Propellers for EEDI Compliant VLCC s

Transcription

1 Introduction Propellers for EEDI Compliant VLCC s Jack Devanney Center for Tankship Excellence, USA, djw1@c4tx.org CTX has undertaken a study of the impact of Energy Efficienct Design Index (EEDI) on VLCC design and CO2 emissions. See The Impact of EEDI on VLCC Design and CO2 Emissions. EEDI will result in drastic changes to VLCC powering, which in turn will force drastic changes in VLCC propellers. This paper attempts to predict what these propellers will look like and estimate their performance. The resulting propeller parameters form part of the input to the main study. CTX gratefully acknowledges the help of Brenden Epps of the M.I.T Department of Mechanical Engineering. Not only is he the principal author of OpenProp, an extremely valuable design tool; but his patient help in teaching us how to use OpenProp was absolutely critical to the performance of this study. Method Our approach is to use MIT s Open Source propeller design program, OpenProp, to attempt to design propellers for Phase 1, Phase 2, and Phase 3 compliant VLCC s. OpenProp, developed by Brenden Epps, Richard Kimball and others, is described in OpenProp: An open source design tool for propellers and turbines. It is a modern, lifting line program capable of creating wake adapted propellers. OpenProp takes as input a ship speed, a required thrust at that speed, an RPM, a diameter, a description of the wake, and returns the propeller that generates this thrust at minimum torque while meeting all the other constraints. In all our analyses we held the hull constant. For each of Phase 1, 2 and 3, we held the engine constant. We then searched over diameter, looking for the propeller that gave the vessel the maximum speed without exceeding the engine s torque/power capability. Finally, we checked that combination for EEDI compliance and, if necessary, derated the engine as required to meet the mandated EEDI. 1 In conducting this search we required that the blade loading on the propeller be no higher that that for the standard no-eedi ship. The hull we used is described in Min, K. S., Choi, J. E., Study on the CFD Application for VLCC Hull-Form Design, in 24th Symposium on Naval Hydrodynamics, In the Min/Choi paper, our hull is called Extreme V. Figure 1 shows the hull resistance curve for this hull. This ship used a 9.93 m propeller with a design propulsive efficiency of about For our axial wake profile, we used that measured for SHI hull 1321, a VLLC which has almost exactly the same hull resistance (towing) curve as the Min, Choi hull, and nearly the same thrust deduction and wake fraction. 2 Table 1 shows the SHI 1321 wake profile. 3 Radius(m) Wake Frac The important feature of this profile is that the high wake region extends out to about 5 m from the shaft centerline. This means that smaller diameter propellers must operate at an average wake fraction that is considerably higher than that for the standard 9.9 m prop. We shall see that the improvement in hull efficiency is more than compensated for by reduction in open water efficiency associated with a lower advance ratio. 1 In doing so, we held RPM constant, moving vertically downward in the layout diagram. It is possible that by reducing RPM and going to a little higher pitch, we could come up with a very slightly better fuel consumption for this power and speed. But any such improvement would be insignificant for present purposes. 2 SSMB, Calm Water Model Tests for Hn 1321/22/39/40 300,000 dwt Crude Oil Tanker, Samsung Ship Model Basin, May, 2000,SSMB- S Actually, this is the nominal wake measure behind the model without a propeller converted to the full scale, effective wake by first converting the model effective average wake, as obtained by comparing the self-propulsion test results to the open water results, to the average ship wake per the ITTC method, and then adjusting the measured nominal wake at each radius to obtain an average effective wake that matches the ship figure. 1

2 No EEDI Since we already have a diameter and propulsive efficiency data for the no-eedi ship, running OpenProp for the no-eedi situation is a calibration exercise. The engine used for the no-eedi case is a Sulzer 7rta84t with a full set of slow-steaming mods. The SFC curve fot this engine is shown below. #7rta84t_wm #Sulzer 7rta84t-D cam shaft but using cylinder cutout , , , , , , , , , , , , , , , , , , When we input the hull described in Table A with this engine (27,516 kw at 76 RPM) and a 9.93 m, four bladed propeller to OpenProp, imposing no limit on EEDI, we obtain Table 1. Table 1: Diam=9.930 Blades=4 Hull=../hull hhi vee engine=7rta84t wm rpm=76.0 MCR= max eedi= Diameter Speed MCR η d EEDI 2.31 EAR Loading 53.5 Speed is loaded, calm water (trial) speed. Despite being labeled EHP, this column is the hull s bare hull towing resistance in kilowatts at that speed. The column labeled t is the hull s thrust deduction factor which for both the Hyundai and Samsung hulls was very close to 0.20 regardless of speed. The column labeled Thust is axial force required from the propeller. The column labeled w is the wake fraction which we assume depends only on the propeller diameter. For both the Hyundai and Samsung VLCC s and the design 9.9m propeller, this was close to over the full range of speeds tested. The column labeled η o is the open water efficiency of the minimum torque propeller for this situation according to OpenProp. The column labeled η h is the hull efficiency, (1 t)/(1 w). The column labeled η r is the relative rotative efficiency. which depends on the tangential wake. OpenProp has the capability of 2

3 designing to the actual tangential wake distribution; but Samsung assumed the tangential wake is symmetric about the centerline, which forced all the circumferential means to zero. Therefore, we treated η r as a calibration factor. The 1.03 figure was chosen to generate an overall propulsive efficiency for the BASE ship of 0.73 closely matching both the Hyundai and Samsung numbers. It is also a very reasonable figure, about half way between the Hyundai and Samsung measured numbers. To put it another way, OpenProp does a good job of replicating the performance of the yards propellers. η d is the overall propulsive efficiency, η o η h η r. 4 engine kw is the power required from the engine, and MCR kw is the corresponding installed power, which is 1.15 times engine kw to reflect the design sea margin. EEDI is the resulting Energy Efficiency Design Index per IMO rules. If this column is blank, the engine is not capable of generating the MCR power at the assumed RPM, which is another way of saving the required torque is more than the engine can deliver. The little table extracts the main numbers for the highest speed which generates a legal EEDI. Since the whole analysis was done in 0.25 knot steps, the EEDI associated with this speed can be below the legal max, implying that the actual optimum speed is slightly higher. The little table also shows the average thrust loading which should be no more than 53, to match the BASE ship propeller. We enforced this rule loosely. If the number is above 53, we probably should increase the Expanded Area Ratio (EAR) a bit, resulting in a slight reduction in the calculated calm water efficiency. If the number is below 53, then it may be possible to reduce the EAR, and improve the calculated efficiency slightly. However, since the entire design procedure is based on a deeply immersed propeller in pristine condition, in calm water, all our efficiencies are on the optimistic side. In part, this is dictated by the artifical conditions at which EEDI is to be measured. But we should keep in mind that a real world ship in real world conditions would almost certainly have overall better performance with more conservatively designed propellers. 4 For a direct coupled, VLCC power plant, the amount of power lost in the shaft bearings is negligible. 3

4 Phase 1 Assuming Phase 1 ends up at 10% reduction from the 2.32 baseline EEDI for VLCC s, then to comply with Phase 1, we will need an EEDI of no more than For Phase 1, an engine that matches the allowable installed power is the Sulzer 6rta84t with a normal MCR of 23,585 kw at 76 RPM. The SFC curve for this engine assuming a full set of slow steaming mods is shown below. #6rta84t with full set of slow-steaming mods # but no WHR or orther efficiency improvements , , , , , , , , , , , , , , , , , , As usual, we assumed a 9% reduction in this SFC curve as a result of WHR and other true efficiency requirements. We checked two propeller diameters: 9.93 m and 9.50 m. Since this engine is simply a six cylinder varient of the no-eedi engine, it is no surprise that the optimal diameter remained at 9.93 m, presumably the largest allowed by clearance, immersion (in ballast) and vibration considerations. We allowed OpenProp to adjust the pitch and lift distribution. Table 2 shows the results. Table 2: Diam=9.930 Blades=4 Hull=../hull hhi vee engine=6rta84t wm rpm=76.0 MCR= max eedi= Diameter Speed MCR η d EEDI EAR Loading The 9.93 m alternative got a little unlucky. The required MCR at knots is just above the normal MCR of 23,538 kw. The actual optimum speed is probably around 15.7 knots, so for the Phase 1 comparison runs we will use the full 23,538 kw and an η d of 0.734, slightly better than the MCR propulsive efficiency of the no-eedi ship of In this case, the reduction in thrust required at the same diameter which allowed a narrower blade just outweighed the loss in efficiency due to the reduction in advance ratio. Having said this, an Expanded Area Ratio of for a VLCC is extremely aggressive and untested. Such propellers should be subject to a thorough cavitation, vibration, and strength analysis. This has yet to be done. 4

5 Phase 2 Assuming Phase 2 ends up at 25% reduction from the 2.32 baseline EEDI for VLCC s, then to comply with Phase 2, a VLCC will need an EEDI of no more than For Phase 2, an engine that matches the allowable installed power is the MAN 6S65mc with a normal MCR of 17,220 kw at 95 RPM. The SFC curve for this engine assuming a full set of slow steaming mods is shown below. #Man Tier II program, page 28 with slowstm mods , , , , , , , , , , , , , , , As usual, we assumed a 9% reduction in this SFC curve as a result of WHR and other true efficiency requirements. We ran OpenProp for a prop diameter of 8.0 to 9.0 m in 0.5 meter increments. Table?? shows the results for the 8.5 m prop. Table 3: Diam=8.500 Blades=4 Hull=../hull hhi vee engine=6s65me wm rpm=95.0 MCR= max eedi= Diameter Speed MCR η d EEDI EAR Loading It appears that the optimum diameter is right around 8.5 meters with a propulsive efficiency of 0.68 resulting in a trial loaded speed of a little over 13.5 knots. This diameter is much higher than that assumed in the original EEDI/VLCC paper. Interpolating to get an EEDI of 1.74, we end up with a trial loaded speed of about kts, a propulsive efficiency of 0.682, and an allowable MCR of 16,800 kw. Both the PE and the MCR numbers are a little higher than those used in the original paper which were 0.67 and 16,500 kw. 5

6 Phase 3 Assuming Phase 3 ends up at 35% reduction from the 2.32 baseline EEDI for VLCC s, then to comply with Phase 3, we will need an EEDI of no more than For Phase 3, an engine that matches the allowable installed power is the MAN 6S60mc with a normal MCR of 14,280 kw at 105 RPM. The SFC curve for this engine assuming a full set of slow steaming mods is shown below. #6s60mc with full set of slow steaming mods , , , , , , , , , , , , , , , As usual, we assumed a 9% reduction in this SFC curve as a result of WHR and other true efficiency requirements. We ran OpenProp for a prop diameter of 7.0 to 8.5 m in 0.5 meter increments. Table 4 shows the results for the 8.0 m prop. Table 4: Diam=8.000 Blades=4 Hull=../hull hhi vee engine=6s60me wm rpm=105.0 MCR= max eedi= Diameter Speed MCR η d EEDI EAR Loading It appears that the optimum diameter is about 8 meters. Interpolating to get to an EEDI of 1.51, we end up with a trial loaded speed of about 12.4 knots, a propulsive efficiency of 0.647, and an allowable MCR of 13,240 kw. This diameter is much higher than that assumed in the original EEDI/VLCC paper; but the resulting PE is pretty close to the 0.64 that the original paper used as is the ship speed. The max legal MCR is essentially the same as the 13,200 kw used in EEDI/VLCC. 6

7 Conclusion Table Z summarizes the results of this initial propeller design effort. No EEDI Phase 1 Phase 2 Phase 3 EEDI Limit Max MCR 27,420 23,538 16,800 13,240 Trial speed Prop Diam η d EAR Loading The input for the CO2 emissions study are the allowable MCR and the corresponding propulsive efficiency. For comparison, here are the numbers that were used in the original paper. No EEDI Phase 1 Phase 2 Phase 3 EEDI Limit Max MCR 27,420 23,538 16,500 13,200 η d In hindsight, it is a miracle that the new Phase 2 and 3 numbers are as close as they are to the old. In any event, the Table Z numbers were used in the revised CO2 emissions study. 7