The Advance Electrical System of Submarine Design Capt.Der-Wei Chen,Jee-Ray Wang,LT.CDR.Tzu-Chiang Hwang ABSTRACT¡G This paper is discussing about submarine electrical system design and analysis.There are includes Electrical Load Analysis (ELA) and Electri cal Load calculation for submarine due to design process. The first, i t will be introduce and analysis the advance electrical system charact eristics and technology. Then, in according with the main object to ca lculate and discussion the energy conservation during design process, and base on the analysis report to built up estimate electrical load r equirement model for pre-design period. Finally, all of the relation e ffective factors can be identify and provide a solution scheme. The re sults can offer the Underwater Vehicle design and developments, also i t helpful for practical work in the future. Keywords: underwater vehicle, electrical distribution system, solution scheme. 1. Introduction The powering of an Underwater Vehicle(submarine) is one of the most im portant factors in the determination of its size. But, in recent liter ature are very less to discuss about the powering of submarine design. As with every other design feature the powering assessment stems from the user¡¦s requirements. Power and energy storage are almost entirel y dependent on the vehicular performance requirements of the submarine . Though the operational equipments place some demands for energy prov ision and must be included in what is known as the Hotel Load (that re quired to keep the submarine ¡§alive¡¨). However, to extend the time f ully submerged the propulsion demand may have to be reduced to a minim um and then it is the Hotel Load drain on the energy storage which wil l determine the actual time submerged. The primary factors which govern the power and energy storage requirem ents are: (1) the maximum speed and for how long; (2) the range of ope ration. The maximum speed of a submarine can be the most difficult and content ious aspect of the dialogue between operator and designer. This is bec ause it is difficult to find logical reasoning to arrive at a required maximum speed. The propulsive power requirements for submerged submar ine of given displacement vary as the cube of speed. In essence, to do uble the maximum speed of a submarine from 20kts to 40kts would requir e 8 times the propulsive power. In fact, as the submarine would have t o be considerably larger to accommodate an eightfold increase in power plant, the speed achieved would fall short of the 40kts and so even m ore power and size would be call for. From military viewpoint having h igh speed available is always desirable, because with a speed advantag e the submarine may be able to outrun its attackers or, if attacking, close rapidly on its target. Since submarine warfare tactics are large ly involved with stealth it can be argued that use of high speed in ev asion or attack is not the only, or necessary the optimum, tactic. The risk of an operator/designer impasse could be avoided by the designer adopting a pragmatic approach and attaining as high speed as possible with the limited available choice of power plants and energy storage devices. Whilst maximum speed requirements determine the size of the propulsion plant needed, it is the time at speed which can govern the energy cap acity requirement of a submarine. For conventional submarine may only be able to run at maximum speed for half an hour before the battery is exhausted, when it would have to surface or snort on diesel engines t o recharge the batteries, and it is important to establish the require ments for time associated with speed pattern. The other requirement to be determined is the submerged endurance of t he vessel which depends upon the energy storage capacity and the speed s required whist submerged. If the submarine is not attempting to move then the drain on energy storage is just that to meet the Hotel Load, which includes any energy consumption needed to sustain a breathable atmosphere. 2. Electrical Load Analysis for Naval Surface Ships 2.1 Electrical Load Analysis Method for Surface ships The current Electrical Load Analysis(EAL) is formatted to reflect the guidance of NAVSEA Design and Practice Criteria Manual Chapter 300 (NA VSEA 1992) , Design Methods for Naval Shipboard System, Section 310-1( 1998) , and Electrical Load and Power Analysis for Surface Ships (MIL- HDBK-2189 1999) . The guidance requires four operational scenarios be reviewed at two environmental conditions. The operational scenarios ar e Anchor, Cruise, Mission, and Shore. The two different environment co nditions are at 100F and 900 ambient air temperatures. The output of t he ELA is a prediction of the maximum operating load under the worst-c ase environmental conditions. The largest ship is then used to size th e electrical generation plant after design and service life margins ar e applied. The ELA contains the power requirement for each electrical load and th e percentage that the load would be on during a 24 hour time period, k nown as load factor. For example, a fire pump might have a load of 101 kW and an operating percentage or load factor of 90 percent. This load factor is converted into the probability of occurrence and kept at 90 percent or 0.9. But the fire pump would randomly be on in 90¢H of the simulations and off in 10¢H of the simulations. In some instances the load factor and probability will not coincide because of inter and in tra zonal coordination that is required for probabilistic analysis. Th ese modifications are described in more detail later in this paper. The previous methods averaged the power output from generators over ea ch connected load and calculated load based on a demand factor. This t echnique is not accurate enough in the contract design stages or for z onal architectures since it artificially averages the electrical loads . Table 1 represents the PFG 2 Class ship ELA data and exact electrical loads in cruise condition. It shows the practical cruise condition ele ctrical loads only half of the ELA. The results will impact the choice of capacities and number of the ship service generators installed. 2.2 Monte Carol Analysis Method The proposed methodology uses a systems engineering approach, applying a probabilistic (Monte Carol) analysis of the electrical loads at eac h switchboard, based on the Electrical Load Analysis. The opportunity to perform a more accurate estimation of connected loa ds reflects the fact that as we continue to apply systems engineering methods to our design and verification process we have more tools at o ur disposal. Combined with advances in computational power, probabilis tic simulations can be performed in a fraction of the effort previousl y required. A probabilistic approach was chosen for this analysis in order to more realistically depict the electrical loading seen by each connected lo ad. This approach allows the designer to evaluate each connected load individually as opposed to the current method of averaging the total e lectrical load. This project used a Monte Carlo Method, which is based on probability theory and is widely used to analysis queuing problem. To determine the representative electrical load profile for each conn ected load, a Monte Carlo analysis as performed on the individual elec trical loads and their respective probability of occurrence. The loadi ng and probability data was based of the ELA data for electrical loads and power factor, respectively. The U.S. Navy has used a radial electrical architecture design in one form or another for nearly a hundred year. However, with ever increasi ng electrical loads, both from new mission applications and the moveme nt from steam, air, and hydraulic auxiliaries to electric auxiliaries, a new electrical architecture is needed to distribute power throughou t the ship. Zonal distribution provides this new architecture while pr oviding a greater reliability and survivability. However, the empirica l demand curves and deterministic processes that have worked on radial architectures, fail to address the uniqueness of high voltage zonal a rchitectures using a limited number of large transformers. The zonal electrical architecture has several characteristics that are different from a radial electrical architecture. Only one of these ch aracteristics will affect the sizing of the switchboard and its supply ing transformer. Loads that support distributive systems, such as fire pumps, or air conditioning plants need to be reflected on each transf ormer. The radial system doesn¡¦t have this concern, as its loads woul d be fed from a ship level distribution, instead of the zone basis. Th e level of conservatism in the traditional guidance though, still affe cts the radial system. The US Navy General Specification for Ships Sec tion 314 (GENSPEC 1995), mentioned earlier, requires capacities of not less than 100¢H of the connected load and usually that number needs t o include spares. While providing 100¢H capacity of connected load sho uld ensure 100¢H confidence of meeting demand, at what cost in dollars , weight, and volume, is this confidence bought? 2.3 Methodology The ELA data was sorted by each switchboard and a Monte Carlo analysis was performed on each switchboard data set. The Monte Carlo analysis used the ELA probability to determine if a specific load was on or off . For example, the fire pump would randomly be on in 90¢H of the simul ations and off in 10¢H of the simulations. If the load was on-line, th e power requirement for that load was added to the other ¡§on-line¡¨ l oads within the transformer data set. If the load was off-line, the po wer requirement for the load was zero and thus it was not added into t he total power required for the transformer. This analysis was perform ed 2440 times per transformer and the total power requirement for each transformer was plotted versus the probability of occurrence based on the 2440 runs. The analysis was performed 2440 times to reflect a 99¢H level of confi dence that the calculated values are within 30kW of the actual value. The number of iterations is based on a confidence equation. The actual 95¢H values from the analysis are used, instead of approxim ated values obtained by normalizing the data. While a normal approxima tion is fairly accurate, deviations as large as 3¢H or 87kW were possi ble. The more accurate analysis data was used rather then increase the margin of error to 120KW. Also, the power factors values were calcula ted to the fifth significant decimal place, but rounded to the second significant decimal place. This was done because the power factor valu es are only known to the third significant decimal place. This allows another 0.5¢H deviation from what could be the actual value. Therefore , the results have a 50kW margin error with a 99¢H confidence level. 2.4 Applications Compare with the FFG-7 class ships actual electrical load and ELA, tha t we can find the actual electrical load only half of ELA (see Table 1 ). That¡¦s mean the ship always running under inefficient conditions c ause low effective. Base on this reason, using the Monte Carol Analysi s Method to recalculate the ELA during in the design process. After ca lculate the FFG-7 ELA is same as the actual electrical load in service life. Then chosen others class ships to verify the Monte Carol Method accuracy, the results in Table 2. 3. Electrical Load Analysis for Submarine The powering of a submarine vehicle is one of the most important facto rs of its size. In the past research, the power requirement are determ ined in conjunction with speed by the size of the vessels, and hence t he designer encounters a loop in the design process whereby the output of the power assessment in terms of a volume requirement for propulsi on plant is itself a significant input to that assessment. In control engineering terms this is a positive feed-back loop, which can all too easily cause a growth in the total size of the design. 3.1 Power Analysis for Submarines For recognizing the dominating objects,which ¡§consume¡¨ a significant part of the available internal volume, mass and electrical energy of a submarine, several classes of diesel-electrical submarines built in the 1980¡¦s have been evaluated. A comparison study by Stenard present s percentages of internal volume and mass of components for submarines with submerged displacement between 1200-2900 tons. Table 3 shows mea n values of four different classes of boats, which are examined in thi s study. The presented percentages can be used to give an indication o f values which are achieved in practice. In general, this comparison s tudy shows that the largest percentage of the internal volume is occup ied by the components fulfilling the Manoeuvring and Energy supply fun ction and considerably less by the Life support and other functions. T he largest percentage of the standard displacement is occupied by the objects fulfilling the Carrying platform, Manoeuvring and Energy suppl y function. Individual classes of submarines can differ from these percentages due to extreme design requirements. Typical examples of these extremes ar e: relative large mass for Carrying platform due to extreme diving dep th requirement, or relative large volume and mass for Manoeuvring and Energy supply caused by an extreme submerged range at high speed requi rement, or relative large volume for Life support caused by extreme to tal mission endurance. An evaluation study by Pel presents percentages of installed nominal l oad of electric energy consumers for submarines with submerged displac ement around 2900 tons. Table 4 presents the results of this evaluatio n. The presented percentages can be used to give an indication of values which are achieved in practice. In general, this evaluation study show s the largest percentage of the electric energy is consumed by the com ponents fulfilling the Manoeuvring and Energy supply function and cons iderably less by other functions. The percentages are based on total n ominal load of electric energy consumers in any type of the electric e nergy. Thus, the nominal loads are not converted to the Direct Current source type of energy. Percentages for burst speed and maximum contin uous speed conditions are presented to show the increasing significanc e of propulsion on the budget when the required submerged speed increa ses. From the above presented percentages can be concluded that the compone nts, fulfilling the Manoeuvring and Energy supply function, determine a significant part of the three budgets, relevant for this thesis. Fur ther the components fulfilling the Carrying platform function are sign ificant for the weight budget and Life support function is significant for the volume budget. In accordance with above discussion and analysis, that we obtained the following results: (1) the conventional submarines electric requireme nt is base on the burst speed during submerged condition, (2) the maxi mum range of continuous speed. So, for to verify the theory is satisfa ctory that analysis the Sea-Dragon class ship due to burst speed and t he requirement electrical load (see Table 5), then comparison with the Table 4. After the analysis and comparison that the result same as the past res earch. 3.2 Design the Aspects of Propulsion Plant 3.2.1 Models for Object Energy Sizing The load is defined as the maximum required load in any of the conditi ons defined in Table 5. These include the nominal loads for the Direct Current (DC), Alternating Current (AC), Hydraulic, Pneumatic, and Hea t release. The purpose of the sizing model for the components fulfilling the Ener gy Conversion function is defined by: calculate the nominal loads for the energy conversion systems for each of the four relevant operationa l conditions. The energy formats are specified in Table 6. The required amount of energy is analysis for each object, defining fo r each condition an operational factor, load factor, and simultaneity factor, (see Table 7). The total required amount of energy for the obj ects is calculated by multiplying these factors with the nominal load of the objects. 3.3 Propulsion Motors These are direct current motors, usually arranged with the direct coup led to the propeller shaft. They are required to be of a size to deliv er the necessary shaft power at the top speed of the submarine. Howeve r, the selection of a suitable motor is not only governed by the maxim um shaft power or the voltage available to drive it. Because it is dir ectly on the propeller shaft its speed of rotation (rpm) is that at wh ich the propeller has been designed to deliver full thrust. Similarly the torque output of the motor must match the torque of the propeller at full power conditions. Hence there is a matching problem to be cons idered (see Figure 1). In general, that high propulsive efficiency calls for a large diameter , low rpm propeller with high torque. Suffice it to say for now that t here are limitations on the field in the gap between rotor and stator of an electric motor and circumferential force that can be generated. Thus a compact design of electric motor for a given power would give o nly low torque at high rpm. The requirement of the propulsor for high torque at low rpm is directly in opposition to this. The propulsor nee ds call for a large diameter rotor, resulting in a heavy, volume-consu ming propulsion motor at the narrow part of the pressure hull stern, a nd so posing hydrostatic balance and space layout problems which are d ifficult to satisfy. There are many in consequence have to be a compro mise in the overall design in which the propulsor rpm are increased an d the torque decreased to match a motor of total size. There may not b e a continuous range of choice, but rather step jumps in the size of a vailable motors, which impact on the propulsor design. A possible alte rnative is the use of two smaller motors on the shaft, which trades di ameter demand for additional length demand, though it also complicates the electric control. The designer could also choose indirect drive of a motor through a gea rbox to the shaft, which overcomes the matching difficulties but at ex pense of cost, weight and space of the gearbox. It is not favored for submarines because a direct drive system can be mad very quiet, where it is difficult to eliminate from a gearbox. For calculated the required space properties of a DC Main Electric Mot or (MEM) by conventional submarines. The following numerical model wil l be approved it. The purpose of the MEM is to convert efficiently Dir ect Current energy into mechanical energy. Two design conditions are d efined for the MEM: a burst and maximum continuous condition. The burs t condition defines the maximum power and associated rotation rate. Th is condition determines the maximum allowable mechanical and magnetic load, while the maximum allowable thermal load may be exceed. The maxi mum continuous defines the maximum allowable thermal load. On the othe r hand, the MEM sizing model is also to determine the space properties of the MEM. The space properties contain width, length, height, area and volume which is required to allocate the MEM to a space. Because s ome of the MEM designs parameters are different by the designers and m anufactures, so it just calculation the width of MEM only in this pape r. Motor diameter is determined by the geometry and topology of the compo nents, represents by a function: ¡K¡K(3.1) To calculate the rotor diameter, its armature winding is divided into a finite number of conductors. For each conductor the 2nd law of Maxwe ll is applied to determine the electromotive force. The flux density u sed in this law is determined using Faraday¡¦s principle and the obser vation that within a magnetic circuit no flux can be lost. Using these first principles and geometric knowledge about DC motor design, the r otor diameter is calculated by the following equation: From the eq(3.2) we can obtain the MEM rotor diameter, then the result s bring into the eq(3.1) and it can be obtain the MEM width size. For verify these equations are satisfactory that author use K.Deleroi stud y model in 1993 to confirm. The model parameters are: (1) maximum outp ut power is 4100kW in submerged condition and (2) the rotor RPM is 200 (others parameters see Table 8). The results are: (1) rotor diameter is 1.94m and (2) motor width is 2.45m. It is as same as the K.Deleroi study model in 1993. Then, the author use above equations to calculate the MEM output power from 1,000 to 7,000kW and the rotor speed is 100, 150 and 200rpm, and try to find the relationship between these parameters. The results sh ows that the rotor and motor diameter is will decrease during the roto r speed higher in the same output power (see Figure 2). Also, the resu lt can be provided to submarine MEM width and space consideration in t he future design. 3.4 Power and Propulsion 3.4.1 Resistance Estimate Once the size of the submarine has been established and the initial ap pendage geometries determined then a more detailed evaluation of the r esistance/speed relationship deeply submerged may be undertaken as des cribed below. The total resistance coefficient may be expressed in the form: ¡K¡K¡K¡K¡K¡K¡K¡K¡K(3.3) Calculate at series of speeds: ¡K¡K¡K¡K¡K¡K¡K(3.4) The correlation allowance is determined ultimately from First of Class trials for a new submarine. However, at the design and model experime nt stage a suitable value of is assumed, based on data from previous classes of submarines. A typical value of the correlation allowance mi ght between . The following equation can be obtained the correlation a llowance resistance: ¡K¡K¡K¡K¡K¡K(3.5) The viscous resistance can be calculated by the following equations: The friction resistance coefficient for the submarine is calculated u sing an empirical formula (the 1957 ITTC Line): So, the viscous resistance is: ¡K¡K¡K(3.8) The total submarine resistance coefficient having been determined as b elow: ¡K¡K¡K¡K¡K¡K¡K¡K¡K(3.9) The figure 3 shows the Sea-Dragon class submarine bare hull resistance vs ship speed. There are two different algorithms for power calculations. One algorit hm built on the resistance calculation and added a 30¢H margin for app endage drag. The second method was developed by MIT. The following equ ations are presented the effective power for submarine bare hull: ¡K¡K¡K¡K¡K¡K¡K¡K¡K(3.10) Power for appendage resistance is: ¡K¡K¡K¡K¡K¡K¡K(3.11) The effective hull horsepower is: ¡K¡K¡K¡K¡K¡K(3.12) Then we use the MIT method to analysis the effective method power for submarine bare hull as below: For the remaining appendages, use the expression for: ¡K¡K¡K(3.15) So, the effective power by the MIT method is: ¡K¡K¡K¡K¡K¡K¡K¡K¡K¡K¡K¡K¡K(3.16) The author refer the Sea-Dragon class submarine and use above two meth ods to calculate, the figure 3and 4 shows the two methods calculate th e bare hull resistance, effective horsepower and the comparisons for t he two methods. Additional calculations were performed a range of numbers for thrust, thrust horsepower, delivered horsepower, open water deliver horsepower , shaft horsepower, and brake horsepower. Based on above results that we can find the maximum output BHP (5,702hp) and kW (4,253kW) shows in Figure 5. The result almost familiar with the Sea-Dragon Class submar ine burst speed output kW (4300kW). 4. Conclusions Looking at the powering of submarines from the overall design aspect, it is a significant drive of size not only because the propulsion plan t occupies so much internal space, but also because the power/size rel ationship is of a self-perpetuating nature, i.e. the more power that i s sought for, the larger the submarine has to be, which necessitates y et further power, and so on. It consequently behaves the designer to p ress arguments for containing operational ambitions for high underwate r speed (in view of the cube law characteristic of the power/speed rel ationship) as well as aiming to achieve a form and propulsor arrangeme nt suited to good propulsive efficiency. It is with regard to the selection of form that powering is an importa nt influence on the designer. For, as we have seen, although other con siderations may well cause designer to depart quite a way from the ¡§i deal¡¨ form, it is concern about the powering consequences which sets boundaries to the extent of departure. This paper provided new methods of estimate the electrical loads for Navy surface ship and powering needed for submarine in design process. ??Michael C. Robinson (AM), Sara E. Wallace (V),David C. Woodward (A M), and Gene Engstrom (AM), “US Navy Power Transformer Sizing Require ments Using Probabilistic Analysis,” SPS 2005. ??“Design Report Littoral Warfare Submarine,” ATLAS,Fall 2004-Spri ng 2005, Virginia Teach Team.