1 - University of Idaho



Ultracapacitor-Based

Vehicle Electric Load Leveling System

Erin Davis

Fred Jessup

Benton O’Neil

November 30, 2004

1.0 Introduction 2

1.1. Executive Summary 2

1.2. Background 2

1.3. Problem 2

1.4. Objectives 2

1.5. Scope 2

2.0 Design Details 3

2.1. System Definition 3

2.1.1. Functional Requirements 3

2.1.2. Specifications 3

2.1.3. Constraints 3

2.2. System Design 3

2.2.1. Design Iterations 3

2.2.2. Design Focus Decision 5

2.2.3. Final Design 6

3.0 Methods 6

3.1. Testing 7

3.2. Modeling 8

3.3. Design 10

4.0 Discussion 11

4.1. Modeling 11

4.1.1. Accuracy 11

4.1.2. Results 11

4.2. Design 14

4.2.1. Results 14

4.2.2. Cost Analysis 15

5.0 Conclusions and Recommendation 15

5.1. Conclusions 15

5.2. Recommendations 15

5.2.1. Implementation on Ford Explorer 15

5.2.2. Future Work 16

1.0 Introduction

Executive Summary

The purpose of this report is to outline the preliminary research and design performed for the implementation of an Ultracapacitor-based electric load leveling system on a 2001 Ford Explorer. Our goal is to reduce vehicle dependence on the high-power capabilities of lead-acid batteries by supplying high starting currents with Ultracapacitors. This will allow for a reduction in battery size and a decrease in vehicle weight, leading to improved fuel efficiency. Details beyond those given in this report are available at seniordesign.uidaho.edu/loadlevelers.

The system described in this report utilizes the high-current capabilities of Ultracapacitors to start the vehicle’s engine, thus eliminating the need for high current lead-acid batteries and reducing the vehicle weight. The Ultracapacitors are recharged by a smaller and lighter battery and are then disconnected to be used during the next engine start. Our modeling and cost analysis suggest that the reduction in weight due to this system will improve fuel efficiency and reduce operating costs.

Background

The American Trucking Association (ATA) currently supports our research and development of vehicle efficiency improvements. The ATA requires a prototype to increase battery life and fuel efficiency on tractor trailers. Battery life for tractor trailers is significant due to the lost profits during vehicle downtime, and also due to costs associated with battery disposal. Vehicle weight is an additional concern for the ATA and any decrease in weight will allow for greater payloads and increased profits.

Problem

Today’s vehicles do not manage the flow of power through their electrical systems for maximum fuel economy and battery life. Large current spikes during engine starting and constantly varying electric loads shorten the life of the battery. This electrical system inefficiency results in heightened operating costs due to battery replacement and disposal. Lead acid batteries used in the majority of vehicles today can be harmful to the environment in both production and disposal[1]. Alternate high-power electric sources exist that can reduce vehicle dependency on lead acid batteries. The vehicle’s alternator both charges the lead-acid battery and supplies the vehicle electric loads at the cost of creating a mechanical load on the engine; however, the alternator functions independent of the power required of the engine. An algorithm could be devised to control when the alternator generates power to increase engine efficiency.

Objectives

The cost of producing this system is offset by its monetary and environmental benefits which allows for the calculation of a payback period. Our ultimate goal at the completion of this project is to achieve a payback period of three years when applied to a 2001 Ford Explorer with additional operating cost savings beyond the three years. We aim to extend battery life to at least that of the manufacturer’s specification. Increased fuel efficiency is an additional benefit of the design and is due directly to the reduction in vehicle weight. The combination of these two improvements will allow us to achieve the required payback period.

Scope

The project deliverables are as follows: design description report, physical implementation of system on 2001 Ford Explorer, technical report, and trade show presentation. At the completion of the project we will have demonstrated the feasibility of an ultracapacitor system that normalizes battery current and reduces battery size and weight.

Design Details

System Definition

1. Functional Requirements

Functional requirements are the things that the system must do. They are the wishes of the customer and are not directly measurable. The system must be able to start the engine under all normal operating conditions or in the same capacity as the stock battery. It must minimize the weight of the starting system, reduce the required battery size, and improve the battery life.

2. Specifications

Specifications are the measurable parameters that define the system. The system must be capable of supplying a peak current of 650 amps with a continuous current of 150 amps for five seconds. The complete system must weigh less than twenty pounds installed.

3. Constraints

The constraints limit the possible design solutions. These constraints come from the customer, the design platform, the federal laws, and the available technologies. The system must meet Federal Motor Vehicle Safety Standards. To do this, it must fit under the hood of the vehicle, all wires must remain a safe distance from moving parts, and all electrical connections must be covered. There must be no driver involvement necessary for normal operation of the system. The total cost for design and prototype construction must not exceed $4000. After one year following implementation the benefits of the system must have exceeded the costs.

System Design

The scope of this project was initially much broader than it has recently become. Initial project goals included controlling current flow into and out of the battery, starting the engine normally, regulating alternator output, and reducing battery power requirements. Through multiple design iterations, journal and article research, and discussions with professionals at the University of Idaho, our scope has narrowed significantly. Our decision to limit the breadth of our design is due to both time limitations (the project must be completed and tested by the end of April) and technological feasibility. The following illustrates the process through which our final conceptual design was reached.

1. Design Iterations

After consideration of the initial goals of the project we decided that there were three possible directions to go with our system design. The first option was to address only the battery to decrease size and increase life. The second option was to address the inefficiencies of the alternator to improve fuel efficiency. The third possibility was to attempt to address both problems in one design.

1. Design #1 – Addressing the Battery

As described previously, the battery supplies very large currents during engine starting. This requires a battery with high-power capabilities found most often in automotive lead-acid technology. With an average life cycle of three to four years [2], these batteries are large, heavy, expensive, and harmful to the environment in both production and disposal. A decrease in dependence on lead-acid technology or an increase in the life of the batteries would therefore be beneficial. The following design was proposed to reduce the size of the battery by starting the engine with Ultracapacitors thus reducing the high power requirement of the battery.

[pic]

Figure 1.1 – Design #1

The DC/DC converter found in this design is necessary to control the current into and out of the battery. The purpose of controlling the battery current was to maintain an optimal charge and discharge rate and in doing so, increase the life of the battery. This DC/DC converter would have to be bi-directional to allow energy to flow both into and out of the battery.

2. Design #2 – Addressing the Alternator

The second design option was to improve the function of the alternator to directly improve the efficiency of the engine. The alternator is connected by a belt directly to the engine. It converts the mechanical force of the engine into an electrical voltage and current that supplies electric loads in the vehicle. When the engine is working hard, in climbing a steep grade for example, the power delivered to the wheels is high. If the alternator were to present a mechanical load to the engine during this time, it would decrease the power delivered to the wheels. This would cause the engine to require more fuel to maintain the current speed and the engine efficiency would decrease. The second design proposed implements a switching algorithm for the alternator that would only allow it to generate electric power when the required engine output power is at a minimum. At times when the engine is running inefficiently, the alternator would shut down and Ultracapacitors and the battery would supply all electric loads. Figure 2.2 below shows this design setup. The DC/DC converter found in this design is required to increase the capacitor voltage. By increasing the capacitor voltage the energy that can be stored in the capacitor bank increases proportionally to the voltage squared. This DC/DC converter would also have to be bi-directional to allow energy to flow into and out of the capacitor bank.

[pic]

Figure 2.2 – Design #2

3. Design #3 – Addressing Battery and Alternator

The third design option was to address both battery and alternator in our design. This would require a more complex control system and more components to complete, but with increased benefits.

[pic]

Figure 2.3 – Design #3

2. Design Focus Decision

Complexity and feasibility of the designs as well as possible benefits were considered in deciding which direction to go with the design. After a fair amount of research, it was concluded that an optimal charging method for the battery would be difficult to determine with absolute certainty. Therefore, the goal of optimally charging and discharging the battery was scrapped and improving battery life would be limited to controlling the battery depth of discharge and limiting the discharge current of the battery to a maximum of fifteen amps. In addition it was decided that the control of the alternator was not technical feasible in that this would require the dismantling and interruption of internal stock vehicle systems to determine real-time engine load values. These values would be necessary to effectively improve engine efficiency. The following decision matrix indicates that Design #1 is the most appropriate for the direction of the project to progress and is a result of team, advisor, mentor, and customer discussion.

|  |Weighting Factor |Design #1 |Design #2 |Design #3 | |

|Benefit to |0.05 |3 |1 |2 |Design #3 assists alternator during peak loads |

|Alternator | | | | | |

|Time to |0.20 |1 |2 |3 |Design #3 is most complex and difficult to complete, |

|Complete | | | | |Design #2 would require a more complex control |

| | | | | |algorithm |

|Cost |0.20 |1 |2 |3 |Two converters in Design #3 would cost most, smaller |

| | | | | |cap bank and converter in Design #1 would cost less |

| | | | | |than in #2 |

|Weight |0.20 |1 |2 |3 |Weight of two converters and two cap banks in Design |

| | | | | |#3, Design #2 still requires full-size battery |

|Size |0.10 |1 |2 |3 |Same as weight issues |

|Efficiency |0.10 |1 |2 |3 |Two converters in Design #3 would use most power, |

| | | | | |high-power converter in Design #2 would be less |

| | | | | |efficient |

|Total |1.00 |1.10 |2.10 |

|BCAP0350 |$216.00 |1.19lb |13.125s |

|BCAP0013 |$175.00 |3.351lb |14.464s |

|BCAP0008 |$444.00 |6.173lb |16.5s |

|BCAP0010 |$606.00 |8.102lb |17.333s |

|PC2500 |$300.00 |11.188lb |16.667s |

Figure 3.7: Capacitor Cost Comparison

Aspects of the design that remain to be addressed include the DC/DC converter, the control system, the battery, and the installation.

In order to address the DC/DC converter we will first determine the necessary input and output voltages and currents. This will dictate the power rating for the converter. Using this power rating, an attempt will be made to purchase an off-the-shelf buck/boost converter suitable for this application. If one cannot be found, the converter will be built in-house using the same specification.

The control system will be designed by first listing all necessary inputs such as battery voltage, battery current, and capacitor voltage. Outputs such as pulse-width modulation and voltage and currents will also be listed. Next, a decision matrix will be used to decide between implementation using digital versus analog control. This decision matrix will be based on the viability of each necessary input and output in that respective implementation. Following this decision, either a digital microcontroller will be selected and programmed, or the analog control will be designed and built.

The next design decision is the selection of a battery. This decision will be based on the required power to charge the capacitor bank for a specified depth of discharge. The allowable depth of discharge will be selected such that the alternator will supply less than maximum current. This maximum current will be determined from the battery data sheet.

The final design decision is the method of system installation. A combination of Federal Motor Vehicle Safety Standards[4] (FMVSS) and DFMEA will be used to address issues such as safety and reliability, engine heat signature, capacitor packaging, wiring and connectors, EMI shielding, mechanical vibration, and the overall harsh environment on the engine compartment. Due to the extreme length of the FMVSS, it will not be followed specifically but will be met by emulating energy system installation already present in today’s vehicles.

Discussion

Modeling

1. Accuracy

The models detailed in this report and those used for design are not claimed to be 100% accurate. There are a number of parameters used in the models that have been approximated. These parameters are primarily electrical and include the starter motor armature resistance and inductance, the induced EMF coefficient Kf, and the electrical torque coefficient Kt. Using these approximate values, the models show results similar to those obtained in testing. These approximations are also justified by the fact that they depend upon a number of variables such as temperature, age, manufacturer, etc. Ultimately, the models are useful regardless of precise accuracy in that they produce results similar to those obtained in testing. The following are a few examples of the modeling results obtained from Simulink and a discussion of the lessons learned from each output.

2. Results

Figure 4.1 is the output of a Simulink model it which a capacitor bank is starting the Ford Explorer engine with an applied voltage of 15 volts. At time t=0s, the capacitors are switched into the circuit and they produce a current spike of 450 amps. This causes the engine speed to ramp up to a steady state value of 200 RPM. As the motor reaches a steady state speed, the capacitor current reaches a steady state level of 80 amps. This is the approximate expected behavior of the design upon implementation.

[pic]

Figure 4.1: Capacitor Applied Voltage of 15V

The following is an output similar to that above, but in this case the applied capacitor voltage has been reduced to eight volts rather than fifteen volts. It can be seen from Figure 4.2 that the peak current is much less than that in Figure 4.1, but it can also be seen that the engine speed never passes 90 RPM. From testing, we found that the engine must reach around 200 RPM in order to start. If the voltage on the capacitors is too low, as in this model, the engine will not start.

[pic]

Figure 4.2: Capacitor Applied Voltage of 8V

The next three modeling outputs illustrate the behavior of the three different converter topologies being used to charge the capacitor bank from the battery. The first output is the source current from the battery and the voltage on the capacitor bank during charging using a buck converter. The buck converter takes the battery voltage, in this case 12.8 volts, and bucks it down to lower voltage depending on the set duty cycle. In this output the duty cycle is set to 80%. The initial capacitor voltage is zero and the switching period for the converter is 0.1 seconds. At this point the three converters are being modeled with a constant duty cycle, but in the design a constantly varying duty cycle will be used to limit the source current.

[pic]

Figure 4.3: Buck, D = 80, Ts = 0.1, Vci = 0, Vbat = 12.8

The source current in Figure 4.3 reaches approximately 300 amps. This is much higher than desired for our design. The reason for the high source current is the constant duty cycle value. If the duty cycle were controlled, then the source current could also be controlled. In the output of Figure 4.3 the source current has a large ripple. This is due to the low switching frequency that used for the model.

The second output is the DC/DC boost converter. This converter has the battery voltage as an input and outputs a higher voltage according to duty cycle. In the case of Figure 4.4 the duty cycle is set to a constant value of 25%. It can be seen from Figure 4.4 that the ripple present in the source current of a boost converter is much less than that in the buck converter. This has important implications to battery life in that highly varying currents can be damaging.

[pic]

Figure 4.4: Boost D = 25, Vci = 0, Vbat = 12.8 Ts = 0.1

The final model output is that of a buck/boost converter, shown below in Figure 4.5. This converter takes the battery voltage as an input and outputs either a lower or higher voltage to the capacitor bank depending on the set duty cycle. This configuration has a source current ripple similar to that of the buck converter, but the advantage to this converter is that it can both boost the capacitor voltage and charge the capacitors when they are completely drained. The ripple shown can be improved by increasing the switching frequency to the point where the source current becomes smooth.

Results from the models are used to observe the behavior of different design alternatives and system configurations without the time-consuming task of building and testing. They allow for quick design validation and are useful in illustrating variables in the design, such as DC/DC converters. The models will also make the transfer from the test vehicle to any final application platform less difficult. They can be changed to model various vehicles with just a few parameter alterations. This method will be used in determining the feasibility of the design for a tractor trailer at the completion of the project.

[pic]

Figure 4.5: Buck Boost D = 55, Vci = 0, Vbat = 12.8 Ts = 0.1

Design

1. Results

There are foreseen advantages in using the chosen design over the stock starting system in the Ford Explorer. Although they will not be fully realized until after system implementation, a prediction of improvements is described below.

The first advantage of this design is decreased weight due to reduced battery size. It is no longer necessary that the battery supply the peak current to start the engine, which allows for the use of a smaller battery or a different type of battery. For example, a typical lead-acid automotive battery similar to that installed in the Ford Explorer, can weigh up to sixty pounds depending on the vehicle. This design uses a Sealed Lead Acid (SLA) battery weighing less than fifteen pounds. The capacitor bank, as it is currently designed, will weigh less than four pounds. The total system weight will give approximately forty pounds of weight savings.

A decrease in vehicle weight will increase the fuel efficiency of the vehicle [5]. For every 1% decrease in vehicle weight, there is a 0.66% decrease in fuel consumption. This means that for a forty pound weight reduction on a 4400 pound Ford Explorer, the decrease in fuel consumption is 0.6%. This result would be much more significant for a tractor trailer because there are four lead-acid batteries, each weighing 60 pounds. According to Dr. Donald Blackketter, director of NIATT at the University of Idaho, a one pound reduction in weight on a semi truck can be sold for between four and eight dollars. This would mean that the projected weight decrease of 200 pounds would be worth between 800 and 1600 dollars over the life of the vehicle [6].

The second advantage to this design is a reduced dependence on lead acid battery technology. One of the main purposes for the use of lead-acid batteries in automotive applications is their capability for high-power output. This design eliminates the need for a high-power battery and therefore the tie to lead-acid technology. Although this design does replace the stock battery with another smaller lead-acid battery, the design creates an opportunity to move to a different battery technology altogether.

2. Cost Analysis

There are two possible ways to perform a cost analysis of this system. The first method is to compare the cost of producing the system with the benefit of implementing it on the test vehicle. This is the method of analysis that has been used thus far in the design. The production cost of the system does not consider the cost as though it were being produced in quantity, but rather the cost of producing a single prototype. This allows for the creation of a budget estimate for the continuation of the project. The preliminary cost analysis indicates that the cost of this system will be approximately $300. This approximation does not consider any tool cost or installation costs, but only component costs. Appendix C contains the complete preliminary cost analysis of the design.

The second method of cost analysis will be completed at the end of the project and will consider the cost of producing the system in mass quantity. This analysis will not consider any research or development costs and will be calculated as applied to the final platform, a tractor trailer. The payback period is the amount of time required for the costs and benefits of a system to balance out. It has been made clear by the customer that in order for this design to be feasible for large scale use in tractor trailers, its payback period must be equal to or less than one year. For this second cost analysis, the benefits of the system will be considered and compared to the costs in order to determine the payback period.

Conclusions and Recommendation

Conclusions

The ATA presented our team with the challenge of reducing the operating cost of a tractor trailer. They require a prototype that would demonstrate the feasibility of the system and an estimation of the benefits if it were implemented on a semi-truck. Through vehicle testing, the team was able to identify several key issues that could be addressed to improve efficiency. After consideration of the main issues, a single direction was for the design. Further testing and extensive modeling was performed in order to validate the chosen design decisions. An analysis of the chosen design was performed to determine its costs and benefits. This analysis has shown that construction of the system is feasible and that there are benefits to be obtained from vehicle implementation.

Recommendations

1. Implementation on Ford Explorer

It is the recommendation of the team that the design described in Section 2.2.3 be constructed and implemented on the 2001 Ford Explorer. This will provide proof of the concept and will allow for further testing to determine the feasibility for implementation on a tractor-trailer.

A testing plan is necessary in order to determine whether the system is effective or not after completion. As discussed in Section 2.1.1, the system must be able to start the engine under all normal operating conditions or in the same capacity as a stock battery. It must also minimize the weight of the starting system, and reduce battery size. It will be difficult to determine whether the capacitors are able to start the engine under all normal operating conditions, so a decision will need to be made as to what conditions, specifically, the capacitors should be able to start under. Testing will then be performed to see if engine can be started under the selected conditions. Reduced starting system weight and decreased battery size will be a simple measurement. A weight measurement will be taken of all components removed from and installed to the vehicle. The difference will between these weights will indicate the weight savings achieved by the system.

The final design presented in this report claims that it will reduce the peak currents seen by the battery. After implementation, current will be measured from the battery during starting to ensure that this is true. This design also suggests that the alternator will provide current for the majority of electric loads during the driving cycle due to the decreased power capability of the battery. Driving cycle tests will be performed and the current from the battery and alternator are measured. This will show the ability of the alternator to supply the majority of power while the engine is running.

2. Future Work

The scope of this project was initially much broader than what it is now. There were many different paths we could have followed during the course of the conceptual design. Though the team chose to work on weight reduction and starting the engine, two earlier topics discussed may be worth further research to improve vehicle efficiency. The first area of interest is the output regulation of the alternator. As discussed in Section 2.2.1.2, the alternator could be used more efficiently during the driving cycle to decrease the fuel consumption. The second area of interest is in the optimization of the battery charging algorithm. Determining what method is most effective for charging a battery is difficult, but the benefit in doing so would be an extended battery life. Preliminary testing showed that the alternator did not charge the battery using a recognizable algorithm, but rather supplied as much current as possible to the battery. There is definite room for improvement in this charging method. These two areas of further work could be continued in an additional senior design project or possibly at the completion of this project, if time allows.

References Cited

1] "Lead Acid Batteries – Hazards and Responsible Use" Integrated Waste Management Board,

Publication #612-00-002. Mar 2000.

[Online].

2] “United Nations Environment Program” EnTA Workshop, Manila Feb. 2000. Page 150-151

3] “CVSP Fuel File: Engine Data from Engine Mapping” Mike Herr. Emailed Document. Mapping Vehicle Database Reference #: CVS 483.

4] “Federal Motor Vehicle Safety Standards and Regulations” U.S. Department of Transportation. DOT publication. Dec 1998. [Online]

5] “EEA’s Methodology to Calculate Fuel Economy Benefits of the Use of Multiple Technologies” US Government Printing Office, 1991 0 – 297-903 QL:3. [Online]

6] “Tractor Trailer Weight Savings” Dr. Donald Blackketter – Int. Director NIATT University of Idaho, Email Document. November 2, 2004

Appendix A – Driving Cycle Event Log

Appendix B – Modeling Parameter Calculation

Appendix C – Preliminary Cost Analysis

................
................

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

Google Online Preview   Download