MONITORING SYSTEMLEVEL AND DEVICELEVEL ANOMALIES IN STEAM GENERATORS AND HEAT EXCHANGERS - PowerPoint PPT Presentation

1 / 43
About This Presentation
Title:

MONITORING SYSTEMLEVEL AND DEVICELEVEL ANOMALIES IN STEAM GENERATORS AND HEAT EXCHANGERS

Description:

Issues related to heat exchanger and steam generator fouling and structural integrity. ... US Department of Energy (NEER) Thank you! Questions? Nuclear Engineering ... – PowerPoint PPT presentation

Number of Views:99
Avg rating:3.0/5.0
Slides: 44
Provided by: kristin133
Category:

less

Transcript and Presenter's Notes

Title: MONITORING SYSTEMLEVEL AND DEVICELEVEL ANOMALIES IN STEAM GENERATORS AND HEAT EXCHANGERS


1
MONITORING SYSTEM-LEVEL AND DEVICE-LEVEL
ANOMALIES IN STEAM GENERATORS AND HEAT EXCHANGERS
Belle R. Upadhyaya University of Tennessee,
USA Evren Eryurek, Emerson Process Management,
USA Presented at the IAEA Technical
Meeting On-line Condition Monitoring of Equipment
and Processes in Nuclear Power Plants Using
Advanced Diagnostic Systems Knoxville, Tennessee,
USA 27-30 June 2005
2
Overview
  • Objectives of Research Development (RD).
  • Issues related to heat exchanger and steam
    generator fouling and structural integrity.
  • Analytical studies.
  • Development of Heat Exchanger (HX) FDI method.
  • Development of Steam Generator structural
    integrity monitoring.
  • Experimental studies of tube-and-shell heat
    exchanger.
  • System and device level monitoring using the
    GMDH method.
  • Concluding remarks and future work.

3
RD Objectives
  • To develop a monitoring, detection, and
    isolation system for system-level and
    device-level degradation of industrial heat
    exchangers and steam generators Integration of
    physics-based and data-based models.
  • To study the particulate fouling progression
    behavior and the effects of different process
    variables Experimental HX loop with flow,
    temperature, and DP instrumentation.
  • To develop a structural integrity monitoring
    technique using transient acoustic signal
    analysis of piezo-sensor measurements.

4
Need for Process Unit Monitoring in Nuclear Power
Systems
  • Provide timely information to operation and
    maintenance personnel.
  • Minimize unscheduled downtime of process units.
  • Track process sensors and minimize de-rating
    effects, and thus improve the unit economy.
  • Meet the needs of longer fuel cycles in next
    generation reactors.

5
Integrated Monitoring and Diagnosis System
6
Particulate Fouling
  • Accumulation of solid particles suspended in a
    fluid onto a heat transfer surface.
  • Suspended particles can be ambient pollutants
    (sand, silt, clay), upstream corrosion
    products, or products of chemical reactions
    occurring within the fluid.
  • Transport of particles.
  • Deposition or adhesion of particles.
  • Removal of particles.

7
Particulate Fouling Behavior
  • Kern Seaton Analysis of thermal surface
    fouling.
  • Müller-Steinhagen et al Influence of operating
    conditions on particulate fouling and HX
    studies.
  • Melo et al Particle transport in fouling by
    KAOLIN-water suspensions in copper tubes.
  • Particulate fouling resistance model
  • Rf Asymptotic fouling resistance, ? time
    constant

8
A Nuclear Power Plant has Many Heat Exchanger
Equipment
9
U-Tube Steam Generator (UTSG)
10
Double-pipe, Parallel-flowHeat Exchanger
11
Heat Exchanger Model
  • heat transfer rate
  • A surface area on which the overall heat
    transfer coefficient, U, is based.
  • logarithmic mean temperature
    difference.
  • For parallel or concurrent
  • flow HX system

12
Schematic of HX Diagnostics System
13
Experimental Study of Time Behavior of
Particulate Fouling in a Heat Exchanger-1
  • Test system a small-scale HX -- 31 copper tubes,
  • shell diameter 2.125 inch, tube length 24
    inch, and tube
  • outer diameter 0.25 inch.
  • Fouling particulate used in experiments KAOLIN.
  • Method of monitoring the fouling deposit Track
    the
  • overall thermal resistance by continuously
    measuring
  • the related process variables and estimating
  • performance parameters.

14
Experimental Study of Time Behavior of
Particulate Fouling in a Heat Exchanger-2
  • Parallel or concurrent flow pattern is used.
  • The electrically heated water, mixed with
  • KAOLIN particles, is designed to flow through
  • the tube-side of the heat exchanger.
  • Cooling water flows through the shell-side
  • of the heat exchanger.

15
Experimental HX Loop Showing Various Measurements
16
HX Experimental Loop
17
Calculation of the Overall Thermal Resistance
  • KAOLIN concentrations 1,647 ppm, 2118 ppm, 2588
    ppm.
  • Constant flow rate.
  • Experimental run time up to 100 hours.
  • Measurements of inlet and outlet temperatures
    (on both sides of HX), and flow rate on the cold
    side.
  • After each run, all the 31 tubes were carefully
    cleaned, and the experiment was run to verify
    the non-fouled state of the HX.

18
Effect of particle concentration on fouling
progression Measured signals
19
Variation of the overall thermal resistance of
the heat exchanger vs. run time
Before heat exchanger tubes were cleaned
After heat exchanger tubes were cleaned
20
Variation of the pressure drop across the heat
exchanger tube-side versus run time
Before heat exchanger tubes were cleaned
After heat exchanger tubes were cleaned
21
Remarks on the Influence of Particulate
Concentration on Fouling Rate
  • Particulate fouling in a small-scale heat
    exchanger exhibits an asymptotic behavior.
  • Results show that the method and design of the
    experimental setup for particulate fouling
    study were correct and successful.
  • Comparing the results of the three runs with
    different particle concentration, we see that
    when the particle concentration increases, both
    the asymptotic thermal resistance and the
    pressure drop across the heat exchanger also
    increase.
  • Longer run time is required for the thermal
    resistance to reach asymptotic values.

22
View of the HX Tubing after Particulate Fouling
(End plate opened for cleaning)
23
Process Variable Prediction Using Group Method of
Data Handing (GMDH)
  • GMDH is a data-based modeling technique that
    predicts one variable as a function of other
    related variables.
  • It creates a general polynomial fit to the data
    by successively increasing the order of the fit
    and generates Gabor-Kolmogorov polynomials.

24
GMDH Model Development by Successive
Approximation (Farlow)
Rational functions of input variables (x1, x2)
may be used to enhance dynamic system
characterization.
25
GMDH prediction of the hot-side outlet
temperature of the heat exchanger as compared
with the experimental data
26
Residual between the GMDH prediction of the
hot-side outlet temperature of the heat exchanger
and the experimental data
27
Remarks on GMDH Modeling of Process Measurements
  • The GMDH estimation residual follows the same
    behavior as the overall thermal resistance as a
    function of the experimental run time.
  • These results suggest that we can use the
    residual trending to monitor and diagnose the
    fouling problem that is occurring in a heat
    exchanger.

28
Structural Defect Monitoring Research Objectives
  • Develop new and innovative methods for the
    detection, location, and classification of
    defects in metal plates and tubing.
  • Study the propagation of Lamb waves in
    structures.
  • Develop piezo-transducer suites for structural
    interrogation and acquisition of acoustic
    signals.
  • Develop advanced signal processing techniques for
    feature extraction and analysis Hilbert-Huang
    Transform for non-stationary and nonlinear
    signals.

29
Detection, Location and Isolation of Flaws using
Piezo-Transducers
Piezo- Amplifier
LabVIEW Interface
Data Acquisition (DAQ)
Active Piezo Sensor
Passive Receiving Sensor
Passive Receiving Sensor
30
Detection, Location and Isolation of Flaws using
Piezo-Transducers
31
Hilbert-Huang TransformationRemarks
  • Hilbert Transform (HT) provides an analytical
    signal that has Fourier transform properties same
    as the original signal (within a scale factor).
  • HT provides amplitude versus instantaneous
    frequency information as a function of time.
  • The HT overcomes the limitations of data window.
  • The EMD technique decomposes a signal into
    intrinsic mode functions that have unique
    frequency features as a function of time.
  • This property provides an effective processing of
    a non-stationary/nonlinear signal using the HHT.

32
Concluding Remarks andFuture Work
  • The objective of the research was to develop a
    combination of physics-based and data-based
    models that could be implemented for on-line
    monitoring and diagnosis of HX systems.
  • Process measurements during tube fouling were
    generated using an experimental HX loop.
  • Group Method of Data Handling (GMDH) has been
    effective in process characterization.
    System-level and device level anomalies can be
    tracked by multiple information integration.

33
Concluding Remarks andFuture Work (cont.)
  • Our contribution is not in illustrating the
    particulate fouling behavior.
  • Experimental data were conclusive for multiple-
    tube heat exchangers.
  • On-line monitoring implementation is in progress
    at Emerson Process Management.
  • Future work will include the extension of this
    technology to feed-water heaters, steam
    generators, condensers, and other process units.

34
Concluding Remarks andFuture Work (cont.)
  • The HHT was used to detect, locate, and classify
    defects in both flat beams and tubular
    structures.
  • A more comprehensive piezo-sensor suite must be
    developed for applications to large plate-like
    structures (such as pressure vessels).
  • Wireless transmission of signals must be
    developed to minimize the cabling requirements
    in industrial applications.

35
Acknowledgments
  • Emerson Process Management
  • US Department of Energy (NEER)
  • Thank you!
  • Questions?

36
Development of HX FDI Method
  • Identify leakage problem by checking the inlet
    and outlet flow rates.
  • Detect sensor problem by checking the heat
    balance between the shell side and the tube side
    or using GMDH model.
  • Develop a knowledge base (e.g. GMDH model) for
    normal operating conditions.
  • Detect fouling problem using the established
    knowledge base.

37
Hilbert Transform
  • Hilbert Transform
  • Define the analytical function
  • Instantaneous frequency

38
Empirical Mode Decomposition (EMD) of the Signal
for Brass tube
39
Empirical Mode Decomposition (EMD) of the Signal
for Brass tube
40
Least-distortion Signal Pre-processing Using EMD
41
Hilbert-Huang Transformation of the Signal for
Aluminum Plate
42
Hilbert-Huang Transformation of the Signal for
Aluminum Plate
43
Innovation Through Collaboration
Write a Comment
User Comments (0)
About PowerShow.com