A Flywheel-Based Voltage Sag Correction System
for the University of Idaho Analog Model Power System
December 12, 2005
Department of Electrical and Computer Engineering
University of Idaho
Sponsors and Mentors
Dr. Herb Hess
Dr. Brian Johnson
Instructor
Dr. Brian Johnson
Flywheel Team
Gavin Abo
Nate Stout
Nathan Thomas
Table of Contents
List of Figures and Tables...... ii
Abstract...... 1
Index Terms...... 1
Introduction...... 1
Background Information...... 1
Project Description...... 1
Problem Statement...... 2
Objectives...... 2
Constraints...... 2
Functional Specifications...... 3
Solution Method...... 3
Status...... 4
Method of Solution...... 4
Technical Description...... 4
Theoretical Basis...... 5
Test Plan...... 8
Appendix A: Figures...... A-1
Appendix B: Pictures...... B-1
Appendix C: Specifications...... C-1
Appendix D: Bill of Materials...... D-1
Appendix E: Parts Ordered...... E-1
Appendix F: Individual Reports...... F-1
List of Figures and Tables
Figure 1. Signal Flow for a Detected Sag...... 3
Figure 2. Space Vector PWM Vector Diagram...... 7
Figure 3. PWM Variable Width Example...... 7
Figure 4. Overall Design Schematic...... A-1
Figure 5. AMPS Side Switch Diagram...... A-2
Figure 6. Induction Motor Side Switch Diagram...... A-2
Table 1: Specifications...... C-1
Table 2: Bill of Materials...... D-1
Table 3: Known Converter Data...... E-1
1
A Flywheel-Based Voltage Sag Correction System for the University of Idaho Analog Model Power System (December 2005)
Gavin Abo, Nathan Stout, and Nathan Thomas
Abstract—A flywheel-based voltage sag correction system for the University of Idaho analog model power system which responds within 2 cycles correcting a 37 percent maximum sag. The system will correct the voltage to 0.95 per unit,0.05 per unit of the rated voltage for a maximum duration of 1.5 seconds for symmetrical sags only. The background, design, implementation, product lifecycle, reliability report, test plan, specifications, budget, user manual, and future work of the flywheel voltage sag correction system are documented in this report.
Index Terms–Flywheels, power systems, voltage sag correction.
(do we need to include words defined in the appendix here?)
A
- INTRODUCTION
flywheel-based voltage sag correction system (FVSCS) for the University of Idaho analog model power system was developed in a two semester senior design class, ECE480/481. Which is a requirement for graduation with a Bachelor of Science in Electrical Engineering degree for the members of this project.
- Background Information
In the mid-1990s, the University of Idaho acquired the analog model power system (AMPS) from Idaho Power for educational and research use [1]. Idaho Power is an electric utility provider to about 895,000 customers in southern Idaho and eastern Oregon [2]. The AMPS was originally constructed by Idaho Power to test relays and relay settings. In addition, the AMPS was used to model part of Idaho Power’s own transmission and distribution system. Over the years, the University of Idaho has made several modifications to the donated system to incorporate the following:
1)A fault matrix that allows three faults to be placed on the system either simultaneously or in an evolving manner.
2)The ability to create faults by loading impedance onto the system.
3)SEL (Schweitzer Engineering Laboratories) protection relays.
The AMPS is currently located in the basement (room G10) of the Buchanan Engineering Laboratory (BEL) on the University of Idaho campus in Moscow, Idaho. It has been and still is a valuable tool for students and researchers as it provides insight into the operation of a power transmission system. Additional subsystems are continually being added to the system such as this project and the supplemental power fuel cell system for AMPS designed during the same time period by the fuel cell team as an overall project entitled, “HydroFly Project.”
A flywheel is well known for its efficient mechanical energy storage in its rotating momentum. This can then be applied to a generation source that converts mechanical energy into an electrical energy output to a system. Thus, a flywheel is a practical alternative power source for temporary inline (series) voltage sag correction for the AMPS.
Satish Samineni, a past graduate student at the University of Idaho, modeled and simulated a flywheel energy storage system for voltage sag correction with PSCAD/EMTDC and MATLAB computer software for his Master of Science degree. His thesis showed that this project was feasible and it was suggested as future work [3]. However, some alternations to his design were made to build and implement the system in hardware since his simulation design was for a shipboard power system rather than the AMPS [4].
- Problem Statement
The AMPS does not have the ability to correct for voltage sags.
- Objectives
Interface a flywheel-based voltage sag correction system to the AMPS. Furthermore, to provide documentation for observation and analysis as an educational and research tool for this system.
- Scope
The FVSCS will only compensate for balanced sags (equal voltages on all three phases), but the design allows for the addition of unbalanced sag correction as a future implementation.
II. Functional Description and Specifications
The theory, design, and implementation of the FVSCS is described is this section.
- Theoretical Basis
In general, a voltage sag is a short-term drop of voltage from its nominal voltage. A voltage sag of only 10% (10% of the nominal voltage) can cause sensitive loads to misoperate or shutdown completely [3]. For example, if a system that operates at a nominal voltage of 120 V (RMS) sees a voltage sag of 10%, it will have voltage sag of 12 V (RMS) or have a voltage of 108 V (RMS). A voltage sag interrupting equipment and services of a process and fabrication plant can result in significant loses of time and money. The most common causes of voltage sags are the energization of loads that draw large starting currents (i.e. a motor) or electrical faults [3].
The energy from a flywheel can be used to correct for voltage sags on a system so the critical loads never see these sags. A flywheel stores an amount of energy proportional to the moment of inertia, I, and the rotational speed, ω, squared.
(1)
The moment of inertia of the flywheel (a solid cylinder) is
,(2)
where m is the mass and r is the radius of the flywheel. The mass of the flywheel (or an object) can be found if the volume, V, and the material density,, of the flywheel are known.
(3)
The volume of the flywheel (a cylinder) can be determined from its geometry and is given as
,(4)
where r is the radius of the flywheel and t is the thickness of the flywheel.
The theoretical estimate of the maximum energy storage of the flywheel for the FVSCS is 91.31 kJ. The parameters used to calculate the maximum energy storage of the flywheel are given in Table I. The maximum energy storage was calculated from using equations (1), (2), (3), and (4) in MathCad [5]. The maximum energy storage was calculated neglecting the axles of the disc (shaft and couplers) and induction motor (shaft and rotor). Intermediate results gave that the flywheel mass is about 257 lbs and moment of inertia is 5.406 kg-m2.
II. Product Lifecycle for a FVSCS
- Product Lifecycle Introduction
The product lifecycle of the FVSCS begins with the Research and Design Phase and is followed by the Development Phase, Introduction Phase, Growth Phase, Maturity Phase, Decline Phase, and Death Phase. The life of the FVSCS in each phase is described in detail in the next section.
B.Product Lifecycle Phases
1)Research and Design Phase
The Research and Design Phase was performed by Satish Samineni. In this phase, extensive background research, modeling, and simulation were performed.
2)Development Phase
Our senior design team undertook the Development Phase of the FVSCS. The project, commissioned by our sponsors, began with basic background research that included reading articles and papers. Detailed specifications were then finalized with our sponsors with input from this research. Design decisions were then made to keep the FVSCS within our budget. The costs of this phase consist of those for a single FVSCS. We found, reviewed, and ordered components and parts for the FVSCS according to our budget and specifications. A demonstration of concept for the SVPWM was created, supporting documents were reviewed for system integration, and additional support came from sponsors/advisors throughout the development process. The FVSCS will be tested according to the materialized test plan of this phase. Updates to design from tester?-Brian
The flywheel interface will require many functions and algorithms to be successful. This will require that a top down modularly designed code be generated, within specifications, to perform these tasks. The system only accounts for balanced sags at this phase. Consequently, it should be made upgradeable to handle unbalanced sag correction for future modifications.
The only hardware developed in this phase is the charging power supply for the capacitors of the DC link. The rest of the hardware combined to form the FVSCS was either provided by the sponsors, donated by different companies, or purchased from various companies.
3) Introduction Phase
At the first introduction of the FVSCS, the primary consumer will be the University of Idaho’s Department of Electrical and Computer Engineering. There are still no limits on the size or weight at this phase of the lifecycle. Also, the software will have limited operating conditions or portability to different DSPs.
The initial customers will not need a lot of support since they are engineers and understand the fundamental engineering concepts. They should need only to read the users manual and make the proper connections to interface the FVSCS, with the power system of interest, as long as it meets the specifications of the FVSCS.
Some limitations on the system in this stage would be:
- Only balanced sag correction
- Limited sag correction duration due to flywheel limitations
- No error output signal indication or display
4) Growth Phase
In the growth phase, the system should be marketed to anyone with a need for an FVSCS system. This would include any customer that requires a consistent voltage magnitude to its loads, in which case the product will have to be modified to meet a broader spectrum of customer needs. It may require some hardware redesign to minimize volume and weight, and the software might also need some modifications depending on the required inputs and outputs to the system. The DSP can easily be updated by reprogramming (flashing) it as long as DSP hardware changes are not required. For example, a European company might need the product, in which case the FVSCS would need modification for a differentoperating frequency.
In this phase, the system would be expanded beyond the limitations established in the Introduction Phase. The biggest improvement would be the ability to correct for closed loop control on unbalanced sags and some kind of output system to display errors. A data acquisition system could also be added for customers that did not already have one. In order to maximize profit in a mass production situation, as many units should be sold and distributed to anyone who has a power system to interface the FVSCS to.
5) Maturity Phase
In the maturity phase, the customer base is not likely to change. The product should not need any major improvements, so there will be less need for product support. Profits should maintain at a fairly stable level, and have little decline.
**added**The customer should perform regular checks on the system such as checking the bearings, induction motor, and connecting shafts to determine wear and/or breakdown. After these checks, the customer should perform any maintenance that is required to repair, reduce, and/or prevent further damage to the system.
If the DSP is discontinued there should be no problem unless for some reason the DSP fails. In the event of a failure and the DSP being discontinued, there is always the chance that some distributors will still have a few in their inventories that can be purchased and replace the failed DSP. If an identical replacement can not be found, another DSP will be lookd into at that time.**added**Required maintenance for users? What if DSP discontinued?-Brian
6) Decline Phase
In the Decline Phase, newer and more efficient systems will most likely be in development and preliminary release. The customer base will be taking an interest in these newer products and moving away from the use of this FVSCS product. Profits from the FVSCS will be declining due to the decline in users and interest.
7) Death Phase
In the Death Phase, there will most likely be various newer, cheaper, and more efficient systems that perform the same job as the FVSCS at a better price. The current customer base will have moved to these newer systems, and the FVSCS will not be a viable product for them. However, the FVSCS may be kept as a historical distribution system, possibly salvaged for other design projects, disposed of, reused for parts, or recycled.
III. Reliability Report for a FVSCS
A. Reliability Report Introduction
The next section is an analysis of the reliability of the FVSCS. It will also explore how the potential failure modes of each component relate to the overall reliability of the FVSCS. Talk more about Table 2. Failure mode is for entire system. Some systems don’t have subsystems, etc.
B. Reliability Analysis
1)Failure Rate Calculation Methods
Failure rate values were found on the Internet for components similar to ours and these values were used for our component failure rates [1]. The mean time between failures (MTBF) was calculated by taking 106 hours divided by the number of failures of the component in 106 hours.
2)Components and Their Failure Modes
The major systems are the building blocks of the FVSCS. They depend on the subcomponents in order to function. Some systems fall into their own category and do not contain sub components with a significant failure rate.
The failure modes listed in Table II as Partial/Total depend on how severe the failure is. For the IGBT’s, one to five of them can fail and the converter can still provide partial correction. For the LEMs, one of them could fail and leave the converter only measuring two of the three phases. The voltage LEMs require a resistor to change the voltage output to a limited current range. If the resistor failed, the LEM would not be measuring within its specifications. A total failure is defined as not correcting on any phase.
3)Overall System Reliability
The overall system reliability as shown in Table III is a function of the reliabilities of the components that make up the entire system. Since a failure of any individual system component will result in the failure of the entire FVSCS, the overall system reliability is the sum of the reliabilities of all of the system components.
List number of each subcomponent. Show how to get these - just sum? Show a fault tree. -Brian
4)Detection Likelihood
The partial failures would not be detected by the code, but the user could detect them by looking at the output of the AMPS. If the voltage graphs do not show sag correction equally on all three phases, then one of the components has failed. At this time there is no plan to make a system for detecting which component has failed.
A total failure could be observed by noticing a lack of sag correction on the output. This could occur by many different paths, such as the flywheel could stop spinning, the LEMs could fail, or the DC link could fail to charge initially.
Table with RPNs shown – severity, occurrence, detectability -Brian
IV. Test Plan for a FVSCS
We might want to just briefly the subcomponent procedures under test in the users test plan and cite a reference to it here [6]. Then, write up and include the short sponsor specification test table here.
V. Budget for a FVSCS
**added**The budget allowed and used for the FVSCS project showing donations, expenditures, and totals is provided in Appendix __. See appendix [7].**added**
_ is the color that Nate is using to denote items he feels should be deleted/changed.
_ is the color that Nathan is using to denote items he feels should be deleted/changed.
_ is the color that Gavin is using to denote items he feels should be deleted/changed.
Appendicies
Table of Appendicies:
Appendix A: Figures...... A-1
Appendix B: Pictures...... B-1
Appendix C: Specifications...... C-1
Appendix D: Bill of Materials...... D-1
Appendix E: Parts Ordered...... E-1
Appendix F: Individual Reports...... F-1
Manuscript received December 9, 2005. This work was supported in part by the University of Idaho under Dr. Herb Hess and Dr. Brian Johnson.
Gavin Abo is with the Electrical Engineering Department, University of Idaho, Moscow, ID 83844 USA (e-mail: ).
Nate Stout was with the Electrical Engineering Department, University of Idaho, Moscow, ID 83844 USA (e-mail: ).
Nathan Thomas is with the Electrical Engineering Department, University of Idaho, Moscow, ID 83844 USA (e-mail: ).