Analysis of energy conversion in ship propulsion system in off-design operation conditions

Energy and Sustainability II 449 Analysis of energy conversion in ship propulsion system in off-design operation conditions W. Shi 1, D. Stapersma 1,2 & H. T. Grimmelius 1 1 Department of Marine & Transport Technology, Delft University of Technology, The Netherlands 2 Netherlands Defence Academy, The Netherlands Abstract In this paper, the ship propulsion system is broken down into components to estimate their individual activities in terms of energy conversion. For common used propulsion system of merchant ships, the efficiency of each propulsion component is presented. In the case study, the interaction between all the propulsion components in energy conversion is investigated by means of computer model simulation. Based on a selected ferry, the results demonstrate that, when operating in off-design conditions, the total energy conversion efficiency is slightly different from that in design condition, whereas, in terms of the ton-mile specific fuel consumption and energy index, the part loading of ship and off-design speed of ship scenarios show different impact, but both are much different from that in design condition. Keywords: energy conversion, ship propulsion system, off-design operation, tonmile specific fuel consumption, energy index. 1 Introduction Industrialization and technological development cause people to use increasing quantities of goods and energy. When looking at the whole transportation system, shipping transportation is a necessary part in the globalization process and is in many instances the only means of transporting the goods. Over the last three decades, the shipping transportation has grown on average by five percent doi:10.2495/esu090411

450 Energy and Sustainability II every year (measured in ton-miles), and shipping is by far the most used transportation mode (90% of all goods measured in ton-miles) [1]. Compared with other transportation modes, shipping transportation is considered the most fuel efficient one [2]. However, with the increase of transportation demand, both from economical and environmental points of view, there is an increasing demand in fuel saving and reducing emissions. This paper will explore the influence of each component in the ship propulsion system and, by means of computer simulation, present the interaction between the individual components in terms of energy conversion from fuel to ship movement. 2 Power transmission from fuel to ship movement From the viewpoint of the overall power transmission system, the input is the amount of fuel (supplied energy), while the output is the ship moving at a specific speed (demand energy), and the energy conversion is corrected by the energy conversion efficiency (η ec ). To go from the consumed fuel to the ship moving, the total energy conversion efficiency (η ec ) could be divided into four parts, as shown in fig. 1: the hull efficiency (η E ), propeller efficiency (η O η R ), the transmission efficiency (η S η GB ) and the engine efficiency (η E ). In the following section, details of each part are presented. Figure 1: Power chain: overview of powers and efficiencies [3]. 3 Operational characteristics of propulsion components In this paper, the common used ship propulsion configuration is considered to reveal the details of energy conversion through the ship propulsion system, which consists of: the diesel engine (as prime mover), the gearbox and the shaft (as transmission system) and the propeller (as propulsor). Furthermore, since the thrust delivered by the propeller is to overcome the ship resistance, which strongly depends upon the ship hull form, as illustrated in fig. 1, the ship hull is also involved in this study.

Energy and Sustainability II 451 3.1 Main diesel engine The diesel engine converts chemical energy to mechanical energy by means of internal combustion. There are two parallel ways to assess the performance of a diesel engine: - From the inside of the engine. When looking at a single cylinder cycle, starting with the fuel injection, with the fuel injected into the combustion chamber an amount of energy (Q f ) is input to the combustion process. The energy contained in the fuel is corrected by the combustion efficiency: η cb. The amount of released energy (heat) that actually is available to perform work is corrected by the heat input efficiency: η q. The cycle efficiency is expressed in the thermodynamic efficiency: η th. Thus, the net work (W in ) produced by a single cylinder cycle (including both output work and gas exchange work) is found: Win = Qf ηcb ηq ηth (1) - From the outside of the engine. The engine performance could be described by the so called effective efficiency: η E, which represent the ratio between the engine effective work (the real output work of the engine) and the total input energy (the energy contained in the injected fuel). Wef η E = (2) Qf Two other important efficiencies, which indicate engine performance, are the indicated efficiency η i and the mechanical efficiency η m. W η in i = = ηηη Q cb q th (3) η W f η ef E m = = (4) Win ηi The indicated efficiency η i represents the thermodynamic performance of the engine including heat and combustion losses, while the mechanical efficiency η m indicates the mechanical performance of the engine. (a) (b) Figure 2: Diesel engine efficiencies in propeller load operation. Fig. 2(a) and (b) illustrate the diesel engine efficiencies when operating in propeller load [4]. Due to the large air/fuel ratio, generally the combustion in a

452 Energy and Sustainability II diesel engine is considered complete, thus, η cb is 100%, as shown in fig. 2(a). Also, fig. 2(a) illustrates that both η q and η th increase when decreasing the engine speed. Because of the lower cylinder temperature in low speed operation conditions, the heat losses due to heat transfer to the cylinder walls during the combustion become a smaller fraction, which makes the heat input efficiency increase slightly with decreasing engine speed. Meanwhile, in low speed conditions, due to the decrease of injected fuel, the total heat input to operation process decrease, and, since the smaller account of air also results in smaller compression work, which increase the net output work relatively. So, although both output work+gas exchange work and input heat decrease in low speed conditions, in terms of the ratio between work and heat, (η th ), the value increases, representing better combustion process. When looking at the overall operation of the engine, due to the improvement of the combustion process in low speed operation conditions, η i slightly increases when decreasing the engine speed, as shown in fig. 2(b). Concerning the mechanical losses, although in high speed operation conditions, the mechanical losses is absolutely larger than in low speed operation conditions, however, in terms of the mechanical efficiency, in low speed operation condition, η m drops dramatically, since the mechanical losses account for a larger fraction of total output power than that in high speed operation conditions. Thus, the benefit of improved combustion in low speed conditions is counteracted by the deteriorated η m, resulting in big drop of η E. In this study, the engine efficiencies map is implemented in the simulation model by means of lookup tables to present the engine operation conditions. 3.2 Gearbox If a medium or high speed diesel engine is installed on a ship, a reduction gearbox is required to provide a speed-torque conversion. In general analysis, the power losses in gearbox is 1~2% for single stage reduction gearbox and 3~5% for complex gearbox, with two or three reduction stages [3]. It is generally true that a gearbox when fully loaded will exhibit higher efficiency than when it is partially loaded. Figure 3: Gearbox efficiencies in design and off-design conditions.

Energy and Sustainability II 453 In [5], it is demonstrated that the most significant sources of power loss in a spur-gear system (one reduction stage gearbox) are the gear-mesh P m (including the sliding frictional component and a hydrodynamic rolling component), gearwind-age P w (the power required to rotate the pinion and gear in the air-oil atmosphere present within the gearbox) and support-bearings P br (bearing friction). As illustrated in fig. 3, η GB is determined by the speed and load. PS ( PB Pm Pw Pbr) ηgb = = (5) PB PB However, one should note that, although the gearbox efficiency decreases fast in low load conditions, as shown in fig. 3, the difference is very small, at full speed about 3% between full load and minimum load. Thus, in general analysis, the efficiency drop of gearbox in off-design conditions could be neglected. 3.3 Shaft Shafts are used to connect prime mover, gear box and propulsor. They transfer both speed and torque through the entire propulsion system. The main source of power loss in shaft transmission is the friction in the support-bearing. Because of the high transmission efficiency of shafts in all load and speed conditions, the η S is set as 99.5% for each single shaft and remain constant in this study. 3.4 Propeller The propeller is used to generate thrust to overcome the ship resistance. Normally, an open water diagram is used to determine the propeller operational behaviour, in particular its open water efficiency η O. However, refer to the open water diagram, the fact is that propellers are tested in an open water tank or tunnel, in which the flow in front of the propeller is uniform during the whole test, apparently, this is rarely the case in reality, thus, relative rotative efficiency, η R is introduced in fig. 1, to convert the open water propeller power to realistic propeller power. Figure 4: Four quadrants open water diagram: D prop = 4.8m, A E /A 0 =0.7. Based on experimental data, refer to [6], the Maritime Research Institute Netherlands (MARIN) developed C T *, C Q * versus β diagram to describe the

454 Energy and Sustainability II entire four quadrants open water diagram of propellers within Wageningen B- Screw Series. Then, the open water efficiency η O could be calculated as: * P 1 Tprop v T A CT ηo = = = 0.35 tan β (6) * PO 2π Qprop nprop CQ Fig. 4 shows an example of the four quadrants open water diagram of a controllable pitch propeller. On the basis of [7], where a neural network prediction is presented to produce four quadrants diagram of any propeller within Wageningen B-Screw Series, in this study, lookup tables are used in the model to simulate propeller operations. In [8], the relative rotative efficiency η R is calculated as: for single screw ship: A η 0.9922 0.05908 E R = + 0.07424( CP 0.0225 lcb) (7) A0 for twin screw ship: η R = 0.9737 + 1.11( CP 0.0225 lcb) 0.06325P (8) D Eqns. (7) ~ (8) illustrate that, η R is dependent upon ship body parameter and propeller properties, thus, it remains constant for a specific ship. 3.5 Ship hull As illustrated in Fig. 1, η H represents the ability of the conversion from thrust power to effective towing power. P Rship v E S η 1 H = = = t (9) PT Tprop va 1 w The thrust deduction factor (t) and the wake factor (w) are mainly dependent upon the ship hull and the ship speed. And also, these two factors could be affected by external disturbances, such as the sea state, the fouling, the ship loading and the depth under keel. Based on [8] and [9], Fig. 5 illustrates an example of the relationship between hull efficiency and ship speed, without any external disturbances. 4 Model simulation 4.1 Model structure In order to investigate the interaction between all propulsion components in offdesign operation conditions, based on the previous analysis on individual propulsion components, a simulation model is built in Matlab Simulink, as shown in fig. 6. Also, refer to [10] for more details of ship operation simulation. There are two input signals to determine the ship operation condition. The desired engine speed (n_engine_set) is the command signal to control the ship speed, and the ship loading factor (x) represents the ship loading condition. To achieve the dynamical balance of this ship propulsion system model, two dynamical systems are introduced to connect the propulsion components: - Shaft rotation system, which deals with the dynamic balance between supplied torque and demanded torque, generating shaft revolution speed.

Energy and Sustainability II 455 Figure 5: Hull efficiency. Figure 6: Block diagram of ship propulsion model. - Ship translation system, which deals with the dynamic balance between propeller thrust and ship resistance, generating ship speed. 4.2 Case study 4.2.1 Ship and modelling profile Ship A cargo/passenger ferry, with which the detail information is available, (see table 1), is selected as an example to execute the simulation and explore the details of energy conversion through its propulsion system. Modelling Voyage Profile A simplified scenario is used to indicate the voyage profile. As shown in fig. 7, the voyage profile is represented by the desired engine speed. Put into words, first the ship is manoeuvring in the original port for 30mins, then, sailing to the open ocean for 4 hours, and then, when approaching the destination, spending another 30 mins manoeuvring before stopping. Thus, total simulation period is

456 Energy and Sustainability II Figure 7: Modelling profile. Table 1: Details of the selected ferry. NAME: Stena Jutlandica CLASSIFICATION: 100A1 DIMENSIONAL PARAMETERS Length of waterline, (m): 169 Breadth, (m): 27.8 Draught, (m): 5.8 Dead weight, (ton): 5640 Mass Displacement, (ton): 17326 GEARBOX Type: NDSHL 2600 Reduction: 550:150 MAIN DIESEL ENGINE Type: MAN 9L40/54 MCR, (kw): 4*6480 Rated speed, (rpm): 550 PROPELLER Type: CPP Diameter, (m): 2*4.8 5 hours, including 1 hour transient operation (20%) and 4 hours steady-state operation (80%). 4.2.2 Operation in design condition The design operation condition in this case study is set as the condition where, the ship is fully loaded, and the engine is running at 80% MCR. As shown in fig. 8(a) ~ (f), it is evident that, in transient conditions, there are large fluctuations in terms of energy conversion efficiencies from the fuel to the ship moving. On the other hand, looking at the steady-state operation, the propeller open water efficiency (η O ) and indicated engine efficiency (η i ) make a larger contribution than others to the entire energy conversion efficiency (η ec ). mf mf tmsfc = = (10) ( w vs) ( w Dis) Q f Qf I E = = (11) ( w vs) ( w Dis) where w is the weight of benefit loading, refer to [11]. When looking at the ton-mile specific fuel consumption, the mean value of the entire simulation voyage is 18.80g/ton-mile, (refer to eq. (10)), while in

Energy and Sustainability II 457 (a) (b) (c) (d) (e) Figure 8: (f) Operation profile in design condition. terms of energy index, based on [2], an extended energy index is used in this study, with the expression as eq. (11). Thus, in the design condition, the mean value of the energy index is 0.041. 4.2.3 Operation in off-design condition According to the definition of the design operation condition in the previous section, the off-design operation conditions consist of 2 categories: ship is part load and the engine is running at off-design speed.

- Off-design 458 Energy and Sustainability II - Ship is part load. Details about ship behaviour and the corresponding fuel consumption, exhaust emission in part loading conditions are presented in [11]. In this section, the effort is to reveal the impact of ship loading condition on the energy conversion. Table 2: Figure 9: (a) (b) Ship operation profile in part loading condition. Fuel consumption and energy index in different loading conditions. tmsfc (g/ton-mile) steady-state/mean value I E steady-state/mean value 100% loading 19.34/18.80 0.043/0.041 80% loading 38.06/37.08 0.084/0.082 60% loading 185.73/182.41 0.409/0.402 The results present that, for the selected ferry, in steady-state condition, when decreasing the ship loading, the ship resistance decreases, resulting in an increase of ship speed at the same engine speed setting (fig. 9(a)), and a slightly increase of energy conversion efficiency (fig. 9(b)). In terms of ton-mile specific fuel consumption and energy index, the results of steady-state condition and the mean value of the whole simulation voyage are both illustrated in table 2. It is evident that, the loading condition has a strong impact on the ton-mile specific fuel consumption and energy index, since they are both dependent upon the benefit loading conditions. In this study, the simulated manoeuvring stages, during which the ship is operated in relative low engine (ship) speed condition, account for 20% of the total duration, accordingly, the consumed energy (fuel) and covered distance by ship only account for small fraction compare to steady-state condition, thus, refer to (10) and (11), the mean value of both the ton-mile specific fuel consumption and energy index are only slight different from those in steady-state condition, in other words, in analysis of energy conversion during long voyage, the manoeuvring stage (or transient operation) could be neglected. speed of engine or off-design speed of ship. It is argued that, the ship often may operate at low speed for fuel saving or at high speed to meet

Energy and Sustainability II 459 Figure 10: (a) (b) Ship operation profile in off-design speed condition. Table 3: Fuel consumption and energy index at different speed. tmsfc (g/ton-mile) steady-state/mean value I E steady-state/mean value 100% eng. speed 31.52/30.47 0.069/0.067 85% eng. speed 21.73/21.09 0.048/0.046 75% eng. speed 16.84/16.20 0.037/0.036 the schedule. Thus, in real life, the ship is always sailing at off-design speed, say different speed from the recommend service speed. The impact on energy conversion is explored in this section. As shown in fig. 10(a), the low speed operation strategy could save fuel in terms of the fuel consumption rate, in unit of g/s. When looking at the ton-mile specific fuel consumption, (ship remains full load in this case) and the energy index, as shown in table 3, it is demonstrated that, the low speed operation could improve fuel saving and lead to better energy index, only with small penalty in energy conversion efficiency, as shown in fig. 13(b), say, 1.5% decrease of energy conversion efficiency when decreasing engine speed by 15% ( or 13% of ship speed), and 3% decrease of energy conversion efficiency versus 25% decrease of engine speed (or 27% of ship speed). 5 Conclusion The ship propulsion system is broken down into components to estimate their individual influence in terms of energy conversion. The results demonstrate that, in off-design conditions, (low brake power condition for the engine, low load for the gearbox, low speed for the propeller and low speed for the ship), the engine and gearbox efficiencies are lower than those in design conditions; for the propeller, the relative rotative efficiency is independent upon operational condition, while the open water efficiency is dependent upon propeller revolution speed and ship advance speed; in terms of ship hull efficiency, it is

460 Energy and Sustainability II determined by the ship hull and the ship speed, and also could be affected by external disturbances. In the case study, by means of a computer simulation model, the interaction of each propulsion components on energy conversion is presented. When operating in off-design conditions, the overall energy conversion efficiency does not change a lot. In other words, roughly, in analysis of ship operation, the energy conversion efficiency through the entire propulsion system could be treated as a constant value. Concerning with the ton-mile specific fuel consumption and energy index, the simulation results illustrate that, due to they are strongly dependent upon the weight of transferred cargo, the part loading operation is not recommended, but, in low speed condition, the results show good fuel economy and energy index, which agree with realistic ship operation experience. When investigating a long voyage, in which the transient operation only accounts for a small fraction, the mean value could be used in analysis of steady-state operation. References [1] International Council on Clean Transportation (ICCT), Air Pollution and Greenhouse Gas Emissions from Ocean-Going Ships: Impacts, Mitigation Options and Opportunities for Managing Growth, Mar. 2007 [2] Shi, W., Stapersma, D., & Grimmelius, H.T., Comparison study on energy and emissions of transportation modes, Computer and simulation in modern science, Vol 2, WSEAS Press, pp.186~195, Oct. 2008 [3] Woud, J.K., & Stapersma, D., Design of propulsion and electric power generation system, IMarEST Publication, London, reprinted in 2003 [4] Stapersma, D., & Grimmelius, H.T., Comparison, the influence of turbocharger matching on propulsion performance, in Proc. INEC08, Hamburg, German, Apr. 2008 [5] Anderson, N.E., & Loewenthal, S.H., Spur-Gear-System Efficiency at Part and Full load, NASA Technical Paper 1622, Technical Report 79-46, Cleveland, US, Feb. 1980 [6] Kuiper, G., The Wageningen Propeller Series, MARIN Publication 92-001, The Netherlands, 1992 [7] Roddy, R.F., Neural network predictions of the 4-quadrants Wageningen propeller series, Hydromechanics Department Report, NSWCCD-50-TR- 2006/004, 2006 [8] Holtrop, J. & Mennen, G.G.J., An approximate power prediction method, International Shipbuilding Progress, Vol. 29, 1982 [9] Holtrop, J. & Mennen, G.G.J., A statistical power prediction method, International Shipbuilding Progress, Vol. 25, 1978 [10] Shi, W., Stapersma, D., & Grimmelius, H.T., Simulation of the influence of ship voyage profiles on exhaust emissions, in Proc. IMECE08, ASME conference, Boston, US, Oct. 2008 [11] Shi, W., Stapersma, D., & Grimmelius, H.T., Simulation of the influence of loading fraction on operational shipping fuel consumption and emissions, in Proc. WMTC2009, Mumbai, India, Jan. 2009