Ice Class Rule Abs

MEDIA EMBARGO UNTIL MARCH 12 This technical paper is embargoed until March 12, 2008 as it will be presented that morning at Technical Session 10, “LNG Shipping: Development, Technology, Crewing, Operations and Trends” during the Gastech 2008 Conference. If you wish to use this material, you must acknowledge it comes from a Gastech 2008 conference paper.

COMBINING ICE CLASS RULES WITH DIRECT CALCULATIONS FOR DESIGN OF ARCTIC LNG VESSEL PROPULSION

Sing-Kwan Lee American Bureau of Shipping

ABSTRACT Ships navigating in ice areas perform quite differently due to ice resistance as compared to their open sea operations. It is a challenge for ship designers to find a design solution which not only optimizes the propulsion performance in open sea but also provides the ship with good ice performance. Traditionally, many ship designers have used rule formulae in Finnish-Swedish Ice Class Rules (FSICR) to estimate the required engine power. While the use of these formulae have proven to be satisfactory thus far, recently increased growth in the activities in oil and gas production in the Arctic area call into question the continued applicability of these Rules for a very obvious reason. The transportation of these oil and gas products requires ice tankers and ice LNG ships to operate in ice conditions much thicker than those previously assumed in FSICR. In such cases, more and more ship designers are turning to direct calculations and model tests for propulsion designs. In this paper, the two most critical design issues in ice ship propulsion will be discussed, namely engine powering and propulsor performance. Comparisons of propulsion designs based on FSICR and on direct calculation are presented for different propulsors. These include fixed pitch, controllable pitch, and ducted propellers for their performance in ice operation. 1. INTRODUCTION Approximate estimates indicate about one third of the world’s known, though not yet developed, reserves of natural gas are in Russia. The overwhelming majority of these reserves are in Artic and Subarctic areas. Russia is not alone in having large natural gas reserves located in Arctic areas, but other countries like Canada and the USA share the same situation with having natural gas reserves in harsh, ice-covered sea environments. As a consequence, the LNG ship technology is moving towards Arctic LNG carriers. New developments in ice navigation and the winterization issues are generating a new challenge for shipping and shipbuilding industries all over the world. 1.1 Design Challenges in Ice Propulsion The economics of LNG transportation by ships from the Arctic are dependent on the efficiency of operation in two areas-- the efficiency of the ship when traveling through the difficult ice conditions and its efficiency in open sea transit. Of special note is the fact that on a direct route from the Arctic to Europe or North America the vessel will spend more than 90% of the time in open sea. In other words, if direct transportation operation is adopted whereby the LNG is loaded on a carrier and transported all the way from the Arctic to its destination, an Arctic LNG vessel is not allowed to sacrifice too much of its open sea performance in order to fulfill its good ice performance. However, designing efficient open sea propulsion and designing good ice propulsion are always in contradiction to each other. In the resistance-propulsion point of view, the contradiction is mainly caused by the fact that high resistance is encountered while operating in low ship speed in ice condition. In such an operation, a propulsor has difficulty performing with its full propulsion capability and efficiency, yet the extreme resistance created by the harsh ice conditions demands that a large thrust be generated enabling the ship to break through the ice. There are many innovative designs proposed to solve this design dilemma, for example, DAT (Double Acting Tanker), podded propulsion, and hull air bubble ejection, to name some of them. Indeed, new design experience continues to grow in industry. Research and development activities have been quite active in recent years in the area of new ice propulsion design. In this study, however, we put more focus on the traditional propulsion concepts such as CPP (Controllable Pitch Propeller) and ducted propeller (nozzle propulsor) instead of the edge-cutting design concepts, while trying to explore more of their potentials in ice propulsion. 1.2 Integrated Design Consideration In ice going ship design, the strengthening of the hull, rudder, propellers, shafts and gears is clearly related to the safety of navigation in ice. In principle, all parts of the hull and the propulsion machinery exposed to ice loads have to be ice-strengthened. Hence, an ice propulsion design cannot be successful with just the use of the propulsion capability, but should include both the thrust and the propulsion power. A comprehensive design should include strength design of the propulsor under ice load. In addition, propeller cavitation and its induced vibration must also be considered. It should be noted that an ice-

Lee 1

strengthened blade design makes it much easier to trigger cavitation problems due to its increased thickness as compared to a non-ice propeller design. For blade strength assessment, finite element analysis is usually requested to be performed based on the ice loads and failure criteria proposed in IACS Polar class rules (IACS, 2008; Lee, 2007). Analyses of cavitation for propeller blade and its induced hull vibration are relied upon more in the direct calculation approach. For more details of the issues refer to (Lee, 2006). 2. ICE RULES FOR SHIP POWERING 2.1 Power Requirement in Ice Rules Generally, FSICR (Finnish-Swedish Ice Class Rule) is the accepted Rules used for vessels trading in the Northern Baltic Sea in winter. The Rules primarily address matters directly relevant to the capability of ships to advance in ice. Ships are required to have a certain speed, 5 knots, in a brash ice channel in order to ensure the smooth progress of traffic in ice conditions. The required minimum speed, 5 knots, is based on the maximum waiting time, 4 hours, for icebreaker assistance. According to Section 3 in FSICR (2002), for ships entitled to ice class IA Super, IA, IB, or IC, the engine output is to be not less than determined by the following formula.

P = KC

( RCH / 1000)3 / 2 Dp

(1)

where, KC is a propeller-related constant factor, Dp is propeller diameter and RCH is resistance, which can be calculated by the following rule formula. 3 RCH = C1 + C2 + C3 C (HF + HM)2(B + CØHF) + C4LPARHF2 + C5 ⎡ LT ⎤ AWL (2) ⎢ 2⎥ ⎣B ⎦

L

Detailed meanings of the symbols used in the formula can be found in Section 3.2.2 of FSICR and are not repeated here. When using equation 2, the following restriction is to apply: 3

⎡ LT ⎤ 20 ≥ ⎢ 2 ⎥ ≥ 5 ⎣B ⎦ where L is length of perpendiculars, B is maximum breadth, and T is draught. In fact, this restriction constrains the use of the engine output formulae for large size ship. For Arctic ships, although IACS Polar Rule for Machinery Requirement, URI3, will soon be published (IACS, 2008), the powering for Arctic navigation is not included. 2.2 Difficulties in the use of FSICR Difficulties in using FSICR for a rational design of propulsion power are learned by shipyards and classification society such as ABS when applying FSICR to large ships (Bezinovic, 2005, Kim et al., 2005, Lee, 2006). For a FPP (Fixed Pitch Propeller) design for an Aframax IA ice class tanker, it is found that the engine power can be largely reduced based on a more accurate direct calculation for propulsor thrust and model test results for ice resistance. To illustrate the general idea of the conservativeness of FSICR, Figure 1 shows a comparison done by an engine manufacturer (MAN B&W, 2002, 2006) for the non-ice class tankers’ engine powering and FSICR requested engine powering for ice tankers. As seen, the Rule requests an Aframax size IA ice class tanker to install the engine power with a non-ice VLCC tanker. It not only causes the extensive increase in cost in building the ice tanker but also creates an extremely difficult situation in trying to fit the huge engine into an Aframax-size engine room. In fact, in determining the ice propulsion powering, ice resistance and propulsor performance are two dominant factors, which are related to the coefficients, RCH and KC in FSICR (see the formula (1)). The

Lee 2

coefficients are empirically estimated based on average ice resistance and propeller performance in FSICR. With the continuously, ever-increasing introduction of new ice ship designs such as the new hull and bow forms designed to reduce ice resistance as well as more efficient propellers, FSICR ice powering formulae have come into a questionable review on the current validity of the rationale.. Furthermore, in the case of these Arctic ships, it has not been verified that the ice resistance and ice powering formulae in FSICR are applicably suited to meet the demands of thicker multi-year ice propulsion.

1a) Engine power demand for non-ice tankers

1b) Engine power demand for Finnish/Swedish ice classed vessels – based on FSICR calculations Figure 1 Ice and non-ice tanker engine power comparison

3. ICE POWERING DETERMINATION BASED ON DIRECT CALCULATION A comprehensive design for ship ice propulsion involves many different elements such as hull resistance, propulsor design, engine character, shafting/gear/bearing lubrication and friction loss, CP mechanism operation, and sometimes propeller cavitation induced by machinery vibration etc.. Although all of them are more or less influencing the finial ice propulsion performance, ice resistance, propulsor performance and engine character are the three most critical areas in ice powering determination. They are addressed briefly in the following sections. 3.1 Resistance In addition to the concern of open sea resistance when working on ice propulsion design, ice resistance is another factor to be considered for power determination. Ship resistances in ice and open sea can be estimated by numerical simulation or model test measurement. As CFD simulations have become more and more practical for the study of open sea resistance in recent years (MARNET, 2002), it is anticipated that these simulations will play an increasingly more important role in resistance prediction. However, in the current stage, CFD is still state-of-the-art and used as a compromising method with model test results for resistance prediction. For ice resistance simulation, some progress has been made in numerical simulation for level ice resistance (Valato, 2001) but the brash ice resistance under waterline is still not certain and needs to use the semi-empirical model of Lindqvist (1989). In general, model tests are recommended by the Guidelines of FSICR (2005) for ice resistance assessment, especially when vessel displacement is greater than 70,000 tonnes. As model test uncertainty is always an issue, it should be noted that the difference in ice modeling and measurement procedure may come up with a large discrepancy on ice resistance. To assist designers in performing ice model test, some aspects of the ice model tests such as the geometry of the ice channel, ice friction coefficient, determination of the propulsion power in full scale, and model test documentation are

Lee 3

addressed in the Guidelines of FSICR (Appendix 4). Ice resistance can usually be expressed as a quadratic curve.

Rice = d0 + d1VS + d2VS2

(3)

where Rice is ice resistance, VS is ship speed and d0, d1, and d2 are curve fitting constants. Resistance tests usually are performed based on experiences and calibrations in different towing tanks. ITTC also recommends some standard procedures for resistance test and performance prediction method (ITTC, 2000 and 1999). If open sea resistance for designed speed is obtained, resistance at other speeds can be calculated based on cubic law as follows:

ROW = c1VS + c2VS2 + c3VS3

(4)

where ROW is open sea resistance, VS is ship speed and c1, c2, and c3 are curve fitting constants, which can be determined based on resistance and ship speed measured in model tests. Figure 2 is the typical outlook of Rice and Row curves. As seen, ice resistance usually has higher value than open sea resistance under the same ship speed. However, the increasing rate of the ice resistance curve is smaller than the open sea resistance curve. It is also the reason behind the use of a quadratic curve for ice resistance instead of a cubic curve.

Figure 2 Typical ice and open sea resistance curves 3.2 Propulsor Performance As is well known, propeller thrust and torque are dependent on its geometry and operating condition (rpm and inflow velocity toward propeller). They can be expressed as non-dimensional forms as follows:

K T = a1 + a 2 J + a3 J 2 + a 4 J 3

(5a)

K Q = b1 + b2 J + b3 J + b4 J

(5b)

2

Lee 4

3

where KT is thrust coefficient defined as T/ρn2D4 and KQ is torque coefficient as Q/ρn2D5, in which T and Q are thrust and torque. J is advanced coefficient defined as V/nD; ρ, n, V, and D are fluid density, propeller rps, inflow velocity, and diameter. For a specific design, {a1 a2 a3 a4} and {b1 b2 b3 b4} are constant vectors. In additional to propeller open water model test, KT and KQ curves can also be constructed by CFD simulation (Chen and Stern, 1999, Chen and Lee, 2005) or regression formulae (for standard series propellers, see Oosterveld and Ossanen, 1975). In general, equations 5a and 5b can be extended to any higher degree polynomial based on the propeller open water measurement data, however, around the range [0, 1] of J, third order polynomial usually provides sufficient accuracy for KT and KQ values. This can be seen obviously from the typical pattern of KT and KQ curves shown in Figure 3 for both model and full scale Reynolds number Rn.

Figure 3 Typical propeller performance curves (KT and KQ) 3.3 Ship/Engine/Propeller Interaction Propeller absorbing power (see propeller demand curve shown in Fig. 4) is usually limited by engine power curve (see torque limit line in Fig. 4). For an appropriately designed propeller, it should absorb all engine power at MCR rpm when advancing with the design ship speed. For Controllable Pitch Propeller CPP), it can adjust its pitch to keep its original hydrodynamic performance at different ship speed. Thus, CPP can approximately maintain its rpm as the MCR engine speed and absorb the MCR engine power at different ship speed. However, for Fixed Pitch Propeller (FPP), when ship speed becomes slower (operating in ice condition), in order to maintain the rpm as MCR speed, the propeller needs to be supplied with more power from the engine to overcome the higher hydrodynamic torque on the blade compared to the original ship speed condition. This leads to the original propeller demand curve ‘shifting up’ to curve A as shown in Fig. 4. As seen in Fig. 4, due to the limitation of the engine torque limit curve, when the propeller operates off its design condition (design ship speed along with MCR rpm), the engine can only provide the power along the torque limit curve and accordingly the propeller will rotate slower (see point x of curve A in Fig. 4). At point x, engine power P is equal to propeller absorbing power, i.e., Pp = 2πnKQ ρn2D5. Using equation 5b and definition of J, we have

Lee 5

2

3

V ⎛V ⎞ ⎛V ⎞ P = 2πρD [b1 n + b2 n 2 + b3 ⎜ ⎟ n + b4 ⎜ ⎟ ] D ⎝D⎠ ⎝D⎠ 5

(6)

3

Usually, engine power along the torque limit line follows the following relationship with rpm (MAN B&W, 1994): P ∝ rpm2 (7a) or more precisely

P = PMCR (

n nMCR

)2

(7b)

where P = engine power corresponding to rotation speed n PMCR = engine power corresponding to MCR rotation speed nMCR Based on (7b) and (6), if propeller rotation speed n at off design operating condition is given, the engine MCR power can be determined by the following equation.

⎛ n PMCR ⎜⎜ ⎝ n MCR

2

⎞ V V V ⎟⎟ = 2πρD 5 [b1 n 3 + b2 n 2 + b3 ⎛⎜ ⎞⎟ n + b4 ⎛⎜ ⎞⎟ ] D ⎝D⎠ ⎝D⎠ ⎠ 2

3

(8)

Figure 4 Engine and propeller demanded curve For ship under ice operating condition, the determination of actual propeller rotation speed n in equation (8) is related to the balance between propulsor thrust and ice resistance. Details are given in the following.

Lee 6

When the ship operates in ice condition, the propeller thrust generated should be used to overcome both hydrodynamic resistance (or open sea resistance) and ice resistance, i.e., T ≥ ROW + Ri = Rice, where T is propeller thrust, ROW is open sea resistance, and Ri is net ice resistance, and Rice is the total resistance under ice navigation. In other words, T should be large enough to overcome the total resistance, Rice. Accordingly, we have

T ≥ Rice ⇒ ρn 2 D 4 K T ≥ d 0 + d 1V S + d 2V S2 ⇒ ρn 2 D 4 [a1 + a 2

V V V + a 3 ( ) 2 + a 4 ( ) 3 ] ≥ d 0 + d 1V S + d 2V S2 nD nD nD

(9)

It should be noted that V and VS in equation (9) represent different physical meanings. The former is the actual velocity toward the propeller and VS is the ship speed. Usually, V can be roughly estimated as VS (1-w), where w is a wake fraction due to hull boundary layer effect. In equation (9), the thrust reduction effect is not considered as it becomes smaller when the ship is in slow speed. Based on equation (9), if total resistance Rice, ship velocity VS, wake fraction w, are given, the actual propeller speed n under ice condition can be calculated. 3.4 Procedure for Power Determination In summary, the overall procedure for determining required engine power is as follows: Perform model test and measure the open sea resistance at design ship speed Construct total resistance curve – equation (3) based on ice resistance model test results Construct propeller KT and KQ curves – equations (5a) and (5b) For FPP, use equation (9) to determine the actual propeller rotation speed n; for CPP, n is same as the engine MCR rotation speed but propeller performance curves (KT and KQ) will be based on those at new turning position (more details for CPP will be illustrated in the Case Study section). Use equation (8) to determine the required engine MCR power. 4. CASE STUDY The aforementioned powering determination procedure is used in this section for an Arctic LNG carrier. Different propulsor designs, namely FPP, CPP, and ducted propeller, are compared for their propulsion powers for this LNG carrier in Baltic ice sea and Arctic sea operations. 4.1 Twin Screw LNG Vessel LNG ships have traditionally been a single screw ship driven by a steam turbine. The size of the steam turbine plant is relatively large and the turbine is expensive both to manufacture and to maintain. Therefore, a twin steam turbine design has not been an attractive solution. New and improved techniques have resulted in more efficient reliquefication plants to take care of the LNG boil off and more efficient dual fuel engines. The expansive steam turbine is no longer the only choice for the LNG ship. An Arctic LNG Shuttle with twin screws (Figure 5) and the following design parameters are selected for this propulsion power comparison study. In this study, the selected LNG ship is assumed to be driven by two Diesel engines. Length = 200 m Breadth (water line) = 28 m Draft (max.) = 9 m Cargo capability = 40, 000 m3

Lee 7

Figure 5 A twin screw Arctic LNG Shuttle (after http://www.hoegh.com) According to the information posted in the internet site (http://www.hoegh.com), the LNG shuttle is designed to transport LNG through areas with severe ice, such as the Russian Arctic. It has the capability to operate independently (without ice breaker escort) in up to 3.0 meter nominal ice thickness. Model tests for ice resistance and open sea resistance are conducted for this shuttle. The ice thickness range for ice resistance is from 0.6 m to 2.4 m with ship speed range between 0 to 18 knots. The resistance curves are reproduced in Figure 6. Later, these resistance curves will be used to estimate the requested engine power and propulsor operation conditions such as rpm for FPP and turning angle for CPP under different ice conditions.

Figure 6 Ice and open sea resistance curves In Figure 6, Baltic ice classes – IA, IB and IC are indicated as the curves corresponding to the ice thickness Hice = 0.6m, 0.8m and 1.0m. Due to the limited data releasing in the public internet site, the ice thickness for the ice resistance curves is only up to Hice = 2.4m although the ship is designed to operate with ice thickness up to 3 m. It is noted that ice resistance increases rapidly with the ice thickness and compared to the open sea resistance ROW, the net ice resistance Ri dominates the total resistance Rice. To have a general idea of percentage for different resistances, the comparison among ROW (open sea

Lee 8

resistance), Ri = Rice – ROW (net ice resistance), and Rice (total resistance) under ship speed 5 knots is summarized as follows: Ice thickness Hice 0.6 m (IC class) 2.0 m

Ship speed VS 5 knots 5 knots

Open sea Resis. Row 157.0kN 157.0 kN

Total Resistance Rice 1214.4 kN 5000.0 kN

Net ice Resistance Ri = Rice – ROW 1057.4 kN (87% of Rice) 4843.0 kN (97% of Rice)

4.2 Powering for Baltic Ice Sea Navigation As explained in section 3, in determining the engine power for ice navigation a sequence of calculations needs to be performed based on the information of ice/open sea resistance, propulsor performance and engine character. To initiate this calculation procedure, propeller diameter and its design rpm need to be determined first. For the selected LNG carrier, a propeller (including FPP, CPP and Ducted propeller) with 6 m diameter and 102 rpm is a reasonable choice. For rotation rpm = 102, one option of the engine can be the following. MCR rpm =102 Cylinder. number MCR power in kW 5 5825 6 6990 7 8155 8 9320 Table 1 A example engine for MCR rpm = 102 4.2.1 Fixed Pitch Propeller Design In order to compare the engine power requested due to the performance of FPP and CPP, the same propeller geometry – AU-CP 4 blades EAR = 55% – is used for both FPP and CPP cases. Figure 7 is an outline plot of the propeller blade with two different EAR 55% and 70%.

Figure 7 AU-CP Propeller geometry - 4 blades with EAR 55% and 70% Lee 9

4.2.1.1 Propeller curves – Kt and Kq In FPP case calculation, although the propeller is a CPP design, propeller turning angle is fixed as θ = 0o, which is corresponding to the pitch ratio P/D = 1.0. The propeller performance curves, KT , KQ, and efficiency η, can be found from the pervious model test results (Yazaki, 1964, Yazaki and Sugai, 1966). For design pitch ratio P/D = 1.0 under θ = 0o position, the KT and KQ curves (Figure 8) can be expressed as follows:

K t = 0.47 − 0.34 J − 0.136 J 2 + 0.034 J 3

(10)

10 K q = 0.69 − 0.46 J − 0.137 J 2 + 0.0012 J 3

(11)

Figure 8 Kt, Kq and η for AU-CP propeller (4 blade, EAR = 55%) 4.2.1.2 Ice resistance for IA, IB and IC classes Under 5 knots speed (FSICR requirement), the ice resistances for IA, IB and IC class can be estimated using the model test results shown in Figure 6. The following table summarizes the results of the total ice resistance curves (based on 2 order curve fitting) and the value of Rice at ship speed Vs = 5 knots. Ice resistance curve IA: Hice = 1.0 m

Rice = 1114 .8 + 191 .1V S + 3.5V S

2

IB: Hice = 0.8 m

Rice = 846.3 + 143.7V S + 4.19V S

2

IC: Hice = 0.6 m

Rice = 592.4 + 103.9V S + 4.1V S

2

Table 2 Ice resistance for IA, IB and IC classes

4.2.1.3 FPP rpm under ice operations

Lee 10

Rice at 5 knot speed 2157.8 kN 1669.55 kN 1214.4 kN

To determine the actual rpm for the FPP under ice operation, equation (9) is used for the three scenarios of IA, IB and IC ice resistance. Since a twin screw design is used, the balance of one propeller thrust should be equal to half of the Rice, i.e., 0.5Rice = ρn2D4Kt. Therefore, we have (see also Figure 9) IA: IB: IC:

1 2157.8 / 2. = 1078.9 = 0.369[0.47 n 2 − 7.49n − 68.24 + 364.01 ] n 1 1669.55 = 834.775 = 0.369[0.47 n 2 − 7.49n − 68.24 + 364.01 ] n

1 1214.4 = 607.2 = 0369[0.47 n 2 − 7.49n − 68.24 + 364.01 ] n

(12a) (12b) (12c)

By solving the equations (12a), (12b) and (12c), the rpms for the AU-CP FPP under IA, IB and IC ice conditions are obtained as follows: Class IA IB IC

Actual rpm of FPP 0.5 Rice 1078.9 kN 88.0 834.7 kN 79.0 607.2 kN 69.0 Table 3 Actual rpm of FPP

Figure 9 Thrust and Rice curves 4.2.1.4 Engine power for ice navigation Once the actual propeller rpms under the ice operations are known, the engine power needed to reserve in MCR speed can be calculated based on equation (8). For the MCR rpm = 102, the requested engine powers for IA, IB and IC classes are 11.98 MW, 10.48 MW and 8.82 MW. The details of the calculations for these engine power values are summarized in the following table.

Lee 11

Class

n actual rpm

IA IB IC

88.0 79.0 69.0

J

10Kq (eq. 11)

Propeller absorbing power P = 2πρn3D5KQ

0.25 0.566 8.942 MW 0.28 0.550 6.287 MW 0.32 0.530 4.036MW Table 4 Engine powers for IA, IB and IC classes

nMCR= 102 rpm (nMCR /n)2 PMCR (eq. 8) 1.34 11.982 MW 1.667 10.48 MW 2.185 8.819 MW

It is also interesting to compare these direct calculation results with FSICR calculations to see how much reduction can be obtained based on the current approach. In FSICR, the engine power is calculated based on equation 1. For the FPP twin case, Kc is taken as 1.6. The following is a table summarizing the comparison. Class Rch FSICR Power IA 2157.8 kN 26.729 MW IB 1669.55 kN 18.191 MW IC 1214.4 kN 11.285 MW Table 5 FSICR requirement for engine powers for IA, IB and IC classes Through this comparison, it is found the engine powers requested in FSICR for this FPP case are 223% of the power from the direct calculation for IA class, 174% for IB class and 128% for IC class. 4.2.1.5 Open sea performance To ensure this FPP and engine powering combination to be suitable for open sea operation, the propeller performance needs to be checked at the MCR condition in open sea. In this study, the MCR condition assumes that a full speed of the LNG is 17 knots under a propeller rotation speed = 102 rpm. Under the 17 knots speed, the flow velocity in front of the FP propeller is estimated to be 14.57 knots. Here, the wake factor is taken as 0.143 (Manen and Oossanen, 1988) based on the LNG ship block coefficient range, 0.7 ~ 0.8. Accordingly, the advance coefficient J is calculated to be 0.73 for rpm = 102. From Figure 8 (propeller performance curves), we notice the FPP selected is operated in good efficiency (see the η curve). The Kt and Kq values under the J = 0.73 can be obtained from equations (10) and (11). Based on the Kt and Kq values calculated, the thrust, T, and propeller absorbing power, P, can be determined. The details are summarized in the following table: rpm 102

J 0.73

Kt Kq T = ρn2 D4Kt Q = ρn2 D5Kq 0.16 0.0278 614.24 kN 640.36 kNm Table 6 Propeller thrust and absorbing power at open sea operation

P =2πnQ 6.840 MW

Based on the propeller thrust value, the twin propulsion thrust is 1228.48 kN. In open sea operation, the resistance R at 17 knots can be calculated by the following equation. It is about 1192.55 kN.

R = 25.44V + 0.59V 2 + 0.12V 3

(13)

In summary, the selected FP propeller can generate enough thrust to propel the LNG ship with 17 knots speed. The requested engine power is in the level for the ship operating in 5 knots in IB class ice condition (see table 4). Basically, the engine powers (IA, IB and IC classes) requested to reserve in MCR condition are large enough for open sea condition (see Table 4). 4.2.2 Controllable Pitch Propeller Design Unlike FPP, CPP can change its pitch by turning the blade in certain angle Δθ (Figure 10) to obtain more thrust for propulsion. Usually the propeller rpm can be maintained as the MCR speed even at low ship speed in ice navigation. These advantages make CPP more capable and efficient in ice operation.

Lee 12

Figure 10 Turning angle Δθ for CPP to adjust its pitch 4.2.2.1 Blade turning angles at ice operations To determine what θ should be turned for the blade to generate enough thrust to overcome ice resistance, propeller performance (Kt and Kq) curves at different turning angles should be obtained first. Figure 11 shows the thrust, torque and efficiency curves for the AU-CP propeller at different blade turning angles (Yazaki, 1964).

11b) Kq and η at different θ 11a) Kt and η at different θ Figure 11 Thrust, torque and efficiency curves for the AU-CP propeller at different θ To interpolate the turning angle value from the above J-θ-Kt curves, the operation point (J-Cice) for ice resistance should be determined first. Here Cice is the ice resistance coefficient non-dimensionalized by 2ρn2 D4; here factor 2 is taken for twin screw propulsion consideration. For the three Baltic ice classes, IA, IB and IC, their values are given in the following table. In Cice calculation, the ice resistance formulae in Table 2 are used for the ice resistance Rice calculations. The J is calculated based on inflow velocity in front of the propeller which takes into account the wake factor w = 0.143. Class IA IB IC

rpm

J

Rice/2 1078.9 kN 102 0.216 834.8 kN 607.2 kN Table 7 Cice values for IA, IB, and IC ice classes Lee 13

Cice 0.281 0.217 0.158

12b Kq values for IA, IB and IC classes 12a J-Cice points for IA, IB and IC classes Figure 12 CPP turning angle interpolations In Figure 12a, the J-Cice points for IA, IB and IC classes are plotted in the J-θ-Kt curves. Through interpolating the Cice values into the Kt values, the turning angles θ can be calculated. They are -5.09o for IA class, -8.27o for IB class, and -11.48o for IC class. Based on these θ values, Kq values corresponding to the θs are calculated and plotted in Figure 12b. 4.2.2.2 Engine power for ice navigation The engine powers required for the ice classes, which are equal to the propeller absorbing powers, are calculated based on the Kq values obtained, i.e. P = 2πρn3D5Kq. They are summarized as follows: Class IA IB IC

n CPP rpm

J

10Kq

θ

Propeller absorbing power P = 2πρn3D5KQ

-4.18o 0.38372 8.632 MW o 102.0 0.216 -7.44 0.2685 6.040 MW -10.72o 0.17615 3.963 MW Table 8 Engine required powers for IA, IB and IC classes for CPP designs

As done in the pervious FPP design case, FSICR formulae are used to check the Rule required engine powers for this CPP design. For a twin CPP propeller, Kc value is taken as 1.44 for equation 1. Comparisons between FSICR and the direct calculations are given in the following table. Class

Rch

FSICR Power

% the ratio of FSICR power to direct calculation power IA 2157.8 kN 24.056 MW 279 % IB 1669.55 kN 16.372 MW 271 % IC 1214.4 kN 10.156 MW 256 % Table 9 Power comparison between FSICR and direct calculations

For open sea operation, since the design is same as the FPP case, the engine power for the LNG operates in 17 knots is with the same value as in the table 6, i.e. 6.84 MW, which is 79.2% of the IA class engine power in this CPP design. 4.2.3 Ducted Propeller Design It is well-known that a ducted propulsor has better propulsion capability when operating in slow speed compared to an open propeller. This is not only due to the better design of the propeller at low J operation but also due to the contribution of the thrust generated by the duct (see Figure 13). Lee 14

Figure 13 Thrust coefficients of duct and propeller for a ducted propeller. As seen in Figure 13, a typical thrust generated by the duct is more than half of the propeller thrust at low J condition (J = 0.1). However, it also needs to be noted that a duct will become a disadvantage when a vessel navigates in high speed. As seen in Figure 13, thrust over J = 0.6 from the duct becomes negative, i.e. extra resistance is generated by the duct. Later, in this case study it will be shown that the use of the ducted propeller can create the problem whereby the LNG carrier cannot reach a 17-knots speed in open sea operation although it does reduce quite a lot of the engine power required for ice propulsion. 4.2.3.1 Ducted propeller 19A Ka-4-70 A fixed pitch ducted propeller selected in this study is a standard design of a 19A nozzle with a Ka 4-70 propeller (4 blade with EAR = 70%). In order to compare the final engine power results with the pervious open propellers under an approximate base, the same propeller diameter, 6 m, as the open propeller is adopted to this ducted propeller. Figure 14 shows the outline of the 19A Ka-4-70 ducted propeller (blade and nozzle profile).

Figure 14 Geometry of ducted propeller 19A Ka 4-70. 4.2.3.2 Ducted propeller curve – Kt and Kq

Lee 15

Figure 15 Kt, Kq and η for 19A Ka 4-70 ducted propeller (4 blade, EAR = 70%) The propeller performance curves of this ducted propeller are available from the well-known model tests done by van Gent and Oosterveld (1983). Figure 15 shows the performance curves. It should be noted that the Kt curve represents the total thrust from propeller and duct. Similar to the pervious open propeller, the Kt and Kq curves can be expressed as cubic curves. Their mathematical forms are as follows:

K t = 0.526 − 0.599 J + 0.285 J 2 − 0.391J 3

(14)

10 K q = 0.443 − 0.0124 J − 0.285J 2 − 0.116 J 3

(15)

4.2.3.3 Operating rpm determination Since this ducted propeller is a fixed pitch design, propeller rotation speed at ice operation will be dropped due to the heavy loading condition encountered. Based on the thrust-ice resistance balance principle (equation 9), the ice operation rpms for IA, IB and IC class can be determined through the similar procedure as the FPP case. The solutions are summarized in the following table. Class IA IB IC

Actual rpm of FPP 0.5 Rice 1078.9 kN 87.0 834.7 kN 78.0 607.2 kN 68.5 Table 10 Actual rpm of ducted propeller

To crosscheck the solutions, a plot for the ice resistances (IA, IB, and IC classes) and the ducted propeller thrust curve is drawn (Figure 16). The intersection points of the thrust curve and the ice resistance lines provide the actual rpm values. As seen, the rpm values of the intersection points are close to the values in the Table 10.

Lee 16

Figure 16 Thrust and Rice curves 4.2.3.4 Engine power for ice navigation Based on the actual rpms in table 10, the engine powers needed to reserve for MCR condition can be determined based on the equation (8). The details are summarized in the following table. Class

n actual rpm

IA IB IC

87 78 68.5

J

10Kq (eq. 15)

Propeller absorbing power P = 2πρn3D5KQ

0.255 0.419 6.397 MW 0.285 0.413 4.544 MW 0.324 0.404 3.010 MW Table11 Engine powers for IA, IB and IC classes

nMCR= 102 rpm (nMCR /n)2 PMCR (eq. 8) 1.37 8.764 MW 1.71 7.770 MW 2.22 6.682 MW

Similar to the FPP case, a comparison of these direct calculations with FSICR calculation is preformed. In FSICR, engine power is calculated based on equation 1. For a FPP twin screw, Kc is taken as 1.6. For a ducted propeller, a further 70% reduction from the open propeller result can be applied according to the ‘Guidelines for the Application of FSICR’. The following is a table summarizing the comparison. Class

Rch

FSICR Power Power for ducted propeller Open propeller 70% power of open propeller IA 2157.8 kN 26.729 MW 18.71 MW IB 1669.55 kN 18.191 MW 12.73 MW IC 1214.4 kN 11.285 MW 7.90 MW Table 12 FSICR requirement for engine powers for IA, IB and IC classes

From Table 11 and Table 12, it is found that FSICR requests 213% engine power of the value from direct calculation for IA class; 164% for IB class; and 118% for IC class.

Lee 17

4.2.3.5 Open sea performance Although the ducted propeller has good performance in ice operations, since its performance is usually bad in fast speed, a crosscheck of its open sea performance is necessary. Assuming that 17 knots is the design full speed and 102 rpm engine is used same as before, the thrust and the engine power can be calculated based on the approach used in the open FPP case. Table 13 summarizes the details of the calculations. rpm 102

J 0.73

Kt Kq T = ρn2 D4Kt Q = ρn2 D5Kq 0.08 0.02366 337.8 kN 555.0 kNm Table 13 Propeller thrust and absorbing power at open sea operation

P =2πnQ 5.821 MW

As known earlier, at 17 knots speed, the ship resistance is 1192.55 kN. From Table 13, the twin screw propulsion from this ducted propeller design is 675.6 kN which is almost 50% smaller than the requested thrust. In other words, the LNG ship can not reach the 17 knots speed by this propulsor design. It should be noted that the problem is not from the engine power but mainly due to the available thrust at the full speed operation. In fact, the engine requested power in this case is smaller than the open FPP case. To estimate the ship speed for this ducted propeller design, the thrust curve for 102 rpm rotation speed based on 2×Kt curve (2.0 × equation 14 – a twin crew propulsion) and the open sea resistance curve (equation 13) are plotted in a same diagram (Figure 17). At the intersection point of the two curves, the propeller thrust becomes equal to the open sea resistance and the ship speed at the point is 12.5 knots. For this speed, the engine power requested is about 8.394 MW for each ducted propeller. Details of the calculations are summarized in the following table. Rpm 102

Vs Thrust per propeller Torque per propeller Power per propeller 12.25 knots 620 kN 785.82 kN-m 8.394 MW Table 14 Thrust, torque and power for open sea operation

Figure 17 Propeller thrust and open sea resistance curves 4.3 Powering for Arctic Sea Navigation To illustrate how the ice powering rapidly increases with the ice resistance, the powering for Arctic ice, which is with thicker ice thickness than Baltic ice class, is also studied. As the ice resistances in thick ice conditions are extremely high and if the FSICR ship speed 5 knots is used, it may come up that the ice powering is irrationally high to select a reasonable size engine to fit into an engine room. Hence, in determining the Arctic ice powering, a slower speed 2.5 knots is used as the operating speed for the Arctic ship design. Basically, all the calculations are the same as that were previously performed for the Lee 18

FPP, the CPP and the ducted propeller. The final results for the FPP, the CPP and the ducted propeller for the ice thickness range from 1.2 m to 2.4 m are summarized in the following tables. Hice (m) 1.2 1.4 1.6 1.8 2.0 2.2 2.4

Total Rice (kN)

Propeller rpm

Propeller absorbing power (MW)

nMCR = 102 rpm (nMCR/n)2 PMCR (MW)

1994.69 80.25 7.465 1.62 12.093 2456.41 88.6 10.142 1.33 13.489 2905.26 95.7 12.883 1.14 14.629 3400. Required propeller rpm > 102; this cause the engine 103.3 being overloaded to generate the thrusts for 2.5 3950. 111.0 knots propulsion in 1.8 m upper ice thickness.. 4450. 117.6 5000. 124.2 Table 15 Engine power for 2.5 knot Arctic sea navigation – FPP design

nMCR = 102 rpm; PMCR (MW) CPP turning θ Total Rice (kN) Hice (m) o 7.001 -7.9 1994.69 1.2 o 9.399 -5.0 2456.41 1.4 o 12.685 -2.4 2905.26 1.6 o 16.554 0.5 3400. 1.8 o 21.946 3.9 3950. 2.0 o 27.964 6.9 4450. 2.2 o 35.208 10.2 5000. 2.4 Table 16 Engine power for 2.5 knot Arctic sea navigation – CPP design nMCR = 102 rpm Total Rice Propeller Propeller absorbing Hice 2 rpm power (MW) PMCR (MW) (n (kN) MCR/n) (m) 77.98 4.782 1.71 8.177 1994.69 1.2 85.77 6.378 1.41 8.993 2456.41 1.4 92.74 8.081 1.21 9.778 2905.26 1.6 99.82 10.1 1.04 10.50 3400. 1.8 107.11 3950. Required propeller rpm > 102; this cause the engine 2.0 being overloaded to generate the thrusts for 2.5 113.22 4450. 2.2 knots propulsion for 2.0 m upper ice thickness 119.73 5000. 2.4 Table 17 Engine power for 2.5 knot Arctic sea navigation – Ducted propeller design According to these case studies, it has been found that the ducted propeller is the most energy-saving design for ice operations for Arctic sea propulsion in the high ice resistance (Hice = 1.4 – 1.8 m ). However, due to the limitation of its fixed pitch design, the ducted propeller cannot operate in the condition with ice thickness thicker than 2 m. For the CPP design, although the power-saving capability is not as good as the ducted propeller, the propulsion capability in ice is unlimited for any ice thickness providing that enough power can be supplied. As expected, the FPP design has the worst performance in ice navigation as it needs the largest amount of power among the other designs for the same ice thickness, and it becomes unable to operate in the ice thickness equal to 1.8 m and upwards no matter if the engine power continues to increase. However, it should be noted that for the open propeller design, both CPP and FPP, the open sea performance is much better than the ducted propeller. 5. CONCLUDING REMARKS Driven by the large reserves of oil and natural gas and their exploitation in Arctic and sub-arctic areas, ship propulsion in ice-covered sea becomes a critical concern. The design dilemma of ice propulsion

Lee 19

performance against open-water performance makes ice going ship design more challenging. In this paper, commonly used propulsors, namely FPP, CPP, and the ducted propeller, were investigated to compare their performance in ice propulsion conditions. A detailed procedure based on direct calculation for determining propulsion thrust and powering was documented. To illustrate how to process the calculation procedure to obtain the required engine MCR power, an Arctic LNG carrier with ice resistance information available was used as an illustrative example. The overall results of powering for the FPP, the CPP and the ducted propeller for the LNG vessel are summarized again in the following table to highlight the findings in this study. Ship speed Vs Hice (m)

17 kts Open sea

IB 0.6

5. kts IC 0.8

IA 1.0

1.2

1.4

1.6

2.5 kts 1.8

2.0

2.2

2.4

1192.6

1214.4

1669.5

2157.8

1994.7

2456.4

2905.3

3400.0

3950.0

4450.0

5000.0

PMCR

6.8

8.8

10.5

12.00

12.1

13.5

14.6

x

x

x

x

Pio rpm PFSICR

NA

4.0 69

6.3 79

8.9 88

7.5 80

10.1 89

12.9 96

x 103

x 111

x 118

x 124

NA

11.3

18.2

26.7

23.8

32.5

41.8

52.9

66.2

79.2

94.3

PMCR

6.8

4.0

6.0

8.6

7.0

9.4

12.7

16.6

21.9

28.0

35.2

Pio θ

NA

4.0 o -11.5

6.0 o -8.3

8.6 o -5.1

7.0 o -6.9

9.4 o -3.9

12.7 o -1.1

16.6 o 2.1

21.9 o 5.6

28.0 o 8.8

35.2 o 12.2

PFSICR

NA

10.2

16.4

24.1

21.4

29.3

37.6

47.6

59.6

71.3

84.9

PMCR

x

6.7

7.8

8.8

8.2

9.0

9.8

10.5

x

x

x

Pio rpm PFSICR

NA

3.0 69

4.5 78

6.4 87

4.8 78

6.4 86

8.1 93

10.1 100

x 107

x 113

x 120

NA

7.9

12.73

18.7

16.7

22.8

29.3

37.0

46.3

55.4

66.0

Resistance (kN)

FPP Power (MW)

CPP Power (MW)

Nozzle Power (MW)

Table 18 Comparison of powering for FPP, CPP, and Ducted propeller based on FSICR and direct calculation In the Table 18, both the direct calculations and FSICR calculations are included. PMCR represents the final engine power required to be installed; Pio is the power required under different ice operation conditions, for instance, ice operation conditions of IC class are ship speed 5 knots and 0.6 m first year ice; PFSICR is the power calculated based on the FSICR. The FSICR calculations are extended to higher ice condition beyond the FSICR classes, in which ice thickness is 1.0m. In principle, FSICR can be applied if ice resistance and propeller size (diameter) and type are known (see equation 1). The rpms provided in the table for the FPP and the nozzle (the ducted propeller) are the propeller rotation speed at ice operation. For the value θ, it is the turning position of the CPP blade at ice propulsion. As well-known, the CPP can maintain its MCR rotation speed even in high load condition under ice operation. The symbol “x” means the propeller design cannot be applied. For example, “x” in the table for PMCR of the ducted propeller (nozzle) means the open sea speed 17 knots cannot be fulfilled due to the shortage of the thrust generated by the ducted propeller. And “x” in the FPP for Pio means no 102 rpm engine can be used to fulfill the ice propulsion for the ice conditions – Hice > 1.8 m, due to the rpms requested are higher than 102 in that ice condition. From Table 18, the concluding remarks of this study are summarized as follows: Design dilemma of ice propulsion against open sea propulsion is obvious in the FPP design. As seen, despite the good performance in open sea for the FPP (the propeller fulfills the full speed and is operated in high efficiency), its ice propulsion is not satisfactory. First, it requires quite high power to reserve in the engine in order to perform the ice propulsion at low ship speed and low propeller rpm. Second, for more severe ice conditions (Hice > 1.8m), no matter how large the engine power provided, the FPP still cannot generate enough thrust to propel the ship at the requested speed (2.5 knots). The ducted propeller has the best performance for ice propulsion in severe ice conditions (see Hice = 1.4 ~ 1.6m) but its open sea operation is the worst. If the CPP design is further applied to the ducted

Lee 20

propeller, its engine power PMCR will be the value of Pio and becomes much smaller compared to open CPP for ice operation (see Pio for Nozzle and Pio for CPP in the table). In fact, there are more advantages associated with a ducted CPP in Arctic ice propulsion. For instance, in the strength point of view, the use of a duct to protect propeller blades will make the CPP mechanism safer in ice operation. The CPP has the best overall propulsion performance among the designs. It fulfills the open sea operation well and its ice performance is also good. As seen in Table 18, its required engine power is the smallest among the other designs for the Hice 0.6m ~ 1,2m range. The biggest advantage of CPP is noted when the ice condition becomes more severe. The CPP design can still generate enough thrust (by changing its pitch through turning blade position) to propel the ship to overcome the large ice resistance. However, it should be noted that the strength of CPP is a serious design concern as the large ice impact can easily damage the CPP mechanism. Compared to the direct calculation for ice propulsion power determination, FSICR is quite conservative. Although FSICR did consider the differences of propeller designs through using different Kc values for FPP and CPP (single, twin and triple propellers), it seems the Rule is still quite conservative for engine powering. Even more, FSICR cannot identify the limitation of a FPP design for high ice resistance condition. At high resistance, in some FPP designs such as the one selected, no matter how great the amount of engine power that can be provided to a propeller, the propeller still cannot generate enough thrust for ice propulsion. REFERENCES Bezinovic, D. 2005, ‘Design of an AFRAMAX tanker Baltic ice class IA acc. to new Swedish-Finnish rules 2002 and model testing results – basic hydrodynamics problem’ Marine Transportation and Explotation of Ocean and Coastal Resources, 2005. Chen, H.C. and Lee, S.K., “Time-Domain Simulation of Four-Quadrant Propeller Flows by a Chimera Moving Grid Approach”, Proc. COE VI Conference, Baltimore, MD, Oct., 2005. Chen, B. and Stern, F., “Computational Fluid Dynamics of Four-Quardrant Marine-Porpulsor Flow”, Journal of Ship Research, Vol. 43, No. 4, 1999. Finnish Maritime Administration and Swedish Maritime Administration, “Guidelines for the Application of the Finnish-Swedish Ice Class Rule’, 2005. Finnish Maritime Administration, ‘Finnish-Swedish Ice Class Rule’, FMA Bulletin No., 2002 (www.fma.fi). IACS, 2008, URI3 Machinery requirement for Polar class ships ITTC, “Testing and Extrapolation Methods Resistance – Resistance Test”, ITTC Recommended Procedures, 2002. ITTC, “Performance, Propulsion 1978 ITTC Performance Prediction Method”, ITTC Recommended Procedures, 1999. Kim, H.S., Ha, M.I., Ahn, D., Kim, S.H., Park, J.W., 2005, ‘Development of 115K tanker adopting Baltic Ice Class IA’, the 15th ISOPE, Seoul, Korea, June 16-24, 2005. Lee, S.K., 2007, ‘Engineering practice on ice propeller strength assessment based on IACS polar ice rule – URI3’, pp. 1337-1345, PRADS 2007, Oct.1-5, 2007, Houston, Texas, USA. Lee, S.K., 2006, ‘Rational approach to integrate the design of propulsion power and propeller strength for ice ships’, IceTech 2006, July 16-19, Banff, Canada. Lindqvist, G., “A Straightforward Method for Calculation of Ice Resistance of Ships”, Proc. 10th Int. Conference on Port and Ocean Engineering under Arctic Condition (POAC), Lulea, Sweden, 1989. MAN B&W, “S70MC Project Guide”, 1994 MAN B&W, “Propulsion Trends in Tankers”, 2002.

Lee 21

MAN B&W 'Diesel Facts', pp 14-15, 2006 Feb. Manen, J.D. van and Oossanen, P. van, “Principles of Naval Architecture”, Vol. II, Chapter 6, Propulsion, 1988. MARNET-CFD Thematic Network, “Best Practice Guidelines for Marine Applications of Computation Fluid Dynamics”, 2002. Oosterveld, M.W.C and Ossannen, P. van, “Further computer-analyzed Data of the Wageningen B-Screw Series”, ISP, 22, July, 1975 Valanto, P., “The Resistance of Ship in Level Ice”, SNAME Transaction, vol. 109, 2001 van Gent, W. and van Oossanen, P, 1983, ‘Ducted propeller systems and energy saving’, NSMB publication no. 756a. Yazaki, A. and Sugai, N., 1966, ‘Further model testa on four-blade controllable-pitch propellers’, Ship Research Institute, Paper no. 16 Yazaki, A., 1964, ‘Model testa on four-blade controllable-pitch propellers’, Ship Research Institute, Paper no. 1

Lee 22