Stanag 1008 Edition 9 Citation
May 12, 2008 STANAG 1008 (Edition 9, 24 August 2004): Characteristics of Shipboard Electrical Power Systems in Warships of the North Atlantic Treaty Navies. Power Quality and EMC in a Naval Environment. For military naval equipment: STANAG 1008 curve (red) 9.TELECOMMUNICATION ENGINEERING.
This paper investigates the ability of a gas turbine alternator to provide power to an electromagnetic (EM) railgun in surface combatants whilst simultaneously maintaining acceptable levels of power quality for other consumers. Following the justification of the investigation and a description of the research methods, which includes a proposed EM railgun charging control system, this paper discusses simulated results obtained from a validated power system model for both single shot and salvo operations and identifies factors that limit the rates of fire in each case. The findings from the investigations suggest that a large gas turbine alternator is able to maintain quality of power supply within acceptable limits for the majority of EM railgun operations. However, the transient quality of power supply limits was found to be exceeded for heavy firing rates but would remain within tolerable limits as defined for exceptional loads by North Atlantic Treaty Organization (NATO) standardization agreement (STANAG) 1008 Edition 9.
A method to increase the maximum rate of fire whilst maintaining the quality of the power supply within acceptable limits is investigated by increasing the size of the energy storage device and retaining a residual charge. The paper concludes by suggesting that the EM railgun system could be integrated into the ship's electrical system without the need for any additional prime movers to achieve rates of fire commensurate with a surface combatant's requirements.
Introduction Electromagnetic (EM) weapons, including technologies such as high energy lasers and EM railguns, will have a potentially significant impact on a warship's electrical system because they demand high amounts of energy over short periods of time, thus creating a pulse power-demand characteristic (Petersen et al. Navy will test its electromagnetic rail gun aboard DDG 1000. [Online; 2015 July 1].
Available from: 1 July 2015. Recently, the announcement from the United States (US) Navy that an EM railgun may feature on an integrated full electric propulsion (IFEP) warship in the near future (LaGrone LaGrone S. Navy Considering Railgun for Third Zumwalt Destroyer. [Online; cited 2015 February 12]. Available from:.
) has accelerated the need to understand the impact of a high pulse power demand on a warship's electrical power system, so as to facilitate the integration of this revolutionary weapons technology. Previous publications have addressed concerns regarding the impact of pulse power demands on the quality of power supply (QPS), especially to other electrical consumers when configured in IFEP architectures (Kanellos et al. Microsoft Office 2007 Enterprise Fully Activated Rarlab. Kanellos FD, Hatzilau IK, Prousalidis J, Styvaktakis E. Simulation of a Shipboard Electrical Network (AES) Comprising Pulsed Loads.
Engine As A Weapon II. London., Kanellos FD, Tsekouras GJ, Prousalidis J, Hatzilau IK.
An effort to formulate frequency modulation constraints in ship-electrical systems with pulsed loads. IET Electrical Systems in Transportation. 1: 11– 23.; Lewis Lewis EA. Optimising the AC interface of high power pulse loads on combatants with integrated electric propulsion. Engine As A Weapon II.
London.; Sedghisigarchi et al. Sedghisigarchi K, Feliachi A, Fernandes A, Karl S. Gas turbine control and load sharing for shipboard power systems. Power Engineering Society General Meeting, Tampa.; Steurer et al. Smith NS, Butcher MS. Warships weapons and waveforms – practical challenges to meeting quality of power supply requirements on current and future warships. Engine as a Weapon II, London.; Salehi et al.
Salehi V, Mirafzal B, Mohammed O. Pulse-load effects on ship power system stability. 36th Annual Conference on IEEE Industrial Electronics Society, Glendale.; Tsekouras et al. Steurer M, Andrus M, Langston J, Qi L, Suryanarayanan S, Woodruff S, Ribeiro PF. Investigating the impact of pulsed power charging demands on shipboard power quality. Electric Ship Technologies Symposium, Arlington. In general these publications take a simplified approach towards analysing the performance of the power system, with Lewis ( Lewis EA.
Optimising the AC interface of high power pulse loads on combatants with integrated electric propulsion. Engine As A Weapon II. ) employing a predictive method and Kanellos et al. ( Kanellos FD, Hatzilau IK, Prousalidis J, Styvaktakis E.
Simulation of a Shipboard Electrical Network (AES) Comprising Pulsed Loads. Engine As A Weapon II. London., Kanellos FD, Tsekouras GJ, Prousalidis J, Hatzilau IK. An effort to formulate frequency modulation constraints in ship-electrical systems with pulsed loads. IET Electrical Systems in Transportation.
), Steurer et al. And Salehi et al. Employing notational prime mover models. While Sedghisigarchi et al. And Tsekouras et al. Employ improved gas turbine models, the magnitude of the pulse loads investigated does not appear representative of EM railguns.
This paper presents the results of research conducted by University College London (UCL) and supported by Rolls-Royce into the ability of a Rolls-Royce MT30 gas turbine alternator (GTA) to maintain an acceptable standard of power-system performance when meeting the demands of EM railguns. The GTA model employed in this research has been developed and validated using actual engine performance data and has been configured in an electrical power system model that is representative of a warship's power system topology so as to give credible simulation results that aid the understanding of performance. Furthermore, building on past research (Kanellos et al. Kanellos FD, Hatzilau IK, Prousalidis J, Styvaktakis E. Simulation of a Shipboard Electrical Network (AES) Comprising Pulsed Loads. Engine As A Weapon II.
London.; Sedghisigarchi et al. Sedghisigarchi K, Feliachi A, Fernandes A, Karl S. Gas turbine control and load sharing for shipboard power systems. Power Engineering Society General Meeting, Tampa.; Steurer et al. Smith NS, Butcher MS. Warships weapons and waveforms – practical challenges to meeting quality of power supply requirements on current and future warships. Engine as a Weapon II, London.; Salehi et al.
Petersen LJ, Ziv M, Burns DP, Dinh TQ, Malek P. U.S Navy efforts towards development of future naval weapons and integration into an All Electric Warship (AEW). Engine As A Weapon IV.
), this paper presents an EM railgun charging control system that has been designed to control power system transients in a way which ensures that a full charge is always delivered to the EM railgun before a shot is taken. Problem formulation The energy required by the EM railgun projectile is to be provided by the warship's electrical power system, meaning that it must be designed for arduous pulse power conditions. Due to the inherent impedance of the power system and the power limits of current prime movers, it is impractical and unrealistic to draw a 160 MJ, 10 ms pulse of energy demanded by EM railguns directly from the ship's generators (Bernardes et al. Bernardes JS, Stumborg MF, Jean TE.
Analysis of a capacitor-based pulsed power system for driving long-range electromagnetic guns. IEEE Transactions on Magnetics.; Lewis Lewis EA. Optimising the AC interface of high power pulse loads on combatants with integrated electric propulsion.
Engine As A Weapon II. Instead, an intermediate charging circuit can be employed to draw and store energy provided by the GTA in an energy storage device (ESD) from the ship's power system, prior to it being supplied to the EM railgun via a pulse-forming network (PFN; Petersen et al. Navy will test its electromagnetic rail gun aboard DDG 1000.
[Online; 2015 July 1]. Available from: 1 July 2015. The energy transfer for this procedure with the associated stage losses is (1) where is the mass flow rate of fuel, is the inertial energy of the GTA, is the energy stored in the ESD, is the EM energy in the rails of the weapon and is the kinetic energy of the projectile. Stored chemical energy in the fuel is converted into mechanical energy by the gas turbine (GT), which is then converted into electrical energy by means of an alternator. The GT and alternator retain a portion of the energy as rotational energy (inertia). Electrical energy is transferred to an ESD. The energy is stored in a capacitor, completing the first stage of the energy transfer process.
Once the ESD has been charged, energy can be released from the ESD into the EM railgun via the PFN, with the energy being converted into EM energy in the rails of the weapon and kinetic energy in the projectile by means of the Lorentz force (Hodge et al. Hodge CG, Buckingham J, Macalindin A, Amy J. Some practical and integration aspects of electric rail guns. Engine As A Weapon II, London. This forms the second stage of the energy transfer process. Energy is lost in each conversion and transfer stage.
Hence, EM railgun firing can be considered a two-stage energy transfer process. In the first stage, energy is transferred from the GTA to the ESD.
In the second stage, energy is transferred from the ESD to the rails of the weapon via the PFN and ultimately to the projectile. The ESD is considered in Equation (1) to be a capacitor bank but equally it may be a flywheel, battery, superconducting coil or some combination of these, with each solution having its particular advantages, disadvantages and charging and discharging characteristics (McNab McNab IR. Pulsed power options for large EM launchers. 17th International Symoposium on Electromagnetic Launch Technology, La Jolla. The relationship between the first and second stage is the rate of charge of the ESD, which dictates the firing rate of the EM railgun rather than the capacity of the ESD. The charging of the ESD is decoupled from the firing of the EM railgun, meaning that the ESD cannot be connected to the generator and the PFN at the same time.
This arrangement has several advantages and disadvantages. A significant advantage is that the generator is not exposed to the high pulse demand of the EM railgun directly but instead experiences the arduous task of charging the ESD before the firing of the EM railgun and recharging the ESD before subsequent firings. The rate of charge of the ESD by the GTA essentially governs the firing rate of the EM railgun – hence, a rapid charge is desirable. The disadvantage is that the GTA will be subjected to a load change after each ESD (re)charge and following each EM railgun shot. Energy transfer is therefore split into two distinct stages; the ESD ‘charging stage', during which energy is transferred from the ship's power system to the ESD, and the ‘firing stage', during which energy is transferred from the ESD in order to fire the projectile. In this paper it is the first part of this transfer that is of interest, since the charging of the ESD impacts the design and operation of the ship's electrical power system. When sustained continuous or salvo firing is required it might be desirable for the GTA to continuously charge the ESD for the duration of the salvo, thereby maintaining the GTA at its maximum output power.
Under such conditions the generator would supply the maximum rated power during the charging of the ESD and when firing the EM railgun. Integrating EM railguns with electric warship power systems Consider the electric warship power system represented in, the characteristics of which have been selected based upon the power and propulsive requirements of a candidate warship (LaGrone LaGrone S. Navy Considering Railgun for Third Zumwalt Destroyer.
[Online; cited 2015 February 12]. Available from:.
The system incorporates a power generation capability comprising 2 × 36 MW GTAs and 2 × 5 MW GTAs which can be connected to a common bus. Transformers step down the main bus voltage from 11 kV to 440 V for the ship and ancillary service loads.
With a total installed power of 82 MW the vessel is able to achieve approximately 30 kt whilst maintaining ship and ancillary service loads totalling 4 MW in the Action State 1 (four island) configuration with prime mover availability maximized and the ship's vulnerability to power loss minimized. Integrated into this electrical power and propulsion system is an EM railgun with its associated charging system.
The parameters of the EM railgun have been derived from Petersen et al. Navy will test its electromagnetic rail gun aboard DDG 1000.
[Online; 2015 July 1]. Available from: 1 July 2015. ), Lewis ( Lewis EA. Optimising the AC interface of high power pulse loads on combatants with integrated electric propulsion.
Engine As A Weapon II. ), Hodge et al.
( Hodge CG, Buckingham J, Macalindin A, Amy J. Some practical and integration aspects of electric rail guns. Engine As A Weapon II, London. ), Wolfe et al. ( Tsekouras GJ, Kanellos FD, Prousalidis JM, Hatzilau IK. STANAG 1008 design contraints for pulsed loads in the frame of the all electric ship concept.
Nausivios Chora:3: 115– 154. ) and Andrews et al. ( Andrews DJ, Bucknall R, Pawling R. The impact of integrated electric weapons on future warship design.
International Naval Engineering Conference 2010 The Affordable Future Fleet, Portsmouth. ), which all offer correlating specifications (summarized in ). The ESD can be charged from either the port or starboard high voltage (HV) propulsion switchboard, providing a level of redundancy. When the ship's power system is configured as shown in each prime mover generator supplies its own independent electrical distribution system. The 5 MW GTAs supply the ship's low voltage (LV) load and one 36 MW GTA provides power for propulsion whilst the other 36 MW GTA provides power for the EM railgun. The rectifier considered to regulate the ESD charge cycle is the six-pulse fully-controlled thyristor bridge.
This type of rectifier can be connected directly to a warship's three-phase electrical power system without the need for bulky transformers, although isolation transformers may be required for other reasons. Importantly, it uses thyristor semiconductor devices which are available as high-power robust devices which have a low voltage drop in the on state. This means they have lower conducting losses when compared to other types of power electronic devices such as insulated gate bipolar transistors (IGBTs) Bradley ( Bradley DA. Power semiconductors. Power electronics. Boca Raton: CRC Press; p. Furthermore this type of converter is simple to control and has a low device count, meaning that it is power dense (Lorenz Lorenz L.
Key power semiconductor devices and development trends. Hindi Typing Test Book In Pdf more. International Conference on Electrical Machines and Systems, Wuhan. However, the six-pulse fully-controlled rectifier introduces harmonic waveform distortion on the AC supply side during the charging of the ESD, in particular 5th and 7th harmonics at 20% and 14% of the fundamental respectively, which would need to be filtered to meet North Atlantic Treaty Organization (NATO) standardization agreement (STANAG) 1008 Edition 9 (NATO NexansAmerCable. NexansAmerCable marine medium voltage cables. [Online; cited 2015 January 14]. Available from: 27 January 2015. ) harmonic waveform distortion limits.
From the GTA's perspective the load characteristic of the six-pulse fully-controlled thyristor bridge is important with regard to the power factor (PF). Consider the input real ( P) and apparent ( S) power, defined as: (2) (3) Hence by taking the ratio of P and S, the PF is defined as: (4) Since the input displacement factor of the fundamental current and the characteristic current drawn (ignoring overlap) is a quasi-square wave defined as (5) the PF of the rectifier can be simply defined in terms of the product of the ratio of fundamental current to total current drawn and the cosine of the firing delay angle, hence the input displacement factor depends on the firing delay angle ( α).
This means that at large firing delay angles the PF is low and the rectifier predominantly draws reactive current from the generator, whilst at small firing delay angles it predominantly draws real current. This specific characteristic impacts directly upon the required automatic voltage regulator (AVR) and governor responses, as these control the reactive and real power delivery from the GTA respectively. The ESD is defined as a capacitor with an energy storage capacity of 160 MJ, which is the overall required energy per shot (). A capacitive ESD was selected based on the results of the simulation-based (Bernardes et al.
Bernardes JS, Stumborg MF, Jean TE. Analysis of a capacitor-based pulsed power system for driving long-range electromagnetic guns. IEEE Transactions on Magnetics.; Wolfe et al. Tsekouras GJ, Kanellos FD, Prousalidis JM, Hatzilau IK. STANAG 1008 design contraints for pulsed loads in the frame of the all electric ship concept. Nausivios Chora:3: 115– 154.
) and experimental (McNab et al. McNab IR, LeVine F, Aponte M. Experiments with the green farm electric gun facility. IEEE Trans Magn. 31: 338– 343., ) research, which has demonstrated that capacitor-based ESDs are capable of powering EM railguns.
The size of a capacitive ESD (as detailed in Bernardes et al. Bernardes JS, Stumborg MF, Jean TE. Analysis of a capacitor-based pulsed power system for driving long-range electromagnetic guns. IEEE Transactions on Magnetics.
) is 160 m 3, which with a mass of 210 t is acknowledged as being too large and heavy when considered for installation on a warship not designed to field EM railguns from the outset – but the energy densities of capacitors are improving. The charging characteristics of any capacitor are (6) and since I = d Q/d t, (7) where V DC is the DC supply voltage, C is the capacitance and R is the equivalent series resistance of the capacitor plus any line resistance. Hence the charge level and current depends upon the DC voltage applied to the capacitor. The rate of charge depends upon the time constant RC, hence the most rapid charge is achieved with an appropriate balance of storage capacitors connected in series (to minimize C) and connected in parallel to minimize the equivalent series resistance R. In practice this paradox is resolved by the maximum available voltage V DC, which in turn depends upon the GTA voltage and the maximum allowed current; also, the equivalent series resistance tends to be small. The output voltage from the rectifier is defined as (8) where V o is the mean DC output voltage, V max is the maximum voltage and V RMS is the root mean square (RMS) voltage of the supply.
The charge of the capacitor is dependent upon V DC, which is approximately the mean of V o, hence (9) and (10) As such, the capacitor charge and its associated current are dependent upon the firing delay angle α. Therefore as the GTA must operate within both its P-Q capability curves and its voltage and frequency stability limits, the control of GTA when charging is facilitated as follows: the real power demand is controlled by the governor, the reactive power is controlled by the AVR and the current drawn is controlled by the rectifier firing angle delay, which also dictates the charge rate of the ESD. Operational context It should be recognized that the quality of the electrical power generated on board a surface combatant is governed by QPS standards, taken in this paper to be NATO STANAG 1008 Edition 9. The compliance with said standards ensures compliance with and proper operation of the electrical machines and equipment connected to the supply.
Consider the case shown in, whereby it is assumed the starboard 5 MW GTA is offline due to damage or a fault. While it may be argued that the port 5 MW GTA could supply the full LV ship and ancillary service load, this would mean connecting all said loads to a single generator, the loss of which would mean a full loss of the ship and ancillary service loads. To avoid this undesirable scenario, the power system is reconfigured in 3 island mode (see ). It now becomes apparent that any impact the operation of the EM railgun may have on the QPS will manifest at the LV service bus as well as the main HV bus. As such, the EM railgun should only be operated so as not to exceed the QPS constraints of NATO STANAG 1008, summarized in.
Operating within these constraints maintains the integrity of the supply to both the main HV bus and the LV service loads, under certain scenarios. The STANAG 1008 QPS limits presented in only specifically govern the LV, or 440 V and 115 V supply of a warship and, as electric weapons were not a factor when these standards were defined (Smith and Butcher Smith NS, Butcher MS. Warships weapons and waveforms – practical challenges to meeting quality of power supply requirements on current and future warships. Engine as a Weapon II, London. ) do not make provision for EM railguns.
With regards to pulse loads in general, STANAG 1008 stipulates that such loads with a real and reactive power demand exceeding 6.5% and 25% respectively of the rated apparent power supply should be avoided unless corrective action be determined through consultation with the design authority, guidance on which is one of the key aims of this research. For an electric weapon it is necessary to design the electrical system to operate within defined standards and as no specific rules and regulations for electric weapons currently exist, acceptability will be considered against all three deviation levels detailed in, which makes for a useful starting point.
Furthermore, the results of this research can be considered a basis from which to formulate an HV standard, taking into account the operation of EM railguns. Modelling To investigate the capability of the EM railgun in terms of the rate of fire when operating within the QPS constraints of STANAG 1008, the GTA EM railgun performance modelling tool was developed, the top-level system model being shown in. The model consists of a 36 MW GTA coupled to an EM railgun ESD via a six-pulse thyristor charging bridge and 100 m of power cable, which because of practical constraints is divided into two 50 m cables either side of an isolation transformer. The cable length allows for the GTA to be installed midship, with the EM railgun installed towards the bow, in a vessel approximately 200 m in length. Each of the individual component parts are described further in the subsequent sections.
So as to produce credible results, the response of the GTA governor and AVR models have been validated, the results of which are given in Tables and respectively, the actual results for which can be found in Appendices 1 and 2. For the case of the governor response, validation was conducted against the results obtained from a Rolls-Royce MT30 engine model, which itself was validated against real-world testing that included load stepping from 2 to 36 MW. The AVR's transient response was validated against the load rejection section of the Lloyd's Register ( Lloyd's Register. 9.4 Generator control, Lloyd's Register.
) generator control testing procedure. EM railgun ESD The railgun ESD is based on the capacitor characteristics presented in Wolfe et al.
( Tsekouras GJ, Kanellos FD, Prousalidis JM, Hatzilau IK. STANAG 1008 design contraints for pulsed loads in the frame of the all electric ship concept.
Nausivios Chora:3: 115– 154. The ESD is composed of 54, parallel-connected 2.96 MJ capacitor banks, each with a capacitance of 48.96 mF and an equivalent series resistance of 541 µΩ. Each capacitor bank is charged to 11 kV, which is equal to the GTA RMS voltage. The overall capacitor bank is modelled as a single capacitor with a capacitance of 2.64 F and an equivalent series resistance of 10.02 uΩ. The control system set point sets the ESD capacitor voltage required to store the full charge, which in this case is 11 kV when α = 42 deg. The rate of rise of capacitor voltage is restricted by the rate limiter so as to limit the rate at which the GTA transfers energy to the ESD.
The aim of this is to control the GTA load transients so as to maintain the QPS within STANAG 1008 limits. If said limits were relaxed then the generator capability curves would dictate the limit. The control system then computes the thyristor bridge firing angle α as follows: (11) where V DC is the voltage set point and V AC is the supply side AC voltage. The actual ESD capacitor voltage feedback firing angle is computed and subtracted from the set point firing angle to create an error signal. This error signal is then processed through a PI controller before being added back to the set point firing angle via the feedforward control loop.
The output firing angle is then used to fire the ESD charging bridge. Isolation transformer While it is acknowledged that the isolation transformer would decrease the overall power density of the EM railgun system, it is considered a simple and robust method of achieving harmonic attenuation by adding line impedance, performing much the same function as the line reactance employed in Kanellos et al. ( Kanellos FD, Hatzilau IK, Prousalidis J, Styvaktakis E. Simulation of a Shipboard Electrical Network (AES) Comprising Pulsed Loads.
Engine As A Weapon II. ) and Steurer et al. ( Smith NS, Butcher MS. Warships weapons and waveforms – practical challenges to meeting quality of power supply requirements on current and future warships. Engine as a Weapon II, London. Furthermore it provides galvanic isolation between the GTA and the EM railgun charging system. The key transformer model parameters are summarized in.
Constraints and modelling assumptions The following constraints were applied during the investigation. • The EM railgun operation commences with zero stored energy in the ESD, which is considered to be inherently safe. • It is accepted that propulsion power will be sacrificed when operating the gun for significant periods and that top speed will not be achievable, as only one GTA will be providing propulsive power; however, with one GTA dedicated to propulsion it should still be possible to operate the gun from 0 to approximately 25 kt.
• At the end of the EM railgun operation minimal excess energy should be generated by the GTA; this means that there is a lower amount of energy that needs to be dissipated following a shot, which in effect increases the power density of the overall gun system. • The EM railgun shot time is negligible when compared with the ESD charge time.
Figures and demonstrate that the minimum ESD charge time, in order to remain within QPS limits for both voltage and frequency, is 8.25 s, at which time both the voltage and frequency deviations are within the maximum transient limits. For both the voltage and frequency to remain within the transient tolerance limit, the minimum charge time is 9.77 s. Once the ESD has been charged and the shot fired, the GTA must unload, as it is now only supplying the 2 MW LV service load. Shows the voltage deviations for unload ramp times of 2 to 10 s, all of which are within the QPS tolerance limit. To examine the response of the system further, a single charge and unload time based on the QPS compliance curves in Figures – was selected.
The charge time was set at 8.75 s, which maintains the voltage deviation below the maximum transient tolerance and the frequency deviation inside tolerance. For this example case the unload time was set to 4.25 s, which maintains both the voltage and frequency deviation within tolerance. While it is acknowledged that the GTA can unload more quickly than this whilst maintaining the QPS, it was thought reasonable to exploit the initial steep section of the curve, which allows the QPS to remain within tolerance for a relatively quick unload time when compared with the charge time. The charge and unload simulations were then coupled together to demonstrate the power-system performance characteristics over a complete firing cycle, the results of which are shown in. The charging cycle begins at 5 s to allow the simulation to settle to steady state beforehand. Upon commencement of charging there is an immediate demand for charging current, which as can be derived from the GTA real and reactive power plots is predominantly reactive. As such, the GTA power factor is very low and may exceed the GTA field current heating limit.
This is due to the previously discussed characteristics of the thyristor bridge and must be controlled by the AVR. Also owing to the characteristics of the thyristor bridge, high levels of harmonic distortion are introduced throughout the EM railgun operation, as shown in the supply side total harmonic distortion (THD) plot. During the charging cycle the AVR maintains the voltage within the QPS maximum transient tolerance, while the governor controls the mass flow rate of fuel to maintain the frequency within the QPS tolerance, as derived from and respectively.
During the GTA unload the AVR and the governor maintain both the voltage and the frequency within QPS tolerance, as derived from and. It should be noted that due to the nature of EM railgun operations the previously described impacts are quasi-periodic. The results presented thus far have demonstrated that to maintain both voltage and frequency deviations within STANAG 1008 limits whilst operating an EM railgun, a minimum charge time of 8.25 s and a minimum unload time of 2.25 s must be imposed. This would allow one shot to be fired every 10.5 s, or at a rate of 5 shots per min in a sequence of single shots. While this may suffice if a single shot or a low rate of fire is required, there is also a requirement to fire at a rate of 10 rounds per min, or one shot every 6 s (Osborn Osborn K.
Navy will test its electromagnetic rail gun aboard DDG 1000. [Online; 2015 July 1]. Available from: 1 July 2015. Owing to the fact that the GTA can deliver 37.6 MJ per s (when at 110% overload and when allowing for the 2 MJ per s service load) this required rate of fire should be achievable, given that the GTA should be able to supply the required 160 MJ in approximately 5 s, allowing margin for transmission losses.
This assumes that the GTA can supply full power for the duration of the charging sequence and maintain this over multiple shots. However, once a shot is taken and the ESD has been discharged to the EM railgun, the capacitor is empty, or perhaps practically empty.
As a result, when the capacitor is reconnected to the GTA for recharge a large inrush current, at levels approaching those similar to a short circuit fault, results. This is demonstrated in, which shows the GTA current following an attempt to immediately recharge the ESD after the shot at 13.75 s instead of unloading the GTA as in the previous scenarios. The inrush current at the level demonstrated in could trigger the protection system into tripping the GTA.
A further problem with this recharging method is that immediately after the ESD is discharged the capacitor voltage reduces to zero. As the power demand is a function of the capacitor voltage and charging current, the GTA attempts to shed load back to supplying only the 2 MW LV service load. A method of overcoming this problem is to oversize the capacitor so that when a shot is fired it does not completely discharge. This means that when the ESD is reconnected to the GTA for another charge it is not empty, but contains an amount of residual charge. The advantages of this are fourfold. Firstly, the residual charge in the capacitor reduces the inrush current at the point of recharge. Secondly, as the capacitor does not fully discharge following the shot the voltage does not reduce to zero, thus the GTA retains load, viewing the discharge as a step shed in load, rather than a complete load shed.
Thirdly, the supply side current harmonics are substantially reduced due to a smaller firing delay angle at discharge. Fourthly, the recharging of the capacitor is achieved by a combined real and reactive current from the start of the recharging process.
In order to assess the extent to which retaining residual charge can limit the GTA load transients, a series of increasing step load sheds were taken and their corresponding over frequencies recorded, the results of which are shown in. This plot was then used to predict the frequency rise following the GTA step down in load, taken immediately before recharging the ESD.
The load steps were taken in 1 MW increments ranging from 1 to 13 MW, beyond which model restrictions prevent further validated test data being obtained. The voltage transients were found to be within tolerance for the full range of load sheds, as shown in Appendix 3. Consider now, firing a salvo of shots at the previously discussed requirement of one shot every 6 s. In order to maintain QPS within the transient tolerance limit, a maximum load shed of 13 MW may be taken between shots (see ). For the purposes of the simulation a 12 MW step shed was selected, so as to remain within the envelope of validated performance. In order to allow the 12 MW load shed between shots, a total ESD capacity of 320 MJ must be employed (see ).
The corresponding capacitor characteristics can then be calculated based on the data from Wolfe et al. ( Tsekouras GJ, Kanellos FD, Prousalidis JM, Hatzilau IK. STANAG 1008 design contraints for pulsed loads in the frame of the all electric ship concept. Nausivios Chora:3: 115– 154. In order to examine the actual response of the system under this scenario, the firing of three shots at a rate of one shot every 6 s was simulated, the results of which are shown in. As the capacity of the ESD is now double that of the original, the initial charge time was increased from 8.75 s to 17.5 s, assuming a linear relationship between the energy delivered to the ESD and the charge time.
The unload time was set to 4.25 s, as per the previous scenario. Commensurate with, upon commencement of charging there is an immediate demand for charging current, which as it can be derived from the GTA real and reactive power plots, is predominantly reactive. Again, the characteristics of the thyristor bridge introduce high levels of harmonic distortion throughout the EM railgun salvo operation, with the GTA supplying a significant amount of reactive power throughout, which as per the single-shot scenario may periodically exceed the GTA field current heating limit. Once the ESD is charged, at 22.5 s a shot is fired and, as by design, the GTA takes a 12 MW load shed before recharging the ESD for the next shot. While the ESD now contains residual charge at the point of recharge, an inrush current still results, as shown in the GTA RMS current plot. This inrush current also manifests in the GTA real and reactive power plots as short-lived spikes of power. However, when compared with, the inrush current at the point of recharge is much reduced, being 1.5 times the full load current as opposed to 3.75 (see ).
This characteristic is then repeated following the 6 s recharge at 28.5 s when the second shot is fired. The GTA then recharges the ESD for a third time, after which the third shot is fired and the GTA unloads to the 2 MW LV service load. It should be noted that the frequency deviation at the point of recharge is less than anticipated, remaining inside the tolerance limits rather than inside the transient limits. This is because instead of settling to a steady state following the step shed as in, the GTA immediately recharges the ESD and thus the governor immediately increases the mass flow rate of fuel to the GTA, meaning that the frequency is recovered more quickly. Recharging the ESD immediately following a step change is therefore shown to be less extreme than a pure step change. Furthermore, when recharging the ESD between shots the GTA does not attain full power or deliver the same charging current as for the initial charge, which coupled with the better than expected frequency deviation suggest that the GTA may be able to enable a greater rate of fire than 10 rounds per min whilst remaining within STANAG 1008 limits. When firing a salvo of shots, a further design trade-off was introduced due to the need to manage the ESD inrush current and the GTA load shed at the point of recharge.
As demonstrates, the deviation in GTA frequency at the point of recharge can be limited by limiting the allowable load shed immediately before recharge. However, as shown in, limiting the maximum allowable load shed requires an increase in the capacity of the ESD so as to retain residual charge between shots, which is also shown to limit the ESD inrush current. In order to meet the required rate of fire of 10 rounds per min, the capacity of the ESD was increased to 320 MJ so as to maintain the QPS within transient limits. While, as shown in, this method enabled three shots to be fired at the required rate of fire, it is acknowledged that the increase in the size of the ESD is significant and at the current energy density of 1 J/m 3 and mass density of 1230 kg/m 3 (Bernardes et al. Bernardes JS, Stumborg MF, Jean TE. Analysis of a capacitor-based pulsed power system for driving long-range electromagnetic guns.
IEEE Transactions on Magnetics. ) would mean an increase of 320 m 3 and 394 tonnes. As such, would mean ESD of this capacity may not be practical to install on an electric warship, based on current energy and mass density. In addition to this, the initial charge of the 320 MJ ESD takes 17.5 s, which would in practice limit the shots fired in the first min to six. So as to enable both single shots and salvos to be fired, an ESD configuration is suggested whereby two 160 MJ capacitor banks are connected in parallel to form a single 320 MJ bank.
Employing such a configuration would allow single shots to be fired by utilizing a single bank, while both banks could be connected in parallel to enable salvo firing (see ). In opting for the low-loss thyristor bridge it was accepted that waveform quality would be compromised. As shown in and the supply side THD during EM railgun operation peaks at 30%, which is outside of STANAG 1008 QPS limits. Such high levels of THD could lead to classical problems associated with harmonic distortion, such as generator overheating caused by increased harmonic loss in the stator windings, rotor circuits and rotor laminations, and interference with protection, control and communications systems (Arrillaga and Watson Arrillaga J, Watson NR. Effects of harmonic distortion. Power system harmonics.
Chichester: John Wiley & Sons; p. 143– 189.; Bucknall Bucknall RW. On the modelling of hamronics in marine electrical propulsion systems. All Electric Ship 2007 The Vision Redrawn, London. Furthermore, the high level of harmonic distortion present during EM railgun operations impacted on the GTA power factor, owing to the significant amount of reactive power drawn at both fundamental and harmonic frequencies present in the power system. This reactive power demand must be controlled by the AVR and is accepted as a design trade-off. It should be considered that the operation of the EM railgun is quasi-periodic; therefore, the power system would only be subjected to this level of THD when the EM railgun is in operation and not when the system is at steady state.
Conclusions The aim of this paper was to investigate the impact of EM railguns on the performance of a candidate electric warship power system. The research problem posed is, can an existing GTA design meet the operational needs of EM railguns whilst maintaining satisfactory QPS as defined by STANAG 1008? The approach taken to answer this research question was to investigate the operation of the EM railgun when connected to a candidate ship's power system.
The results of the investigation have shown that maintaining the QPS within STANAG 1008 limits places rate-of-fire constraints on the EM railgun. The limits constrain the rate of fire to one shot every 10.5 or 12.77 s when remaining within maximum transient and transient limits respectively. This research went on to demonstrate that the rate of fire can be increased to the required 10 rounds per min by increasing the size of the ESD from 160 MJ to 320 MJ, so as to limit the GTA load transients and maintain the QPS within STANAG 1008 limits. Furthermore, based on the results of the GTA performance presented in, it is suggested that the rate of fire could be increased to above 10 rounds per min while still maintaining an acceptable level of QPS.
In practice the size of the ESD and its discharge depth will be dictated by the pulse power requirement, as demanded by the EM railgun, i.e., the amount of energy and the required discharge rate. In this research an isolated six-pulse thyristor bridge was used for reasons of robustness and efficiency, yet it is appreciated that the high harmonic levels need to be reduced, which may require heavy filtering or a higher pulse number converter. Nevertheless the charging rate and generator current control was achieved with the approach adopted. An ESD configuration, based on the parallel connection of capacitor banks in order to maintain both single-shot and salvo capability was also suggested.
However, based on the current energy density of capacitors, such a system may not be practically realizable on an electric warship and, as such, the size of the ESD should be considered in terms of the extent to which QPS limits must be adhered to. Finally, it was concluded that compliance with the current STANAG 1008 standard may constrain the operation of an EM railgun to a rate of fire less than that achievable from an energy-transfer perspective and thus may reduce its operational capability.
As such, the applicability of the standard during the quasi-periodic operation of EM railguns should be considered and the standard revised accordingly.