Prousalidis JMET Jan

Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 1 Analysis of electric power demands of podded propulsors Analysis of electric power demands of podded propulsors JM Prousalidis and PS Mouzakis, National Technical University of Athens, School of Naval Architecture & Marine Engineering, Division of Marine Engineering, Athens, Greece The authors investigate and analyse the electric power operating conditions of podded electric motor drives, mainly in an attempt to explain the high failure rates of the electric components in pod propulsion installations. The analysis is based on simulations, where the pod torque demands are obtained from experimental results recently published in literature.Thus, the same series of loading scenarios for an entire range of turning azimuth angles of a twin pod configuration are considered. The approach is rather generic as only nameplate data of the motor drives are required while, as the simulations focus on the electric motors, the rest of the ship electric system considered, including generator sets, is represented by a simplified network. Two different sets of simulations are considered, one comprising a synchronous ac motor drive and one with an asynchronous motor of similar rating. In all cases, it is shown that during manoeuvring in certain range of azimuth angles significant overloading occurs exceeding the apparent power capacity of the motors. AUTHORS’ BIOGRAPHIES Dr J Prousalidis (Electrical Engineer from NTUA/1991, PhD from NTUA/1997) is an Assistant Professor at the School of Naval Architecture & Marine Engineering at the National Technical University of Athens, dealing with ship electric energy systems and electric propulsion schemes ([email protected]). PS Mouzakis is a Research Assistant (Naval Architect and Marine Engineer, D-Eng/2009), at the National Technical University of Athens, School of Naval Architecture & Marine Engineering ([email protected]). INTRODUCTION uring the last few decades there have been several significant improvements relating to the vital parts of a ship – hull form, bulbous bow and rudders – and improvements in ship manoeuvrability via thrusters and pods. Pods, in particular, which were introduced some 20 years ago, offer great flexibilities in vessel manoeuvring and machinery arrangements as well as propulsion efficiency. But this progress also introduced new challenges for research and investigation. D No. A16 2010 Journal of Marine Engineering and Technology A podded propulsor is defined as a propulsion or manoeuvring device that is external to the ship’s hull and houses a propeller powering capability. Some ship configurations have a system of tandem propellers with two propellers each connected at each end of the propulsor body. The pod mechanical system comprises a short propulsion shaft on which the electric motor is mounted and a set of rotating elements, radial and thrust bearings. In most cases propulsion is achieved by means of fixed pitch propellers powered by an ac synchronous motor (conventional, permanent magnet or high temperature superconductive) or asynchronous electric motor, fitted inside the pod. Electric power is supplied through cabling connected to the inboard ship grid via slip rings. The loadings generated by the pod are extremely complicated, as a large number of factors and parameters should be taken into consideration (eg, the flow around the fin, the strut, as well as the helicoidal propeller slipstream). In addition, the interaction between the propeller and the pod body(ies) must be also taken into account. The pod body design is of great importance because of the need to minimise the hull boundary layer and the developed vorticity. As the hydrodynamic characteristics of the pod strictly 1 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 2 Analysis of electric power demands of podded propulsors depend on its shape, the body must be designed with the smallest possible diameter along with the greatest possible length, hence, the ratio of pod radius to pod body length must be as low as possible. Furthermore, if the forces and moments developed during pod operation are not accurately estimated, the induced loads to the bearings and the shaft will not be well assessed either, leading to premature ageing and, eventually, failure of most components.1–4 These induced loads strictly depend on the pod azimuth angle.1–4 At each azimuth angle, the thrust and the torque of the pod should be contrasted with the thrust and torque at zero azimuth angle, corresponding to ‘sea-going’ operation. It is noted that these loading conditions have only recently been thoroughly investigated on an experimental basis, as described in a series of papers, conducted by LRS.1–3 In the same work it is reported that amongst the highest failure indices are those of main electric motor parts and, in particular: while cases D-F refer to an asynchronous ac motor, the data of which are tabulated in Table 3. The motors are of almost equal rating and have been taken from actual pod configurations.5 Figs 4–61–3 present the torque measured on the pod propeller for all three pod operation modes considered (A, B and C or D, E and F) for different pod azimuth angles.  Stator windings and core,  Rotor windings,  Slip ring electric connectors. These high failure rates are attributed mainly to high temperature development in the windings, leading to excessive overheating stresses indicating that the motors have not been properly matched with the stressful conditions they are requested to operate in. Hence this paper aims to investigate and analyse the operating conditions of pod motor drives and their effect on electric power. The analysis is based on simulations, where the pod torque demands are obtained from the experimental results published;3 moreover, the same series of loading scenarios, covering the entire operating range of twin pod configurations, are considered. The approach is as generic as possible considering that no specific motor data are required, with the exception of the nameplate data.5 Furthermore, as the simulations focus on the electric motors, the rest of the ship’s electric system, including generator sets, is represented by a simplified network.5 Two different sets of simulations are considered, the first corresponding to a synchronous ac motor drive, the second to an asynchronous one but of similar rated values. Fig 1: Pod configuration in case studies cases A & D Fig 2: Pod configuration in case studies cases B & E POD BEHAVIOUR DURING TURNING MANOEUVRES Figs 1–3 present the twin pod configuration and operating scenarios considered, as obtained from.1–2 In this, when undertaking a full-turn, the resulting forces and moments produced by the propellers located on the port and on the starboard side of the hull, are different. This difference strictly depends on the side boundary layers and on the flow field in the way of the propellers. For the purposes of this paper, simulations of pod synchronous motor behaviour were organised into six different study cases, as shown in Table 1. In the first three cases (A-C) the pod motor is supposed to be a synchronous ac machine with the nameplate data shown in Table 2, 2 Fig 3: Pod configuration in case studies cases C & F Journal of Marine Engineering and Technology No. A16 2010 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 3 Analysis of electric power demands of podded propulsors Case studies Star Board Pod Case Study A Port Pod Star Board Pod Case Study B Port Pod Star Board Pod Case Study C Port Pod Star Board Pod Case Study D Port Pod Star Board Pod Case Study E Port Pod Star Board Pod Case Study F Port Pod x From To P S x P S x x P P S S x x S P P S X P S X X P P S S X S P Pod Drive 3-phase AC Synchronous Motor 3-phase AC Asynchronous Motor Fig 5:Torque vs pod azimuth angle considered (as measured in1–3) in case studies B and E Table 1:The case studies under consideration Nameplate data of synchronous AC Pod Motor Nominal Output Power Nominal Input Active Power Nominal Input Reactive Power Nominal Power Factor Speed at full load Line Voltage Efficiency at Full Load Design Life 25 MW 25.64 MW 19.23 MVAr 0.8 120 RPM 6600V 97.5% 30 years Table 2: Nameplate data of AC synchronous pod motor Nameplate data of asynchronous AC pod motor Nominal Output Power Nominal Input Active Power Nominal Input Reactive Power Nominal Power Factor Speed at full load Line Voltage Efficiency at Full Load Design Life 24.64 MW 25.20 MW 15.27 MVAr 0.855 120 RPM 6600V 97.5% 30 years Table 3: Nameplate data of AC asynchronous pod motor Fig 6:Torque vs pod azimuth angle considered (as measured in1–3) in case studies C and F In the starboard-pod experiment,1–3 positive angles refer to turns of the pod propeller to starboard (open sea), while negative angles refer to turn to port (hull side). On the other hand, with port-pod operation, positive angles refer to turns of the pod propeller to hull side (starboard side), while negative angles refer to turns to open sea (port side). Moreover, regarding twin screw configurations, the following assumption is made:5 The curve representing the torque demand of a separate pod torque demand for different azimuth angles remains unchanged* in the twin screw configuration. This means that it is assumed that there is no interaction between these two propulsive devices, which is, in general, a plausible assumption for the ship structures where twin pods are installed. The electric power grids of the two main configurations are depicted in Fig 7a & b, respectively, as modelled in the PSCAD environment of the Manitoba HVDC Center.6 Specifically, considering that the interest in this work is focused only on the operation of pod motors, the power sources are ideal. In this way, the voltage to pod motor terminals remains constant despite any current fluctuations, so that the pure power demands of the pod motors are identified.5 Further, taking into account Fig 4:Torque vs pod azimuth angle considered (as measured in1–3) in case studies A and D No. A16 2010 Journal of Marine Engineering and Technology * Provided that the angular speed remains constant 3 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 4 Analysis of electric power demands of podded propulsors Fig 7: Electric power grid of the two configurations considered (in PSCAD simulation environment) – (a) the pod motor is an ac synchronous machine (b) the pod motor is an ac asynchronous one that the apparent power absorbed by the motor is the product of the input voltage times the input current, the apparent power waveform is directly proportional to that of the current. Should the source not be ideal, and an actual generator is used instead, with its associated speed governor and Automatic Voltage Regulator (AVR), the behaviour is expected to change. In a future work this interaction between generators and pod motors as well as the rest of the ship grid is intended to be studied. PRESENTATION AND DISCUSSION OF RESULTS The most representative results of all simulated study cases are presented and in each case study the following quantities are figuratively presented: Power and current demands vs azimuth angle (P, Q, S, I vs a-curves)  Active vs reactive power demands (P vs Q -curves).  For the sake of tangible comparison, all waveforms are expressed in percentage (%) using as reference base values the 4 corresponding synchronous and asynchronous motor rated values in Tables 2 and 3. However, concerning the active and reactive power quantities, it is questionable whether they should refer to their corresponding rated values and not the rated apparent power value.5 More specifically, considering that active and reactive power complement one another, should not their %-values both refer to the (total) apparent rated power, ie, the conditions where the electric motor capacity is superseded with subsequent overloading, and are more figuratively identified? This is the reason why two pairs of reactive vs active power demand diagrams are provided in all case studies. The difference between them is the reference values and consequently the motor capacity limit curves. Regarding the time analysis of the simulations, for the first three simulations (powered by the synchronous machine) (Cases A, B & C), the motor(s) starts directly when the starboard motor azimuth angle is equal to -30.0 degrees o o (towards the hull). In addition, the angle range [-30.0 , +30.0 ] is covered in a linear manner, within 60 seconds. However, due to the fact that there is no motor starting procedure before it o reaches pod azimuth angle ±30.0 , a large starting-up transient phenomenon takes place. For that reason, not all recorded simulation results refer to the draft examined azimuth angle area Journal of Marine Engineering and Technology No. A16 2010 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 5 Analysis of electric power demands of podded propulsors o o o [±30.0 ], but to that of [-25.0 , +30.0 ]. On the other hand, for the last three simulations, (powered by the squirrel cage asynchronous machine) (Cases D, E & F), it is considered that the motor(s) starts at azimuth angle 0.0 and then it turns to the o o starting azimuth angle (-30.0 or 30.0 ).1–3,5 Finally, the angle o o range [-30.0 , +30.0 ] is covered in 60 seconds.5 Study case A remarks As already mentioned, in case study A the simulation concerns the operation of the starboard pod only. The starboard pod turns from the port side of the hull to the open sea (starboard). The torque demand as measured in1–3 is shown in Fig 4. From Fig 8 it can be observed that the active power curve strictly depends on the propeller torque required, as depicted in Fig 4. Moreover, significant overloading with respect to the active power is noted, especially in negative angles. Quite the opposite remark is made for the reactive power. However, by inspecting the apparent power and the active vs reactive power diagrams (Figs 9 and 10) it is seen that significant overloading also occurs relating to the apparent power capacity of the motor. Fig 9: Pod reactive power vs active power in case study A. Active and reactive power quantities are expressed in % with respect to their corresponding rated values Fig 10: Pod reactive power vs active power in case study A. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified Fig 8: Pod power (active, reactive and apparent) demands vs azimuth angle in case study A In addition, the starboard pod active power demand o reaches its lowest ever rate at 10 starboard (towards the open sea), while at the same point, reactive power reaches its highest ever rate. On the other hand, active power reaches its o highest rate at 20 port (towards the hull), while at the same point, reactive power stands at its lowest ever rate. o Furthermore, the pod requirements at azimuth angle 22.6 are equal to those at zero azimuth angle, pointing to some o kind of symmetry with respect to the angle of 12.3 . In the P-Q diagrams, Figs 9 and 10, a hysteresis phenomenon is noted, as the trajectory passing from negative to positive angles is different from that of the opposite direction. This hysteresis effect is noticed only in the motor reactive power. Considering that the reactive power is adjusted by the synchronous motor exciter, this phenomenon can be attributed to the integrator module of the motor exciter, which introduces this kind of memory effect. Fig 11 presents the reactive power demands of the examined asynchronous motor of Case A vs its active power demands referred to their rated values. In detail, this graph is the product of four continuous operations with return to its o initial position (starboard pod azimuth angle -30.0 ). As already mentioned, these loops are attributed to memory effects of the exciter integrator. No. A16 2010 Journal of Marine Engineering and Technology Fig 11: Explanatory figure on how the hysteresis loops of Fig 9 are created Study case B In study B, the simulation concerns the operation of both pods. The starboard pod turns from the hull side to the open sea, while the port pod turns from the open sea to the hull side. During this operation, significantly different loading torque is noted between the two motors (Fig 5). More specifically, by inspecting the results depicted in Figs 12–14, the following remarks are made:  By comparing Figs 8 and 11, the loading pattern in case B is more symmetrical with respect to 0 angle than in 5 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 6 Analysis of electric power demands of podded propulsors  case A. The symmetry of this case is due to the operation of both pods. However, especially in large overload o o o o conditions [-30.0 , -17.0 ] and [+30.0 , +17.0 ] small asymmetry has been observed in reactive power requirements and thus in apparent power and current curves. In this case, no overloading in reactive power demand is noted in the entire range of angles.    On the contrary, overloading is noticed in active and apparent power, as well as the current waveforms. At zero azimuth angle, the total active power demands of both pods stand at the lowest ever rate, while the corresponding demands for reactive power stand at the highest rate. The total pod requirements in active power depend on the azimuth angle regardless of the angle direction. However, the total pod requirements in reactive power, total current and apparent power depend on angle direction possibly due to memory effects as well as the collaboration effects of the two exciters. Study case C In study C, the simulation corresponds to the operation of the starboard and port pod, both turning from the hull side to the open sea. In Figs 15–17 the total power demands are depicted. In this case, which resembles closely case A, the following remarks are made: Fig 12: Pod power (active, reactive and apparent) demands vs azimuth angle in case study B   The azimuth angle for each pod where the maximum or minimum total demand for active and reactive power occur, are the same with that of study case A. o Like in case A, the pod power demands at 22.6 are equal o to those at 0 . Fig 13: Pod reactive power vs active power in case study B. Active and reactive power quantities are expressed in % with respect to their corresponding rated values Fig 15: Pod power (active, reactive and apparent) demands vs azimuth angle in case study C Fig 14: Pod reactive power vs active power in case study B. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified 6 Fig 16: Pod reactive power vs active power in case study C. Active and reactive power quantities are expressed in % with respect to their corresponding rated values Journal of Marine Engineering and Technology No. A16 2010 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 7 Analysis of electric power demands of podded propulsors   depend on the applied torque (Fig 4), with the minimum o o values in 10 towards the open sea (+10 ), while the maxio o mum values 20 towards the hull of the ship (ie, -20 ). For similar reasons, ie, the absence of excitation circuit with integrator module, no hysteresis effect is noted in the active vs reactive power diagrams, Figs 19, 20. While apparent and active power overloading is noted in o o o o the range of angles [-30 , -10 ] and [+25 , +30 ], the corresponding range of angles of reactive power overo o o o loading is [-30 , 0 ] and [+20 , +30 ]. Fig 17: Pod reactive power vs active power in case study C. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified   As in case A, a minor overloading in terms of reactive o power is noted in the range between 0o and 22.6 . On the contrary, in this region no overloading is noted in terms of active/apparent power and current; however, significant overloading occurs in the complementary regions of o o o [-30, 0 ] and [+22.6 , +30 ]. In overloading conditions, there are differences noted between motor reactive power demands in Cases A and C, of up to 7% of their rated values. On the other hand, no differences between these two study cases are noted regarding active power requirements. Fig 18: Pod power (active, reactive and apparent) demands vs azimuth angle in case study D Fig 19: Pod reactive power vs active power in case study D. Active and reactive power quantities are expressed in % with respect to their corresponding rated values Fig 20: Pod reactive power vs active power in case study D. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified Study case D Like case A, this case involves only the use of starboard pod, which turns from the port side of the hull to the open sea (starboard). The corresponding results are presented in Figs 18–20, from which the following remarks are made:  Unlike the previous cases, both active and reactive power quantities have identical variation. This is due to the asynchronous type of the motor drive and its lack of independent excitation circuit regulating reactive power. Evidently, apparent power and consequently current follow similar track. Hence, all the examined magnitudes strictly No. A16 2010 Journal of Marine Engineering and Technology Study case E Similar to case B, case E involves two pods (port and starboard), but are both driven by asynchronous pods. The starboard pod turns from the hull side to the open sea, while the port pod turns from the open sea to the hull side. From the simulation results presented in Figs 21–23, the following remarks are made:  As in case D, all quantities examined follow the track of the pod torque demands, see Fig 5. Thus, a symmetry 7 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 8 Analysis of electric power demands of podded propulsors o  with respect to 0 , the point with minimum power demands, is noticeable. While overloading situation with respect to active and o o apparent power is noted in the range of angles [-30 , -10 ] o o and [+10 , +30 ], reactive power overloading is apparent o o in the entire range [-30 , +30 ]. Study case F Like case C, case F simulation concerns the operation of the starboard and port pod with both of them turning from the hull side to the open sea. From the simulation results presented in Figs 24–26, the following remarks are made:  The resemblance between cases A and C is repeated between cases D and F. Fig 21: Pod power (active, reactive and apparent) demands vs azimuth angle in case study E Fig 24: Pod power (active, reactive and apparent) demands vs azimuth angle in case study F Fig 22: Pod reactive power vs active power in case study E. Active and reactive power quantities are expressed in % with respect to their corresponding rated values Fig 23: Pod reactive power vs active power in case study E. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified 8 Fig 25: Pod reactive power vs active power in case study F. Active and reactive power quantities are expressed in % with respect to their corresponding rated value Fig 26: Pod reactive power vs active power in case study F. Active and reactive power quantities are expressed in % with respect to the rated apparent power value so that the operation limits are identified Journal of Marine Engineering and Technology No. A16 2010 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 9 Analysis of electric power demands of podded propulsors    All the examined quantities (active and reactive power, as well as three phase current) follow the pattern of the propeller torque demand (Fig 6). Hence, minimum o values appear 10 towards the open sea, while maxio mum values are noted 20 towards the hull of the ship o referring to pod 0 . o In addition, the pod requirements in azimuth angle 22.6 are equal to those in zero azimuth angle. While apparent and active power overloading is noted o o o o in the range of angles [-30 , -10 ] and [+25 , +30 ], the corresponding range of angles of reactive power overo o o o loading is [-30 , 0 ] and [+20 , +30 ].  RESULTS/DISCUSSION In this section, the results of all case studies are compared to each other and discussed further. For comparison purposes, certain characteristic values concerning overloading are tabulated in Table 4, eg, range of azimuth angles where overloading occurs, as well as maximum and minimum values. Hence, the following conclusive remarks are made:     In general, results of case A resemble to a great extent those of case C. This is due to the fact that the hydrodynamic loading in case A is supposed to be the same as case C. Thus, at each simulation instant, the waveforms of active and reactive power as well as motor currents are approximately the same. For the same reason, regarding asynchronous motor configurations, the simulation results of case D are the same as those in case F. Cases A, C, D and F seem to be significantly worse than cases B and E from the electric power overloading point of view, ie, in terms of both:  maximum values of all power quantities P,Q,S (and I)  range of azimuth angle with overloading in terms of P,Q,S (and I). This overloading can be stressful for the generator sets, too, and has to been taken into account in the electric balance analysis and the generator selection study. It is noted that this situation is not a fast transient, eg, in terms of inrush current during motor starting-up with a duration of 100–500ms, but could last for several minutes and hence could cause several adverse phenomena. Furthermore, it seems that the worst overloading operating condition of the electric motor is in the vicinity of o -20 , as in all cases this is where the maximum apparent power demands occur reaching occasionally, an overloading level of 160% the rated one. Active power P follows the pattern of pod torque demands as it is directly related with it. Thus, due to the significant overloading occasionally occurred, pod motors could eventually encounter adverse consequences, such as premature ageing or failure of the propeller, the shaft and the bearings. In the asynchronous motor cases (D, E, F) the overloading noticed in torque demand is accompanied by a decrease in speed; this No. A16 2010 Journal of Marine Engineering and Technology  explains why the resultant active power demand is less than the corresponding torque demand. Reactive power Q has a more complicated behaviour. In cases A-C with the synchronous motor drive, reactive power is adjusted and controlled by the independent AVR of the motor and, hence, appears to follow a pattern complementary to that of the active power P and torque T; thus minimum Q values coincide with maximum P (and T) values and vice versa, due to the effort of the AVR to keep the total apparent power demand to low level. According to starboard simulation results, the starboard pod for the azimuthing angle range [-25.0 -14.75] requires from the system a negative amount of reactive power. This means that the motor produces rather than consumes reactive power. Moreover, the presence of the AVR in these cases (A-C), results in an intriguing hysteresis effect in the P-Q diagrams, according to which tracking from positive to negative angles is different from the other way round. On the other hand, in cases D-F, with the asynchronous motor drive, where no reactive power regulation device is present, reactive power Q follows an identical pattern to that of the active power. Therefore, in cases D-F, reactive power pattern resembles that of the pod torque demand leading in certain cases to even more significant overloading as the motor operation cannot be adjusted at all. This remark points why the asynchronous motor could be completely inappropriate for pod applications. Finally, in these last three cases where no reactive power regulation can be done, no hysteresis effect in the P-Q diagrams is noticed. Moreover, apparent power S is yielded from the combination of both active and reactive power according to the well known expression: S = P2 + Q2 (1) In general, apparent power demands resemble more the active power, which can be explained considering that this power portion is always predominant. The angle regions, where rated apparent power is exceeded are most significant as the apparent power shows the motor overall capacity to withstand overloading or not. This is why the P-Q diagrams where both active and reactive power refer to the rated apparent power rather than their own rated values are most useful. Finally, current changes in a manner identical to the one of the apparent power. This is explained as the current is directly proportional to the apparent power, considering that they are related via the expression: I= S 3V (2) As already mentioned, in this work the terminal voltage V of the motor is considered to be constant and provided by an ideal source. In a future work, the pod motor will be 9 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 10 Analysis of electric power demands of podded propulsors considered connected to the actual ship grid, and the terminal voltage will be subjected to fluctuations by several grid operation factors including the pod motor itself. Apparently, considering that overloading takes place in an entire range of values, the appropriate selection of the pod motor drive should be the result of a customized optimisation methodology. More specifically, besides any other design constraints, two contradictory concepts have to be taken into account: 1. If the motor rated power is selected according to the mean power demand then significant overloading would occur leading to accumulated insulation stressing, premature ageing and failure of components as already statistically recorded.1 On the other hand, if the motor sizing is based on the overloading operating conditions noticed in certain pod turning angles, then the motor will normally (ie, in ‘sea-going’ conditions) operate in values significantly lower than the rated ones, resulting in lower efficiency and power factor which consequently means higher fuel consumption of the generators and higher total operation cost. 2. In any case, the actual overloading limitation is the apparent power as already discussed on the occasion of the P vs Q diagrams. Furthermore, in synchronous motor drives, where an independent excitation circuit exists, reactive power Q follows a complementary pattern to that of active power P resulting in comparatively lower power demands. This is a major advantage of this motor type, despite even the hysteresis effects, ie, the fact that different overloading is noted when turning from positive angles to negative ones than the opposite. Case Study A B C D E F Range of angles where Overloading occurs P Q S, I o o o o [-25 , 0 ] [-25 , -14 ] o and [0, +22.6 ] and o o o o [+22.6 , +30 ] [+22.6 , +30 ] o o o o [-25 , -18 ] [-25 , -10 ] and and o o o o [+10 , +30 ] [+11 , +30 ] o o o o [-25 , 0 ] [-25 , -14 ] o and [0, +22.6 ] and o o o o [+22.6 , +30 ] [+22.6 , +30 ] o o o o o o [-30 , -0 ] [-30 , -7.5 ] [-30 , -10 ] and and and o o o o o o [+25 , +30 ] [+20 , +30 ] [+23 , +30 ] o o o o o o [-30 , -5 ] [-30 , -13 ] [-30 , -14 ] and and and o o o o o o [+14 , +30 ] [+5 , +30 ] [+13 , +30 ] o o o o o o [-30 , -0 ] [-30 , -7 ] [-30 , -10 ] and and and o o o o o o [+25 , +30 ] [+20 , +30 ] [+24 , +30 ] In the case of asynchronous motor drive where no independent excitation is present, reactive power has a similar behaviour to that of active power, resulting in more significant overloading in terms of apparent power. This is the major reason why this motor type is to be exempted from pod applications. Nevertheless, novel alternative motor configurations, especially those regarding their power demand controllability, have to be sought and investigated in depth – a direction that future work by the authors is focused on. NOMENCLATURE AVR: Automatic Voltage Regulator I: motor current P: active power demand of the motor Q: reactive power demand of the motor S: apparent power demand of the motor T: motor output torque V: terminal supply voltage of the motor CONCLUSIONS In this paper an effort is made to investigate and analyse the electric power operating conditions of pod electric motor drives in an attempt to explain the high failure rates of the electric components in pod propulsion installations. The analysis is based on simulations for a twin pod configuration, the torque demands of which were obtained from experimental results recently published. Two alternative electric motor types are considered, a synchronous and an asynchronous one. Significant overloading is Maximum Power Demand P Q S, I 155% 110% 125% o o o at -20 at 10 at -20 Minimum Power Demand P 90% o at 10 Q -40% o at -20 S, I 98% o at 10 -35% at o -20 o and +25 -33% o at -18 97% o at 15 95% o at 10 89% o at 10 130% o at +30 - 108% o at +30 155% o at -20 110% o at 10 125% o at -20 98% in the range o o [-10 , 10 ] 90% o at 10 135% o at -20 160% o at -20 141% o at -20 87% o at 10 115% at o -30 and o +30 135% o at -20 130% at o -30 and o +30 160% o at -20 120% o at -30 o and +30 142% o at -20 92% in 100% in the range the range o o o o [-10 , 10 ] [-10 , 10 ] 87% 95% o o at 10 at 10 97% o at 10 95% in the range o o [-10 , 10 ] 89% o at 10 Table 4: Overview of overloading critical values in all case studies 10 Journal of Marine Engineering and Technology No. A16 2010 Prousalidis_JMET Jan.qxd 12/30/09 5:17 PM Page 11 Analysis of electric power demands of podded propulsors noticed in certain operating ranges of pod azimuth turning angles, eg, in turning-to-port manoeuvring conditions, exceeding as much as 150% of the motor capacity. Besides the pod torque defining the active power demands, the reactive power demands play a significant role. Thus, it is shown that the situation is milder in the synchronous motor drive due to the presence of the excitation circuit regulating the motor reactive power demands, and hence also its total apparent power demands. Nevertheless, extra research work is required in terms of optimising the motor performance without exceeding its capacity. ACKNOWLEDGEMENTS The authors wish to express their gratitude to Professor Gerassimos Politis for his valuable advice on pod propeller hydrodynamic behaviour. No. A16 2010 Journal of Marine Engineering and Technology REFERENCES 1. Ball WE and Carlton JS. 2006. Podded propulsor shaft loads from model experiments for berthing manoeuvres. International Journal of Maritime Engineering, RINA. 2. Ball WE and Carlton JS. 2006. Free-running model experiments in calm-water and waves, International Journal of Maritime Engineering. RINA. 3. Carlton JS. 2008. Podded propulsors: Some results of recent research and full scale experience. IMarEST Journal of Marine Engineering & Technology (Part A11). April 2008, pp 3–16. 4. Islam MF, He M, Veitch B and Liu P. 2007. Cavitation characteristics of some pushing and pulling podded propellers. International Journal of Maritime Engineering, RINA. 5. Mouzakis P. 2009. Analyzing and resolving electric power supply quality problems in ship electric networks due to large power machine operation. Graduation Diploma thesis. 6. HVDC Center, PSCAD/EMTDC User’s Manual, Manitoba (Canada), 2006. 11