Battery Charging Redesign for Large Scale Vehicle 2, …



University of Idaho Electrical and Computer Engineering – Moscow, IDNavy Acoustic Research Detachment – Bayview, IDBattery Charging Redesign for Large Scale Vehicle 2, CUTTHROATLSV2 Battery Charging OptimizationChristopher Douglas, James Randall, David Hookerautovolt@uidaho.edu12/15/2008AbstractTeam AutoVolt was formed in order to study alternate charging schemes for recharge of the all electrical 1/3 scale Virginia class submarine dubbed the Large Scale Vehicle 2 (LSV2), propulsion batteries in order to increase battery life.Increased performances will be achieved by improving capacity retention throughout the batteries’ life through implementing new charging algorithms. The schemes (discussed in section 4) are designed to prevent both undercharge and overcharge, thus decreasing degradation of the batteries. These separate schemes also include other benefits elaborated upon in their respective sections, which extend the life of the batteries. Battery life will also be increased through automation of the charge. Automation will also result in better performance and lower risk of error in the system. The current manual charging done is performed by technicians, leaving room for human error in adjustments and lack of adequate observation, both of which can lead to battery degradation. Further disadvantages of the existing system will be outlined in section 1. Useful definitionsVRLA- Valve Regulated Lead Acid batteriesCI- Current Interrupt overcharge method CC/CV- Constant Current / Constant Voltage recharge methodZDV- Zero Delta Voltage overcharge termination methodORE- Oxygen Recombination EfficiencyEOL- End of Life BOL- Beginning of LifeVCCS-Voltage Controlled Current Step Down Charge (also see FC)FC- Fast Charge charging methodPSOR-Partial State of Recharge charging methodOC- Over ChargeLSV2- Large Scale Vehicle 2ARD- Acoustic Research DetachmentNSWCCD- Naval Surface Warfare Center Caderock DivisionAh- Ampere HoursTable of ContentsContents TOC \o "1-3" \h \z \u 1.Background PAGEREF _Toc220718530 \h 51.1.Motivation for Work PAGEREF _Toc220718531 \h 51.2.Identification of Need PAGEREF _Toc220718532 \h 61.3.Expected Benefits PAGEREF _Toc220718533 \h 72.Problem Definition PAGEREF _Toc220718534 \h 72.1.Goals and Deliverables PAGEREF _Toc220718535 \h 72.2.Specifications and Constraints PAGEREF _Toc220718536 \h 83.Project Plan PAGEREF _Toc220718537 \h 84.Concepts Considered PAGEREF _Toc220718538 \h 94.1.Zero Delta Voltage (ZDV) PAGEREF _Toc220718539 \h 104.2.Current Interrupt (CI) PAGEREF _Toc220718540 \h 104.3.Partial State of Recharge (PSOR) PAGEREF _Toc220718541 \h 114.4.Fast Charging PAGEREF _Toc220718542 \h 124.5.Catalyst Caps PAGEREF _Toc220718543 \h 145.Concept Selection PAGEREF _Toc220718544 \h 166.Future Work PAGEREF _Toc220718545 \h 17References PAGEREF _Toc220718546 \h 19*due to the fact that the all of the work done to date is research, this section is omitted from this report. BackgroundThe Navy Acoustic Research Detachment operates a Large Scale Vehicle (LSV2) submarine in Lake Pend Oreille to gather acoustic research data. This all electric acoustic research submarine consists of 1680 VRLA primary batteries used for the propulsion system batteries. The 2V VRLA batteries are split into four parallel strings of batteries, summing to a total 840V delivered to the propulsion system. 12192001236345In order to charge the batteries, seven chargers are connected to four parallel strings of 60 batteries. The cabling running from the chargers to the submarine are rated for 50A and main switches are located inside the submarine to reconfigure the batteries for charging or for an underway. The diagram below displays visually how the chargers are connected. Figure 1.1Motivation for WorkThe existing charging system for the LSV2 is suboptimal. The charge cycle is run by technicians who check voltage and current levels every 15 to 20 minutes, which may result in damaging over voltage portions of the charge. The charging routine is explained below and graphically represented in Figure 1.mence charging at initial charging rate of 45A per string (180A total).Continue until battery voltage increases to 2.35V per cell for each string.Maintain 2.35V per cell by reducing the charging rate as necessary and allow the charging current to taper to 6.25A per string (25A total).666751118235Maintain the charging rate for a period of 3 hours and until a minimum of 10% overcharge is achieved based on the ampere-hours discharged from the battery. At no time during the charge should individual cell voltage be allowed to increase above 2.50Figure 1. SEQ Figure \* ARABIC 2As stated in 4, the overcharge portion of this routine is a constant voltage time limited overcharge, which allows possible overcharge of the batteries, leading to a considerably shorter lifespan.Identification of NeedThe useful life of the propulsion energy source, made up of the 1680 VRLA batteries, is replaced every four years. The cost to replace the entire system is roughly $593,000 plus labor cost and 2 months of downtime where the submarine is out of commission.Expected BenefitsOne expected benefit of this project is the reduction of capacity loss over the batteries’ service life. If the proposed solutions are successful, they will allow the battery to maintain a high State of Charge (SOC) throughout its cycle life. This allows the LSV2 to maintain longer data-gathering runs (“underways”) throughout system life. More benefits will be achieved through automation which must be implemented due to the complexity of the proposed solutions. Automation will solve the problems linked to human charge addressed in section 1.1 and will allow the technicians to work on other tasks while the charge is in progress. Finally, the most important benefit will be the extension of the useful life of the batteries, which will reduce expenses over the long term and reduce submarine downtime, resulting in more testing time per year. There are several other benefits specific to individual charging algorithms to be discussed in their respective sections.Problem DefinitionLSV2 is an all electric 1/3 scale Virginia class submarine operated by the Acoustic Research Detachment (ARD) of the Naval Surface Warfare Center, Caderock Division (NSWCCD) to study acoustic properties of propulsion systems. The current system’s battery charger configuration and charging scheme were inherited hardware from an earlier craft, the LSV1. The ARD would like to optimize and improve battery capacity over the course of the main batteries’ life cycle, and possibly reduce charge time by improving the method used to charge the system.Goals and DeliverablesMaintaining the battery capacity will solve two main problems. First, if successful, we will be able to offer the ARD longer mission run times as the batteries approach the 4 year expected end of life. Secondly, if battery life can be extended by what our research states to be around 50%, the savings to the ARD would be approximately $50,000 per year in battery costs alone. This cost does not take into account the 2 months of downtime while batteries are being replaced. This also does not take into account the man hours required to accomplish the replacement.The current algorithm used to charge the LSV2 utilizes a constant current (CC) charge followed by a constant voltage (CV) charge in order to restore the 1680 Valve Regulated Lead Acid (VRLA) batteries to a full charge state. This charging scheme has been the standard since VRLA batteries were introduced in industry. Recent research by several others experienced in the study of VRLA battery life has been conducted to find better ways of performing a charge and this is the focus of the project.Our research has pointed us to several promising charging algorithms that will be tested in the course of this project. Among these are Zero Delta Voltage (ZDV) and Current Interrupt (CI). We will be testing these charging algorithms to determine what benefit would result in charging the batteries using a new scheme. We are also looking into using a technology called catalyst caps on new batteries that would potentially extend the battery life regardless of the algorithm used to charge the batteries. Specifications and ConstraintsThe specifications for this project will include documenting the current charging configuration to show advantages and disadvantages. We will be researching potential changes to extend the capacity retention. Laboratory tests will be conducted to verify benefits of proposed charging systems.Proposed solutions are subject to several constrains. The first constraint is that any solution must not reduce the battery capacity, which would lead to shorter mission run time. Secondly, we are limited by the facility hardware and existing infrastructure. Each charger can only supply 45 amps per battery string, limiting the charging options that can be used.Project PlanTeam AutoVolt investigated how the current system was configured to determine what the strengths and weaknesses were. A visit to the facility enabled us to comprehend the size of the project and what items would impede our progress towards completion. Most of the fall semester was spent researching what methods for charging VRLA batteries were available and evaluating them in lift of the constraints of the infrastructure to determine feasibility. Continued research was conducted on potential alternate charging methods, looking at how they could be tested and modeled. To reduce cost and the amount of physical lab testing, the team utilized part of the fall semester to research methods to simulate VRLA battery capacity over the course of their intended life. Researching software modeling has proven fruitless, as the simulation of VRLA batteries with respect to their capacity retention and state of health over expected cycle life has been found to be non-existent. Because of this, the only option available to test actual maintenance of performance is to physically test the possible charging schemes against a control of the current charging algorithm. In order to test the algorithms, team AutoVolt will be designing and constructing a battery cycling testing apparatus to accomplish this task. Figure 3.1 shows progress of the team through the end of this semester. Once the research portion -5238751495425was completed, we moved into conceptualizing the ideas into feasible and non-feasible paths. Figure 3.1Concepts ConsideredWe have researched several new charging schemes for the VRLA battery. These include Zero Delta Voltage (ZDV), Current Interrupt (CI), Partial State of Recharge (PSOR) and Fast Charging. When analyzing each charge method, the potential gain, cost/benefits, and feasibility were analyzed to determine the likelihood of success.Zero Delta Voltage (ZDV)ZDV [1] charging is simply a method of terminating the system charge at the correct point during the charging cycle. This will prevent undercharging or overcharging of the battery cells, which in turn prevents battery performance degradation. Using this charge method requires that the batteries be charged at maximum current until 70% of the previously discharged ampere-hours have been returned to the battery string. The batteries are then charged at 20% of max current until ZDV is reached.In order to determine ZDV, voltage measurements are taken on the battery string every second for 30 seconds. This value is averaged and then subtracted from the previous 30 second average. This new value is then compared to a limit that will be determined during the testing phase. If the limit is not exceeded for five consecutive iterations, then the charge is complete. Figure 4.1.1 shows an example of how ZDV is detected in the charging cycle.ΔVoltageFigure SEQ Figure \* ARABIC 4.1.1Current Interrupt (CI)Current Interrupt (CI) [1] is a charging algorithm that is used during the overcharge portion of the charging cycle. As stated in [1], research has indicated that CI is typically used after a fast charge algorithm has restored 100% of the depleted amp-hours. At this point, the batteries are charged using a pulsed current at approximately one fourth of the batteries rated current. Pulse durations are set to a 15 seconds on and 20 seconds off cycle as can be seen in Figure 4.2.1. This charge continues until batteries reach 10% overcharge. . . . Figure 4.2.1Current interrupt overcharging has several advantages. CI can help reduce the gassing effect. This reduction will prolong the battery life by retaining the chemical composition of the battery. During the overcharge phase of charging, a certain level of current is required solely to maintain the oxygen recombination reaction. High pulsed currents will still feed this cycle but will also provide enough current for continued battery charging. Pulsing the current allows for a cooling period which will minimize thermal degradation and gives time for chemical reactions to stabilize resulting in increased charging efficiency. As referenced in [1], CI charging was not tested independently of a fast charge algorithm. Laboratory research will be conducted to determine the stand alone benefits of using a CI charging algorithm.Partial State of Recharge (PSOR)PSOR [2] is a charging scheme that limits the amount of overcharge delivered to the batteries. Using PSOR, batteries are charged to 100% capacity for approximately 9 charging cycles. The tenth cycle is used as a conditioning charge, charging to 120% of nominal capacity. PSOR has been shown to extend the operating life of VRLA batteries by a significant amount by limiting the overcharge the batteries receive. Reducing the charge delivered to the batteries will have the negative effect of reducing operation run time on the LSV2. For this reason, PSOR will not be considered as a viable candidate for further testing and is not included in section 5.Fast ChargingFast charging, also known as voltage controlled current step down charging, provides large current pulses on the order of 4C where C is the 6 hour Ah rating of the battery. During these large current impulses the cell voltage is monitored and compared to a voltage limit, usually about 2.5V per cell. This voltage measurement must also take into consideration the voltage drop caused by large amounts of current flowing through the batteries’ internal resistance estimated at the given SOC at which the measurement is taken. This charging algorithm is further explained and developed in [3] and is summarized below. The current impulses follow the maximum charge acceptance curve of the specific battery. In the case of VRLA batteries, this closely matches an exponentially decaying function. Looking at figure 4.4.1, it can be seen that the current pulses roughly follow this curve. Not only does this allow charge in a minimal amount of time, but as can be seen by looking at figure 4.4.2, it extends the life of the battery considerably when compared to CC/CV charging.Fast Charging Algorithm for 12V 70Ah VRLA batteryFigure 4.4.1Comparing life of battery using fast charge vs. CC/CV Figure 4.4.2There are several benefits to fast charging, including decreased charge time and increased capacity retention throughout life, but application to the LSV2 charging system will be nearly impossible. Given the Ah rating of the propulsion batteries, the initial current values for this scheme would be upwards of 800A, requiring extremely large wire, and would overload the power system on base. This is also not a suitable charging method for many batteries in a confined space because the large amounts of current cause a large amount of heat. Further comparison of this and other schemes can be found in section 6.Catalyst CapsProvided by Philadelphia Scientific as elaborated upon in [4], these devices replace the native vent caps and introduce a catalyst into the system which aids in the recombination of oxygen and hydrogen. These devices cost $35.00; since the LSV2 batteries will require 2 per battery, the total cost will be $70 per battery. Because of the cost, these devices have been shown to prolong VRLA batteries life by up to 100%. Figure 4.5.1 shows how the Catalyst Cap decreases the loss of H2 molecules by absorbing them before they are released through the vent cap. The United States Air Force now requires catalyst caps to be installed in all new VRLA batteries installed in their communication systems. They are also working to replace old batteries with new ones containing catalysts caps [5]. A more in-depth explanation of catalyst caps and how the Air Force is using them can be found in [4] and [5].Equation Typical gas cycle of a 100Ah VRLA battery [5]Figure 4.5.1Concept SelectionTable 5.1 shows the decision matrix utilized to determine which schemes will be feasible. The solutions shown were down-selected from the initial research phase. In order to weight each item, they were rated out of a 100% pool from most important taking up 25%, down to 1% being the least important item to worry about when considering the different algorithms. We determined that the best course of action for this project was to save money. In order to save money, the system needs to be improved, meaning longer underways until the batteries are decommissioned, increasing the useful life of the battery system, and not requiring a major infrastructural upgrade to implement a new method of charging.The first option, CC/CV is the current configuration, and the fourth option is the fast charge solution which will not be considered due to poor scores. Because the other choices, ZDV, CI, and Catalyst Caps are close in scoring, and because with many unknowns within the matrix, all three have been chosen for lab testing.Item DescriptionMethod Weight??CC/CVZDV CI FCCaps ?Software Complexity Less lines of code for greater rating0.270.20.20.20.33%Power Requirements Lower kW rating for better rating0.450.50.30.10.66%Shore Power Considerations T = 0/F = 1 - Requires Infrastructure upgrade1.51.51.501.515%Rewiring of Barge and Vessel Less rewiring for better score0.20.20.200.22%Difficulty of ImplementationReplacement with on-hand or new units0.450.40.40.20.45%Capacity available for Underway The more charge available for runs the better1.21.61.821.620%Expected EOL Capacity More capacity at the end of life 0.30.60.910.810%External Interfacing of Controls T= 0/F = 1 - Requires external interfacing? ? ? ? ? 8%Reduction in Charge time Greater reductions for better score? ? ? ? ? 1%Cost of Implementation Higher rating for lower costs? ? ? ? ? 5%Long term Costs reduction Higher rating for greater reduction possibility? ? ? ? ? 25%?Higher score is better ?4.3755.23.45.4100%Table 5.1 Future WorkDue to the availability of critical components required to test the capacity retention over a battery’s lifetime, team AutoVolt has developed a future work plan with two paths. The University of Idaho Battery Lab has given the team access to one channel on an Arbin BT2000 battery cycler. It is a 300A testing machine capable of handling any solution proposed. Three choices will be tested for 100 cycles as a proof of concept over the months of February to April. Figure 6.1-7493001445895The other path involves building a cycler for higher cycle testing. The cycler will include the use of SEADAQ, a data acquisition system designed for the SEAJET to collect data during the cycling of the batteries. Because of the complexity of designing a charge and discharge testing apparatus, the construction of the system will require a significant amount of the spring semester. Depending upon the time constraints, it may be designed to enable all solutions to be tested, or only one system to be tested with the system setup.References[1] Matthew A. Keyser, et al., “Charging Algorithms for Increasing Lead Acid Battery Cycle Life for Electric Vehicles”[2] Elizabeth D. Sexton, Robert F. Nelson, and John B. Olson, “Improved Charge Algorithms for Valve Regulated Lead Acid Batteries”[3] V. Svoboda “The Influence of Fast Charging on the Performance of VRLA Batteries.” June 2004[4] “Catalysts Save US Taxpayers Millions” Batteries + Energy Storage Technology, Winter 2005 pg 67-70.[5] Teresa Hanson “Catalysts on VRLA Batteries Save Air Force Millions of Dollars” , Oct 2005 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download