flow modeling and testing of esps - airflow sciences
TRANSCRIPT
![Page 1: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/1.jpg)
1
Flow Modeling and Testing of ESPs
ICESP 2018
Charlotte, NC, USA
Robert Mudry, P.E.
Airflow Sciences CorporationLivonia, MI, USA
![Page 2: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/2.jpg)
2
Outline
Introduction
ESP Fluid Flow Basics
Flow Modeling
• Physical scale
• CFD
Questions
![Page 3: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/3.jpg)
3
Boiler ESP ID Fan Stack
ESP Fluid Flow Basics
Primary goal: Maximize particulate capture
Image: Gerry Klemm, Southern Company
![Page 4: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/4.jpg)
4
ESP Fluid Flow Basics
Flow Related Goals
• Uniform gas velocity through
collection fields
• Equal flow balance between
multiple chambers
• Avoid hopper re-entrainment
• Avoid particulate deposition / drop out
• Minimize peak velocities / erosion
• Minimize pressure drop
• Uniform temperature distribution
• Sorbent injection / mixing
![Page 5: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/5.jpg)
5
Gas Velocity Distribution
Uniform velocity within collection region
Industry standards
• ICAC
• RMS/CV
deviation
ICAC: 85% of velocities ≤ 1.15 * Vavg
99% of velocities ≤ 1.40 * Vavg
Other: % RMS Deviation ≤ 15% of Vavg
![Page 6: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/6.jpg)
6
Gas Flow Balance
ICAC goal: flow within ±10% per chamber
21 %
35 %
26 %
18 %
Percent of total mass
flow through each
chamber
ICAC: Flow within each chamber to be
within ±10% of its theoretical share
![Page 7: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/7.jpg)
7
Hopper Re-entrainment
Avoid gas flow in hoppers
Avoid sweepage
Particles escape due
to gas flow in hopper
Image: Gerry Klemm, Southern Company
![Page 8: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/8.jpg)
8
Particulate Deposition
Duct floors
Turning vanes
Perforated plates
![Page 9: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/9.jpg)
9
Particulate Deposition
Extreme deposition = structural failure
![Page 10: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/10.jpg)
10
Peak Velocities / Erosion
Erosion potential depends on
• Impact velocity (to the 2.5 power)
• Flow rate
• Hardness, shape
• Angle of impact
Erosion issues
• Vanes
• Baffles
• Perforated plates
• Ductwork
• Trusses
![Page 11: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/11.jpg)
11
Pressure Drop
General goal:
• Minimize DP
Ductwork redesign saves
2.1 IWC (520Pa)
To
tal P
res
su
re
![Page 12: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/12.jpg)
12
Gas Temperature
Temperature stratification
• Resistivity
• Corrosion
• Sorbent performance
Temperature
Res
isti
vit
y
![Page 13: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/13.jpg)
13
Injection Systems
Gaseous injection
• SO3, NH3, others
Particulate injection
• Activated carbon
• Trona, SBS, lime, etc.
![Page 14: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/14.jpg)
14
Sorbent Injection
Goals
• Uniform distribution
• Surface contact / residence time
• Mixing / dispersion
![Page 15: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/15.jpg)
15
ESP Modeling
Physical Model Testing
Computational Fluid Dynamics (CFD)
![Page 16: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/16.jpg)
16
ESP Modeling – Physical Models
Background
Theory
Simulation Parameters
Fabrication Details
Results Analysis
Flow Visualization
![Page 17: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/17.jpg)
17
Physical Models – Background
Utilized for fluid flow analysis for a century … or more?
Applied to ESPs for decades
Underlying principle is to reproduce fluid flow behavior in
a controlled, laboratory environment
Image: NASAImage: Smithsonian
![Page 18: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/18.jpg)
18
Physical Models – Theory
Key criteria is to generate “Similarity” between
the scale model and the real-world object
• Geometric similarity
Accurate scale representation of geometry
Inclusion of all influencing geometry elements
Selection of scale can be important
• Fluid dynamic similarity
Precise Reynolds Number (Re) matching is not feasible
General practice is to match full scale velocity but ensure
that Re remains in the turbulent range throughout the model
Re =ρ v Dh
μ
![Page 19: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/19.jpg)
19
Physical Models – Parameters
ESP geometry
• Typically 1/8th to 1/16th scale
• Include features affecting flow patterns
Flow conditions
• Scaled air flow rate (at
ambient temperature)
• Simulate injections with a
tracer gas
• Simulated particle tracking
In flight
Deposition
![Page 20: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/20.jpg)
20
Physical Models – Fabrication Details
More detail is better
• Vanes, splitters, baffles
• Trusses, gusset plates
• Dampers, columns
![Page 21: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/21.jpg)
21
Physical Models – Fabrication Details
Perforated plate selection
Collection plate representation
• Geometry / loss coefficient
• Re consideration
![Page 22: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/22.jpg)
22
Physical Models – Results Analysis
Quantitative data
• Available at discrete measurement points
• Velocity magnitude, directionality
• Pressure (corrected to full scale)
• Tracer gas concentration
• Mass balance between chambers
• Comparison to ICAC conditions
• Correlation to test data
Qualitative data
• Flow directionality (smoke, tufts)
• “Dust Test” for particle behavior, drop-out, re-entrainment
![Page 23: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/23.jpg)
23
Physical Models – Velocity Results
Uniformity vs. ICAC goals
![Page 24: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/24.jpg)
24
Physical Models – Dust Testing
![Page 25: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/25.jpg)
25
Physical Models – Visualization
Smoke flow
Helium bubbles
Images: Gerry Klemm, Southern Company
![Page 26: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/26.jpg)
26
Physical Models – Visualization
Video footage
![Page 27: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/27.jpg)
27
ESP Modeling – CFD
Background
Theory
Simulation Parameters
Mesh Details
Results Analysis
Flow Visualization
![Page 28: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/28.jpg)
28
CFD – Background
Developed in the aerospace industry c.1970 (with the
advent of “high speed” computers)
Applied to ESPs for 30+ years
Underlying principle is to solve the first-principles
equations governing fluid flow behavior using a
computer
Image: NASAImage: USAF
![Page 29: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/29.jpg)
29
CFD – Theory
Control Volume Approach
• Divide the flow domain into distinct control volumes
• Solve the Navier-Stokes equations (Conservation of
Mass, Momentum, Energy) in each control volume
Inflow Outflow
Control Volume or
“Cell”
ESP model with
12,500,000 cells
![Page 30: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/30.jpg)
30
CFD – Mesh Details
A good mesh is critical for accurate results
• Cell count
• Length scale
• Local refinement
• Shape/topology
Hex vs. Tet vs. Poly
• Flow alignment
Why?
• Finer resolution = more accurate results
• Numerical calculation scheme accuracy is influenced
by cell topology
![Page 31: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/31.jpg)
31
CFD Mesh Quality – Cell Topology
Not all CFD cells are created equal
• Hexahedral vs. Tetrahedral vs. Polyhedral
• Tet mesh = easier, automeshing
• Hex or hybrid mesh = harder, hand built
• 5 million Hex cells = 30 million Tet cells
Hexahedral cell(6 faces)
Tetrahedral cell(4 faces)
6 Tet cells fit into 1 Hex cellof the same length scale
![Page 32: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/32.jpg)
32
CFD – Mesh Details
Direct geometry inclusion
• Vanes, baffles
• Structure
Numerical simplification
• Perforated plates
• Collection plates
![Page 33: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/33.jpg)
33
CFD – Simulation Parameters
ESP geometry (mesh) Full scale representation
Include features important to flow, more detail if possible
Flow conditions Full scale gas flow rate
Reproduce velocity & temperature profile at model inlet
Simulated chemical injection
Simulated particle tracking
Solver Azore
Fluent
Star CCM+
![Page 34: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/34.jpg)
34
CFD – Results Analysis
Quantitative data available at all control volumes Velocity magnitude, directionality
Temperature
Pressure
Turbulence
Chemical species concentrations
Particle trajectories
Integrated/reduced data Mass balance between chambers
Comparison to ICAC conditions
Correlation to test data
![Page 35: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/35.jpg)
35
CFD – Results
Velocity patterns
Uniformity vs. ICAC goals
• All cells vs. traverse
Pressure drop
![Page 36: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/36.jpg)
36
CFD – Advanced Modeling
Electrostatics
Transient phenomenon
Large meshes and HPC solvers
Pressure pulses during to rapping
• Hopper “splash”
• Minimizing re-entrainment
Chemistry / absorption
FINE CARBON PARTICLES EJECTED FROM
HOPPERS
LARGE CLUSTERS
SHEARED
DURING RAPPING
RAPPER FIRES
2
1
3
Image: Gerry Klemm,
Southern Company
![Page 37: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/37.jpg)
37
CFD – Flow Animations
Video footage
![Page 38: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/38.jpg)
38
Model Correlation – Field Testing
Velocity Distribution in ESP
• Cold flow conditions
• Vane anemometer
Rides on plates / electrodes
Accuracy 1% in 3-10 ft/sec range
Lightweight, portable
Sensitive to flow angularity, turbulence, dust
• Difficult, expensive test
![Page 39: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/39.jpg)
39
Accuracy can be influenced by
• Proximity to perforated plate
• Collection plate geometry
• Electrode geometry
Scrutinize the test data closely
Field Testing – Collection Region
![Page 40: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/40.jpg)
40
Flow Model Accuracy
Do physical and CFD models provide useful
results, suitable for ESP design?
• Yes, experience has shown that modeling is a
dependable engineering tool
• Useful for new ESPs and improving existing ESPs
• Needs to be performed correctly and to the best of
current modeling practices
![Page 41: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/41.jpg)
41
Flow Model Accuracy
Do the model results match actual plant data?
• Not as well as one would like in some cases
• Error bars are larger than desired
• Correlation can be dependent on ESP geometry
Nozzle style inlets are particularly challenging for models
Square and ladder vane style inlets have stronger correlation
What to do when they don’t match?
• Re-evaluate the modeling, improve the methods
• Scrutinize the test data
![Page 42: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/42.jpg)
42
Flow Model Accuracy
Do CFD and physical models match each other?
• Sometimes but not all
• Correlation of uniformity statistics vs. profile
• Limitations of both methods hamper correlation
• Can be dependent on ESP geometry
Inlet style – nozzle vs. square
Flow devices and complexity
• Research is ongoing
• Speaker’s experience in this area focuses on
1. Perforated plate representation
2. Collection plate representation
![Page 43: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/43.jpg)
43
Summary – Industry Experience
Hundreds of successful projects industry wide
• Performance guarantees met
• Reduced emissions / DP / maintenance
• Field verification not often performed
Improved modeling methods would allow for
additional ESP performance optimization
![Page 44: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/44.jpg)
44
Summary – Key Takeaways
Both CFD and physical modeling have been
widely used for decades
From a general view, the modeling methods
provide beneficial engineering tools for design
Both methods have specific procedures, and
need to be done correctly to yield usable results
Both methods have their limitations, their
supporters, and their critics
Research is ongoing on both methods, striving
to improve accuracy and correlation
![Page 45: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/45.jpg)
45
Acknowledements
Dan Fahrer, DTE Energy
Gerry Klemm, Southern Company
Dr. Jia Mi, Southern Company
Jose Sanchez, EPRI
![Page 46: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/46.jpg)
46
Questions?
![Page 47: Flow Modeling and Testing of ESPs - Airflow Sciences](https://reader036.vdocument.in/reader036/viewer/2022062306/62a6891e6fbb1f674069417e/html5/thumbnails/47.jpg)
47
Contact Information
www.AirflowSciences.com
www.AzoreCFD.com
www.AirflowSciencesEquipment.com
Airflow SciencesCorporation
12190 Hubbard Street
Livonia, MI 48510-1737 USA
Tel. +1.734.525.0300
Robert G. Mudry, P.E.