ENVE 4003

1
ENVE 4003

• PARTICULATE MATTER Primary emission control
  devices
• Wall collection devices

2
PRIMARY PARTICULATES CONTROL DEVICES

• WALL COLLECTION DEVICES
• Settling chambers
• Cyclones
• Electrostatic precipitators
• DIVIDING COLLECTION DEVICES
• Filters (surface and depth)
• Scrubbers

3
GRAVITY SETTLING CHAMBERS

• Gas behaviour can be characterized by limiting
  cases plug flow, mixed flow (see note, next
  slide)
• Particle removal efficiency related to
• residence time in chamber
• terminal settling velocity
• distance to travel before hitting wall

4
Note The names plug flow (or block flow) model,
and mixed flow model have been used differently
from common usage in reactor notation

• de Nevers
• Plug flow or block flow
• No mixing in the direction of fluid flow (x,
  horizontal),
• no mixing in the transverse direction of particle
  motion (y, horizontal, or z vertical)

• Levenspiel
• Plug flow
• No mixing in the direction of fluid flow (x,
  horizontal),
• well mixed in transverse direction (y,
  horizontal, or z vertical)
• de Nevers calls this the mixed flow model

5
Note The names plug flow (or block flow) model,
and mixed flow model have been used differently
from common usage in reactor notation

• de Nevers
• Mixed flow
• No mixing in the direction of fluid flow (x,
  horizontal),
• well mixed in transverse direction (y,
  horizontal, or z vertical)
• Levenspiel calls this plug flow

• Levenspiel
• Mixed flow
• well mixed in all directions (x, y, z)

6
Figure 9.1 de Nevers

• Gravity settler

7
SETTLING CHAMBER CAPTURE EFFICIENCIES

• Plug flow
• Mixed Flow

8

• Example 9.1 calculates efficiencies as a function
  of particle diameter (hence terminal settling
  velocity) for the two models using
• Height 2 m
• Length 10 m
• Vgas 1 m/s
• Results plotted in Fig 9.2

9
Figure 9.2 de Nevers

• Plug flow and mixed flow efficiencies for gravity
  settler, Example 9.1
• H 2 m, L 10 m, V 1 m/s

10
PRIMARY PARTICULATES CONTROL WALL COLLECTION
DEVICES

• Gravity settling is effective for large particles
  ( more than 100 micrometers), in
  reasonably sized chambers
• For smaller particles, the terminal settling
  velocity is too small
• IDEA Impose an external force greater than
  gravity
• Centrifugal - CYCLONES
• Electrostatic - ESP

11
CYCLONES

• At 60 ft/s circular velocity and 1 ft radius
• With corresponding increase in terminal velocity.

12
Figure 9.4 de Nevers

• Cyclone

13
CYCLONES

• Principles similar to settling chambers
• More complex geometry and flow patterns

14
CYCLONES

• If we use the Stokes law for settling velocity

15
IMPROVED CYCLONE EFFICIENCY

• Efficiency increases with increasing Vcircular.
  
• But, pressure drop is proportional to V2circular
  
• Reduce inlet duct Width (and diameter in
  proportion)
• Split flow into multiple cyclones to keep
  Vcircular constant
• MULTICLONES

16
Figure 9.5 de Nevers

• Multiclone

17
DIMENSIONAL RATIOS IN CYCLONE DESIGN

• Optimize cyclone dimensions for increased
  efficiency vs reduced pressure drop
• General types
• High efficiency
• Conventional
• High throughput

18
Figure 4.4, Cooper Alley
19
Table 4.1, Cooper Alley
20
Figure 4.3, Cooper Alley
21
CYCLONE COLLECTION EFFICIENCY WITH PARTICLE SIZE
DISTRIBUTiON

• Collection efficiency varies with particle
  terminal velocity, which in turn varies with
  particle diameter D and density
• Cut Diameter Dcut is the diameter which has
  collection efficiency of 50

22
CYCLONE COLLECTION EFFICIENCY ESTIMATE OF CUT
DIAMETER

• Using Stokes region expression for Vterminal and
  plug flow model (neither of which are
  particularly good representations of the actual
  situation) we can obtain
• This turns out to be a reasonable estimate of
  Dcut
• Empirical data on standard cyclones is required
  for more precision

23
Collection efficiency vs particle diameter
24

• Empirical collection efficiency vs particle
  diameter behaviour of typical cyclones

25
Figure 9.6 de Nevers (Example 9.6)

• Eqn 9.18 plug flow and Stokes law
• Eqn 9.19 mixed flow and Stokes law
• Eqn 9.21 empirical

26
Example 9.6 de Nevers

• Performance computation for a cyclone separator
  of Dcut 5 ?m with log normally distributed
  particle size
• D mass mean 20 ? m,
• ? 1.25
• (log normal distribution previously demonstrated
  in Fig 8.8. (8.10) de Nevers)

27
Table 9.1 de Nevers (Example 9.6)

• Performance computation for a cyclone separator

28
CYCLONE DESIGN

• Standard dimensional ratios based on
  accumulated experience are available for specific
  objectives (high efficiency, high flow
  throughput, or a compromise)
• Cyclone manufacturer may provide empirical Dcut
  vs Qgas and pressure drop vs Qgas data
• An iterative (trial and error) procedure required
  to find Dcyclone for desired collection
  efficiency
• Given gas flow with known particle size
  distribution, choose Dcyclone , calculate
  collection efficiency for each particle size and
  overall, repeat.

29
ELECTROSTATIC PRECIPITATORS(Cottrell
precipitators)

• Principle charge the particles, use
  electrostatic force to attract them to wall

30
Figure 9.7 de Nevers

• Sketch of ESP

31
ELECTROSTATIC PRECIPITATORS(Cottrell
precipitators)

• Note similarities of geometry between settling
  chamber and ESP.
• H the height through which particles must
  travel, at right angles to gas flow, before
  hitting wall
• L distance travelled by gas in the collection
  device.
• The H will be smaller in ESP, the velocity of
  particles much higher because of the
  electrostatic force.

32

• Corona discharge at the wire electrons collide
  with gas molecules, knock out electrons,
  positively charged gas ions migrate to wire and
  discharge particles
• Field charging away from the wire as electrons
  fly towards wall, they collide with particles in
  their path and are captured by particles,
  negatively charged particles attracted to wall
  and discharge there.
• Diffusion charging for particles smaller than
  0.15 µm, the interaction with electrons can be
  significantly due to their random motion as a
  result of electron-gas molecule collisions

33
Maximum charge on particles
34
Drift velocity (I.e. terminal settling velocity
under electrostatic force)

• Force on particle F qE
• Resulting terminal settling velocity (with
  Stokes law for drag force)

35
Collection efficiency

• Block flow
• Mixed flow

36
(No Transcript)
37
ESP Performance and cake resistivity

• High resistivity ash
• - large ?Vcake , small ?Vwire, poor charging,
  low ?
• - electron flow within cake, back corona
• Low resistivity ash
• - small ?Vcake , weak attraction to collection
  plate, re-entrainment

38
Figure 9.10 de Nevers

• Voltage-distance relation for different ash
  resistivities

39
ESP Performance and cake resistivity

• Resistivity 1/conductivity
• Conductivity surface volume
• Volume conductivity determined by chemical
  composition of particle
• Surface conductivity determined by chemical
  composition of gas
• Remedies for high resistivity ash
• - Higher temperatures, hot ESP (improves volume
  conductivity)
• - Gas conditioning, add hygroscopic components
  to gas to improve surface conductivity. SO3 for
  basic coal ash, NH3 for acidic cement ash .

40
Figure 9.9 de Nevers

• ESP collection efficiency for coal of differeent
  S content