of anchoring time before loading compared with before unloading of cargo. 0 20 40 60 80 100 120 1 2 3 4 5 6 7 8 9 10 11 12 Lo ad in g/ U nl oa di ng r at e (t on ne /h ) Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships Loading Unloading ANALYSE 37 Figure 6.13; 7 700 dwt, anchoring time before loading/unloading of cargo per voyage. The 7 700 dwt ships have an average of 47% longer anchoring time before loading than before unloading of cargo. And the group of 12 700 dwt ships have an average of 42% longer anchoring time before loading than before unloading of cargo. Both groups of ships show less difference among the liner ships between anchoring time before loading and before unloading than the ships in tramp traffic. Figure 6.14; 12 700 dwt, anchoring time before loading/unloading of cargo per voyage. The analysis of the tramp shows that the anchoring time waiting for cargo is significantly longer than the anchoring time with cargo on-board. This result could be different in a different period of study. The analysis is carried out with data from 2010-2011 when the global economic situation was in harsh and started to slowly recover after the finical crisis in 2009. An analysis of the anchoring time during the economic peak period (200-2007) before the global recession would most likely show on less time at anchor. 6.4 FUEL EFFICIENCY Fuel efficiency could be improved by reducing the energy intensity. Energy intensity is measured as the amount of energy required per unit output. Lower energy intensity means less energy needed to produce the same output, i.e. improved energy efficiency. In study, energy is measured in fuel consumption (tonne), and the output in terms of transportation work (tonne-km). 0 10 20 30 40 50 60 70 80 90 1 2 3 4 5 6 7 8 9 10 11 12 An ch or in g ti m e / vo ya ge (h ou rs ) Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships Before loading Before unloading 0 10 20 30 40 50 60 70 23 24 25 26 27 28 29 30 31 32 An ch or in g ti m e / vo ya ge (h ou rs ) Liner ships 33 34 35 36 37 Tramp ships Before loading Before unloading 38 Table 6.2; 7 700 dwt ships energy intensity. The ship with lowest energy use per tonne-km was ship 8 (Vliediep), which used 37% less fuel per tonne-km than the ship with the highest energy intensity, i.e. ship 22 (Transandromeda). The ships in tramp traffic used approximately an average of 19% more energy per tonne-km than the liner ships, this difference is a result of the tramp ships higher speed with cargo on-board (Figure 6.3) and less transportation work carried out (Figure 5.3). 6.4.1 CAPACITY UTILIZATION The potential of energy efficiency improvement by reduction of energy intensity could be found by analysing the capacity utilization, which is the ratio between the actual output and the maximum potential output for the ship, expressed in percentage of the potential output. The calculations are based on the AIS draught data and same operational pattern is assumed, i.e. same speed and distance as the ship carried out in 2010-2011. The maximal capacity was set to the deadweight of the ship, i.e. 7 700 dwt. Capacity utilization = transportation work (tonne−km) maximum transportation work (dwtkm) Figure 6.15; 7 700 dwt, capacity utilization. The results of capacity utilization were distributed in the range between 52% and 75%, with an average of 64%. The results could be an effect of individual market conditions and strategy. However, a higher capacity utilization was found among the liner ships, 69%, compared with 59% for the tramp ships. This difference could be derived from more time in 2,00 2,50 3,00 3,50 4,00 4,50 5,00 1 2 3 4 5 6 7 8 9 10 11 12 fu el (g ra m ) / to nn e- km Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships 30% 40% 50% 60% 70% 80% 1 2 3 4 5 6 7 8 9 10 11 12 Ca pa ci ty u ti liz at io n Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships ANALYSE 39 ballast conditions, which might be a result from the lack of return cargo and operational imbalance of tramps. Study of short sea shipping (RoRo) made by Styhre (2009) show on a desirable capacity utilization between 75% and 88%. A maximum utilization might be profitable in the short run. However, a higher capacity utilization could lead to a more vulnerable liner system. An excess capacity is important in order to be able to serve customers even when there is a peak in demand of transportation. Not being able to respond to market fluctuations could lead to loss of customers in the long term. Liner ships in the study has yet some extra capacity in spare before reaching the optimum capacity utilization span, possible as a consequence of the financial turmoil following 2008. In order to increase the capacitation utilization, either the demand of, or the supply of the transportation service has to be adjusted. The demand of an existing ship could be enhanced by promotional activity or by a strategic change, e.g. repositioning, price adjustment, or by adding value to the customers. By reducing supply of transportation, i.e. reducing the capacity, a higher utilization could be achieved if the company has multiple ships serving the same segment. One way of reducing capacity is to scrap ships or simply by slow steaming. 6.5 THEORETICAL SAVINGS The anchoring operation is most often seen prior berthing at port. Congestion is one common reason why not the ship could berth directly upon arrival at port. A just in time/ perfect arrival strategy could help the ship operator to choose the right speed in order to minimize time at anchor and significant fuel savings could be achieved. The theoretical savings calculation assumes same distance and the same number of voyages during the same period of time as previous. You could expect the ship to increase the number of voyages if the time anchoring is reduced; however, this has not been taken into consideration in these calculations. Figure 6.16 and Figure 6.17 show the results from the theoretical savings calculation. Figure 6.16; 7 700 dwt, theoretical fuel consumption savings. (Tonne/hour). The theoretical and potential average fuel saving is 26% for the 7 700 dwt ships, where the tramp ships show on a higher potential saving (32%) compared with the liner ships (21%). Two ships, ship 9 (Nordland) and ship 10 (Wisaforest), with very little time at anchor, could only save 3% of fuel with no anchoring. The most saving, 49%, could ship 16 (Pensilvania) 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 1 2 3 4 5 6 7 8 9 10 11 12 To nn e (f ue l) / h ou r Liner ships 13 14 15 16 17 18 19 20 21 22 Tramp ships Previous New (just in time) 40 achieve, however, this ship had a few voyages with very long anchoring time in relation to voyage time, which could affect the reliability of the result due to other circumstances, for example; no cargo available, maintenance, etc. Figure 6.17; 12 700 dwt, theoretical fuel consumption savings. (Tonne/hour) The 12 700 dwt group of ship could achieve an average saving of approximately 26%. The theoretical saving calculations are based on average speed per voyage, which are an average of 14% lower, for both the 7 700 dwt ships and the 12 700 dwt ships, than the actual fuel consumption calculated per AIS recording, i.e. observed average speed per hour. This difference origin from the ship distribution of speed (see 6.1 Speed), and confirms the importance of an even operational speed in order to minimize fuel consumption. The theoretical potential of a no anchoring strategy show significant opportunities in reduction of fuel consumption. 26% is not possible to achieve as an average among the ships, even though it shows on a great potential and only a few percentage in less anchoring time of an individual ship could result in significant savings in fuel consumption. The importance of collaboration between all parties in shipping is vital to carry out a no anchoring strategy. The economic benefit must be shared between the stakeholders in order to fulfil a sustainable no anchoring strategy. 0,10 0,20 0,30 0,40 0,50 0,60 0,70 23 24 25 26 27 28 29 30 31 32 To nn e (f ue l) / h ou r Liner ships 33 34 35 36 37 Tramp ships Previous New (just in time) CONCLUSION 41 7 CONCLUSION The study highlights two main research questions: • Is it possible to benchmark ship fuel efficiency from AIS data? • What benc