PSERC - PowerPoint PPT Presentation

1 / 19
About This Presentation
Title:

PSERC

Description:

Design of a high speed, MEMS, circuit breaker for the U.S. Navy for on-ship ... Stiction. Delamination. Factors Contributing to Switch Failure. PSERC ... – PowerPoint PPT presentation

Number of Views:166
Avg rating:3.0/5.0
Slides: 20
Provided by: dja27
Category:
Tags: pserc | stiction

less

Transcript and Presenter's Notes

Title: PSERC


1
Simulation for Failure Analysis of a Series and
Parallel Switched Power Circuit Breaker
D. S. James Student Member, IEEE,
Arizona State University Arizona State
University
G. T. Heydt Fellow, IEEE,
October 21, 2003
PSERC
2
Research Motivation
Research Motivation
  • Design of a high speed, MEMS, circuit breaker for
    the U.S. Navy for on-ship fault and overloading
    protection
  • General design specifications include
  • High voltage, DC switching
  • Distribution class industrial ratings
  • Innovative design of microelectromechanical
    systems
  • (MEMS) in high power applications
  • Project funded by NSF and the U.S. DoD under the
    motivation for Electric Power Network Efficiency
    and Security (EPNES)

3
Introduction to Series and Parallel Systems
The concept of series and parallel devices has
been used in the past to increase overall system
reliability and efficiency. Solid state and
MEMS switches of lower ampacity and voltage
ratings may thus be used to build-up a high power
circuit breaker.
Series and parallel switch layout
4
Transient Snubbing
In DC systems, zero voltage switching is
unachievable and stresses caused by switching
transients can destroy the switchgear. Parallel
snubber circuits are often used to dampen the
electrical stresses. An example of an
over-voltage snubber is a RCD (resistor-capacitor-
diode) circuit.
DC over-voltage snubber
5
Design of MEMS Switch Matrix
SIDE VIEW
TOP VIEW
MEMS switch design (side view)
MEMS switch design (top view)
6
Design of MEMS Switch Matrix
60µm
1328 µm
100µm
CLOSER VIEW OF A SINGLE SWITCH (METALIZATION)
1175 ?m
4 x 4 Matrix of switches using aluminum
metalization
7
Factors Contributing to Switch Failure
Modeling failure probabilities for switches
requires broad knowledge of the devices used and
is often found through extensive experimental
data. Select contributing factors to failure of
MEMS include
  • Switching speed
  • Over-voltage
  • Over-current
  • Switching transients
  • Creep
  • Static overload
  • Moisture
  • Mechanical failure of
  • cantilever
  • Thermal stresses
  • Extensive time in service
  • Stiction
  • Delamination

8
Reliability Estimation Methods
Two proposed methods can be used to estimate the
reliability of a series and and parallel
configuration of switches
  • Markov process

2) Monte Carlo simulation
9
Markov Process Notation
  • The failure of multiple switches can be modeled
    by a Markov process.
  • The single switch state transition diagram
    depicts only four possible states of a switch
  • A good switch in the open state
  • A good switch in the closed state
  • A failed switch in the open state
  • A failed switch in the closed state

State transition diagram of a single switch
10
Markov Process Notation
As the number of switches in the system
increases, the total number of transitions
rapidly increase
State transition diagram for 2 series switches
11
Monte Carlo Simulation
A conjoined iterative and Markov approach can be
used in estimating the optimal series and
parallel configuration of switches given any
magnitude of system ratings.
12
Monte Carlo Simulation Assumptions
The key assumptions of this simulation are
  • Synchronized switching
  • Perfect voltage and current snubbing
  • Perfect voltage and current grading
  • Identical components
  • No bouncing of the switch cantilever
  • Time is not considered, only number of
    switching
  • sequences

13
Modeling the Probability of Switch Failure
A cumulative distribution function (cdf) can be
used to model the probability of switch failure,
Q(x), versus a failure-causing parameter, x. A
general model of Q(x) is used to convert
experimental results to the simulation
The probability of failure function, Q(x), versus
switch voltage, current, or speed
14
Circuit Breaker Failure Identification
Failed system in the open state
Failed system in the closed state
15
Mean Time to Failure (MTTF) Estimation
By integrating the cdf of the probability of
failure and a Markov model for a single switch,
the mean time to failure (MTTF) of the circuit
breaker can be estimated for any series and
parallel configuration.
(MTTF)
Series switches
Parallel strings
MTTF for all configurations of a 100 switch
series and parallel circuit breaker
16
Conclusions
  • A complete Markov process is too complex in
  • analyzing large series and parallel systems
    systems
  • A Monte Carlo iterative simulation can be
    integrated
  • with a basic Markov model of a single switch to
  • estimate the system reliability
  • Statistical methods can be used in determining
    the
  • optimal series and parallel configuration of
    any type
  • switch matrix

17
Present and Future Work
  • Integration of transients into failure analysis
  • Asynchronous switching analysis
  • Reliability of MEMS breaker under fault
    conditions
  • Modeling and testing of switch reliability
    curves
  • AC switching

18
Questions
?
19
Switch Failure Mode Identification
The loci represent constant power curves.
Assuming no switching transient is present, the
operating regions lie along the axes. An
isolation between failure modes is assumed by
treating the voltage and current stresses as
separate and independent events
Write a Comment
User Comments (0)
About PowerShow.com