EPA-600/2-76-140
May 1976
Environmental Protection Technology Series
                        RAPPING  REENTRAINMENT  IN  A
                             NEARLY FULL-SCALE PILOT
                         ELECTROSTATIC  PRECIPITATOR
                                   industrial Environmental Research Laboratory
                                        Office of Research and Development
                                       U.S. Environmental Protection Agency
                                  Research Triangle Park, North Carolina 27711

-------
                RESEARCH REPORTING SERIES

 Research reports of the Office of Research and Development, U.S. Environmental
 Protection Agency,  have been grouped into five  series. These five broad
 categories were established to facilitate further development and application of
 environmental technology. Elimination of traditional grouping was consciously
 planned to foster technology transfer and a maximum interface in related fields.
 The five series are:

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research
     4.    Environmental Monitoring
     5.    Socioeconomic Environmental Studies

 This report has been  assigned  to the ENVIRONMENTAL  PROTECTION
 TECHNOLOGY series. This series describes research performed to develop and
 demonstrate instrumentation,  equipment, and methodology to repair or prevent
 environmental degradation from point and non-point sources of pollution. This
 work provides the new  or improved technology required for the control and
 treatment of pollution sources to meet environmental quality standards.
                    EPA REVIEW NOTICE

This report has been reviewed by  the U.S.  Environmental
Protection Agency, and approved for publication.  Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                    EPA-600/2-76-140

                                    May 1976



        RAPPING REENTRAINMENT

        IN A NEARLY FULL SCALE

  PILOT  ELECTROSTATIC  PRECIPITATOR
                    by

           HerbertW. Spencer, HI

         Southern Research Institute
           2000 Ninth Avenue South
        Birmingham, Alabama 35205
           Contract No. 68-02-1875
            ROAPNo. 21ADL-027
        Program Element No. 1AB012
    EPA Project Officer: Leslie E. Sparks

 Industrial Environmental Research Laboratory
   Office of Energy, Minerals, and Industry
      Research Triangle Park, NC 27711
               Prepared for

U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Research and Development
            Washington, DC 20460

-------
                           ABSTRACT

      The results of an experimental investigation of rapping
reentrainment in a nearly full scale pilot electrostatic pre-
cipitator are reported.  The objectives of the study included
a fundamental study of the mechanics of removal of dry dust
by rapping and the quantification of rapping reentrainment in
terms of the percentage and the particle size distribution of
       i  "     *rn»M 4MB ^^ tf>^ tfm
the ree-ntrained*^isjfe^»tDuring this study, the contribution of
rapping reentra£njndn.tjto total emissions ranged from 53% to
18% depending on rapping conditions.  These percentages corre-
sponded to 5.4% and 2.7% of the dust collected on the plates
being emitted from the precipitator during plate rapping.  A
major portion of the reentrained material was observed to
result from hopper "boil-up".  A decrease in rapping emissions
was obtained by increasing the time interval between raps.
The mass median diameters of the size distributions of the
particles emitted during the raps were observed to increase
with increased time between raps.
                               11

-------
                      TABLE OF CONTENTS








                                                        Page




Abstract                                                 ii



List of Figures                                          iv



List of Tables                                           ix



Acknowledgements                                         xi








Sections



    I     Summary                                         1



   II     Introduction                                    5



  III     Experimental Arrangement                       19



   IV     Test Conditions                                46



    V     Results and Discussion                         76



   VI     References                                    149



  VII     Appendix                                      152
                             111

-------
                           FIGURES

No.

 1   Theoretical efficiency of dust removal as a
     function of dust layer thickness                     10

 2   Rapping efficiency as a function of mass per unit
     area for a precipitated fly ash at 150°C, rapped
     with a maximum normal acceleration of 1.3 x 10 **
     cm/sec2 (13 G); measured by Sproull.5                11

 3   Near full scale pilot precipitator at FluiDyne
     Engineering                                          20

 4   Various views of FluiDyne pilot precipitator         21

 5   Design details for pilot precipitator plates         24

 6   Block diagram of experimental layout for rapping     29
     reentrainment study

 7   Sampling locations during the rapping reentrain-
     ment study                                           30

 8   Schematic of diluter for large particle sizing
     system                                               36

 9   Block diagram of large particle sizing system        37

10   Load cell                                            39

11   Precipitator plate weight measurements system
     block diagram                                        41

12   Accelerometer system block diagram                   42

13   Cumulative particle size distribution of the feed
     dust, Banco analysis                                 48

14   Resistivities measured in Southern Research
     Institute laboratory for fly ash used during the
     test                                                 50

15   Precipitator electrical characteristics              56

16   Precipitator outlet velocity profile                 58
                             IV

-------
                           FIGURES

                         (continued)


No.                                                     Page

17   Precipitator inlet velocity profile                  60

18   Precipitator outlet velocity profile                 61

19   Correlation of measured feed rate and hopper
     feed rate                                            65

20   Plate 5-1 accelerations as a functional differ-
     ential pressure on rapper, upstream rapper,
     dust on plate.                                       69

21   Plate 5-1 accelerations as a functional differ-
     ential pressure on rapper, upstream rapper,
     no dust on plates.                                   70
22   Typical traces of plate accelerations for a
     dust covered plate [plate 5-2, 3m from top of
     plate, AP = 6.2 x 10^ Pa (90 psi)].
73
23   Typical traces of plate accelerations for a
     clean plate [plate 5-2; KI, Ylf z\, 0.9 m from
     top of plate; Xa, Y2 3m from top of plate;
     AP= 6.2 x 105 Pa (90 psi)].                          75

24   Mass emission rates for lower 32% of the pre-
     cipitator cross sectional area as a function
     of the time interval between raps                     78

25   Mass emission rates for upper 68% of the pre-
     cipitator cross sectional area as a function
     of the time interval between raps                     79

26   Combined upper and lower mass emission rates as
     a function of the time interval between raps          80

27   Particulate loading - lower 32% of the precip-
     itator as a function of time interval between
     raps                                                  81

28   Particulate loading - upper 68% of the. precip-
     itator as a function of time interval between
     raps                                                  82

-------
                           FIGURES

                         (continued)

No.                                                     Page
29   Combined particulate load as a function of time
     interval between raps                                83

30   Percent of dust collected on precipitator plates
     that is emitted due to rapping                       87

31   Inlet and weighted outlet particle size distribu-
     tions, dm/d log D                                    90

32   Inlet cumulative percent particle size distribution  91

33   Inlet and outlet (lower 32% and upper 68% of the
     precipitator)  particle size distribution,
     dm/d log D                                           92

34   Outlet emission rates between raps for various
     size intervals                                       94

35   Outlet emission rates measured with the impactors
     as a function of the time interval between raps      95

36   Average particulate load during a rap as a function
     of the time interval between raps                    96

37   Particle size distribution for rapping puffs,
     dm/d log D's (average volume of gas sampled dur-
     ing the rapping puffs was determined from the
     sampling rate and the average persistence time
     of the rapping puffs)  tests 5, 4 & 6, 7 & 8
     had 12, 32, and 52 minutes, respectively,
     between-rap intervals                                97

38   Cumulative percent distribution for rapping puffs,
     lower 32% of the precipitator cross sectional
     area, rapping intervals of 12, 32 and 52 minutes     98

39   Mass emission rates for rapping reentrainment,
     lower 32% of precipitator exit for various
     particle size intervals                             100

40   Mass emission rates for rapping reentrainment,
     upper 68% of precipitator exit for various particle
     size intervals                                      101
                             VI

-------
                           FIGURES

                         (continued)

No.                                                     Page

41   Percent emissions due to rapping reentrainment
     as a function of particle size                      102

42   Spatial distribution of between rap emissions for
     various particle sizes                              105

43   Spatial distribution of particles in upstream
     and downstream rapping puffs for 12 minute rap
     interval test                                       108

44   Spatial distribution of particles in combined up-
     stream and downstream rapping puff                  109

45   Rapping puffs lower portion of the precipitator,
     32 minutes between raps.  First set of puffs,
     upstream rap.  Second set of puffs, downstream rap. HO

46   Rapping puffs upper portion of the precipitator,
     32 minutes between raps.  First set of puffs,
     upstream rap.  Second set of puffs, downstream rap. HI

47   Rapping puffs, lower portion of the precipitator,
     12 minutes between raps.  First set of puffs,
     upstream rap.  Second set of puffs, downstream rap. H2

48   Rapping puffs, upper portion of the precipitator,
     12 minutes between raps.  First set of puffs,
     upstream rap.  Second set of puffs, downstream rap. 113

49   Percent contribution of a rapping puff to total
     emissions for several particle size bands as a
     function of the time interval between raps          116

50   Correlation of mass of dust removed from plate 4-1
     and amplitude of Konitest meter current output for
     the upstream rapping puff for test 5 (12 minute
     rap interval)                                       120

51   Load cell data, mass collected on plate 4-1,
     12 minute interval between raps                     123

52   Load cell data, mass collected on plates 4-2
     and 4-3, 12 minute intervals between raps           124
                             VI1

-------
                          FIGURES

                        (continued)

No.                                                     Page

53   Load cell data, mass collected on plate 4-1,
     32 minute intervals between raps                    125

54   Load cell data, mass collected on plates 4-2
     and 4-3, 32 minute intervals between raps           126

55   Collection rate of particulate as a function of
     time interval between raps                          127

56   Dust removal efficiency for dust on plate 4-1
     versus the time interval between raps               130

57   Collection rate versus horizontal location in
     the precipitator                                    135

58   Fractional collection efficiency versus particle
     size, solid line theory, circles experimental
     between rap, diamonds experimental, include
     rapping reentrainment                               136

59   Percentage of total emission due to rapping
     reentrainment                                       139
                             Vlll

-------
                           TABLES

No.                                                      Page

 1     Estimates and measurements, tensile strengths
       of precipitated fly ash layers                      12

 2     Design parameters, FluiDyne pilot precipitator      22

 3     In situ resistivity data                            49

 4     Chemical analysis of Black Dog fly ash, percent
       by weight                                           52

 5     Composition of gas entering the FluiDyne pilot
       precipitator                                        53

 6     Average electrical conditions                       55

 7     Average velocities 76 cm downstream of the exit
       baffles                                             59

 8     Volume flow rates during the test                   62

 9     Inlet particulate load and particulate feed rates   64

10     Composite average of plate accelerations during
       rapping for AP = 6.2 x 105 Pa, dust covered plate   66

11     Composite average of plate accelerations during
       rapping for AP = 4.8 x 105 Pa, dust covered plate   67

12     Composite average of plate accelerations during
       rapping for AP = 6.2 x 105 Pa, clean plates         71

13     Composite average of plate accelerations during
       rapping with modified rappers for AP = 6.2 x 105
       Pa, light dust layer                                72

14     Ratio of relative total emissions between raps
       in five particle size intervals for rapping
       intervals of 12, 32, and 52 minutes                106

15     Comparison of real-time system data and impaction
       data obtained between raps                         118
                               IX

-------
                           TABLES

No.                                                      Page

16     Mass per unit area on plates when rapped, mass
       per unit area collected between raps and mass
       per unit area removed by rap                       129

17     Effect of rapping reentrainment on efficiency
       for rapping losses accounting for 20%,  30%,  40%,
       and 50% of emissions for various overall effi-
       ciencies (first approach)                           141

18     Effect of rapping reentrainment on performance
       for various reentrainment percentages as a
       function of number of sections and a constant
       collection efficiency per section (second
       approach)                                          142

19     Fraction of collected material for a given
       particle size i that was emitted due to rapping
       reentrainment                                      147
                              x

-------
                      ACKNOWLEDGEMENTS

The assistance of the staff of FluiDyne Engineering in
conducting the experiments is gratefully acknowledged.
The descriptions of the FluiDyne facility contained in
this report were drafted by Dr. Henry Hanson of FluiDyne
Engineering.
                             XI

-------
                         SECTION I
                          SUMMARY

An experimental investigation of rapping reentrainment was
performed in a nearly full-scale pilot precipitator at FluiDyne
Engineering Corporation's Rosemount Laboratory.  The objectives
of the experimental study included a fundamental study of the
mechanics of removal of dry dust by rapping and the quantifica-
tion of rapping reentrainment in terms of the percentage total
emissions and particle size distribution of the reentrained dust.

During the experiment the percentage of dust removed from the
plates was observed to depend on the mass per unit area of
dust collected on the plates as predicted by theory.  The
build-up of a residual dust layer was also observed.  A dust layer
developed that could not be removed with the available rapping
intensities (up to 20 G's).

During the study the contribution of rapping reentrainment to
total emissions ranged from 53% to 18%, depending on rapping
conditions.  These percentages corresponded to 5.4% and 2.7%,
respectively,  of the dust collected on the plates being
emitted from the precipitator during plate rapping.  A signi-
ficant decrease in rapping emissions was obtained by increasing
the time interval between raps.  This decrease was due to
the resulting larger mass per unit area collected on the plates
when they were rapped.  Particle size distribution measurements
showed that the mass median diameter of the particles emitted
during the raps increased with increased time between raps.   As
would be expected, this produced lower overall emissions.  The
increase in the size of the particles emitted during rapping
was ascribed to an increase in the agglomeration of the par-
ticulate removed from the precipitator plates with the respec-
tive increase in the mass per unit area collected on the plates
(thicker dust layers)  before the plates were rapped.

-------
A major portion of the reentrained material was observed
to result from hopper "boil-up".  A small portion of the dust
was observed to pass out the precipitator in  a  short
burst  at the velocity of the gas passing through the unit,
while the remaining portion of the material was observed to
fall into the hoppers, then to rebound, and finally to escape
slowly over the baffles and out of the precipitator.  This
produced a significant vertical gradient in the concentration
of the dust emitted from the precipitator due to rapping
reentrainment.

The fractional collection efficiencies of the precipitator
both excluding and including rapping reentrainment emissions
were determined and compared with theoretical values.   Rapping
reentrainment increased the discrepancy between theory and
measured fractional collection efficiencies for particles
larger than 5 ym.  However, rapping reentrainment did not
account for all of the discrepancy.  Reentrainment between
raps and sneakage through the unelectrified regions of the
precipitator appeared to contribute significantly to overall
emissions.

The experiments pointed out some problems in detailed mathe-
matical modelling of rapping reentrainment.  It is difficult
to predict the quantity of dust removed from a plate by an
individual rap.  The calculation of the recollection of the
reentrained material is difficult for several reasons.  The
particle size distribution of the reentrained particles can
be changed significantly by moderate changes in rapping vari-
ables.   Hopper boil-up also makes the calculation of the
recollection of the reentrained dust difficult, since much
of the reentrained material is in unelectrified regions of
the precipitator.  However, the experiments supplied some

-------
general values that could be used to estimate the effects of
rapping reentrainment on the size of a precipitator for a given
collection efficiency.

Two approaches using simplified assumptions were used to make
estimates.  The first assumed that a fixed percentage of rapping
reentrainment emissions was emitted from a precipitator indepen-
dent of the size of the unit, and the second assumed that the
same percentage of collected unclean material was reentrained
and emitted from each section due to rapping and that the recol-
lection efficiency for the reentrained material for all sections
having the same between rap collection efficiency.  The estimates
based on the above approaches showed that the increase in precip-
itator size needed to recover the rapping reentrainment emissions
can account for 6% to greater than 80% of the size of the unit
for the cases considered.

Rapping reentrainment emissions computed on the basis of the
percentages obtained at FluiDyne account for a significant
portion of precipitator emissions.

FUTURE RESEARCH
The parameters studied during the FluiDyne experiment were
somewhat limited.   A data bank for detailed modelling of
rapping needs to include additional data on rapping emissions
as a function of plate accelerations and gas velocity.  The
FluiDyne experiments also show the need for studies with multi-
section precipitators.

The FluiDyne experiments also indicated a need for a study
of the parameters affecting hopper "boil-up" and the develop-
ment of methods for removing it.  Conditioning with SO3 and
continuous removal of dust from the hoppers may be helpful,

-------
along with proper baffling.  Additional studies on the effects
of residual dust layers for high resistivity dust and studies
of dust cohesive properties are also needed to aid in deter-
mining optimum rapping conditions.

-------
                         SECTION II
                        INTRODUCTION

The efficiency of an electrostatic precipitator is governed
by the quantity of dust that passes through the electrified
region without being collected and the amount of collected
material which is reentrained in the effluent.  Reentrain-
ment losses are the result of several factors.  The primary
one is that associated with rapping of the collection plates.

This report discusses the results of an experimental
investigation of rapping reentrainment using a nearly full-
scale pilot precipitator at FluiDyne Engineering Corporation's
Rosemount Laboratory.  The work had three main objectives:
(1) a study of the basic mechanics of removal of dry dust by
rapping and the variations in the removal mechanisms with
changes in dust properties, (2) quantification of rapping re-
entrainment in terms of the percentage of the total losses,
and of the particle size distribution of the reentrained dust,
and (3) modification of the Southern Research Institute com-
puter systems model to include losses due to rapping reen-
trainment.

BACKGROUND
The removal of particulate from a gas stream by a dry electro-
static precipitator involves four separate processes.  The
first is the precipitation of the particulate from the gas
stream onto the collection plate.  A recent Southern Research
Institute report discusses the physical mechanisms involved
and describes a computer model for the removal of the dust
from the gas stream.l  The second process is the removal of
the particulate from the collection plate, and the third is

-------
the transfer of this particulate to a hopper.  The fourth
process is the removal of the collected particulate from the
hopper to a disposal area.

In present-day precipitator applications, two philosophies
are prevalent with regard to the removal and transfer of the
particulate from the collecting plates.  One approach is to
rap often and to provide maximum rapping acceleration to the
collection plates during each rap in an attempt to minimize
the thickness of the residual dust layer.  The other approach
is to vary the intensity and frequency of rapping in an
attempt to minimize the quantity of material reentrained.

The uncertainties with regard to rapping procedure are asso-
ciated with the method by which dust is actually removed and
transferred from the collection plates during a rapping sequence.
Sproull2 suggests that the rapping process occurs in a sequence
whereby the collected dust is removed by what is termed a
slip-stick process.  The concept is that the dust layer builds
up to a sufficient thickness so that the acceleration of the
plate causes a shearing force between the dust layer and the
plate.  The dust then slides along the surface of the plate
until it is recollected on the plate.  The process is repeated
with each successive rap until the dust finally falls into the
hopper.

Another concept of rapping is that the collected dust is dis-
lodged in large agglomerates or clumps and falls directly into
the hopper without being recollected.

The mechanics of the dust removal process varies with the
properties of the dust, precipitator operating conditions,
and rapping parameters.  The dust properties and precipitator

-------
operating conditions affect the adhesion and cohesion of the
dust layer.  The adhesion and cohesion of dust layers depend
upon particle-to-particle forces.  According to Tassicker3,
the forces comprise London-van der Waals, triboelectric,
capillary, surface dipole, and electric-field corona forces.
These component forces are influenced by the following:
particle diameter, porosity and compaction of the layer,
complex dielectric constant, humidity in the gas, adsorbed
surface dipolar molecules, work-function interfaces on the
material, and the electric field and current density in the
dust layer.

The tensile strength of a dust layer is the longitudinal
stress (force per unit area) that the layer can bear without
tearing apart.  Tassicker3 has formulated the following
equation for the tensile strength of a dust layer with a given
degree of compaction by appropriately combining the above
mentioned forces.
Tensile Strength of Dust for Given Degree of Compaction
                                            !9
      P = (A!/d) + (A2/d) + 105(£0/2) [e^Jpk)2 - E2]       (1)
where
      AI = van der Waals and dipole constant
      A2 = the capillary constant
      e0 = permittivity of free space
      e1 = relative permittivity of dust
       J = current density in dust layer
       p = resistivity of dust layer
       k = constant from 1 to 20 depending upon particle
           elasticity
      E  = electric field in gas space, (applied potential)/
       ^   (electrode spacing)
       d = particle diameter

-------
According  to  this  equation,  the  capillary,  van  der Waals,
and  dipole components are  always adhesive and inversely pro-
portional  to  particle diameter.  The electrical components can
be either  adhesive or nonadhesive, depending on the  resistivity
of the  particulate and the current density  in the layer.

The  development of this equation for the tensile strength of
dust layers is discussed by  Tassicker  in his dissertation.1*
Tassicker  estimates that the magnitude of the shear  strength
is probably twice  that of  the tensile  strength.  Data obtained
by Sproull5 also indicates that  shear  strengths are  relative-
ly   larger  than tensile strengths.  Experiments by  Ruckelhausen,6
Brandt,7 and  Watson and Blecher8 emphasized that it  is peak
acceleration  normal to the precipitator plates  which is effec-
tive in dislodging the collected particulate.   Nevertheless,
when shear accelerations are present,  they  must play some
role in particle dislodgement, as indicated by  the data ob-
tained by Koschany,9 who found that the peak normal  accelera-
tion required to attain a  given  percentage  removal of particu-
late from  a precipitator plate decreased as shear accelerations,
applied simultaneously, increased.

Theory of the Mechanics of Dust  Removal
An elementary theory of dust removal which  considers only the
tensile strength (P)  of the  dust layer and  the  acceleration (a)
normal to the plate has been developed by Tassicker.10  The
theory shows  that  the dust layer is removed only when

      a > TO  •  M7A
where 6 is the bulk density,  £ is the dust  layer thickness,
and M/A the mass per unit area.   Equation (2)  shows  that,
for  a given dust thickness,  the rapping intensity must be of
sufficient magnitude to produce an acceleration greater than

-------
the  ratio of  the  tensile  strength of  the  ash  layer  and  the
mass per unit area.  Also, Tassicker  shows that dust is
removed for a given acceleration only when
      M/A > P/a;                                            (3)
that is, when the mass per unit area  is greater than the ratio
of dust layer tensile strength to the normal  plate  accelera-
tion.  Since  the  mass per unit area depends on the  dust layer
thickness, which  in turn  is related to collection time between
raps, the time interval between the raps  is directly related
to the efficiency of dust removal from the plates.

Tassicker also indicates  that the maximum stress occurs at
the boundary  between the plate and the dust layer.  However,
in an operating precipitator, a thin  layer is often left when
a plate is rapped, indicating a stronger  tensile strength
between the plate and the dust than between particles in the
dust layer.

When a thin skin of unremoved dust develops,  100% removal
is not obtained.   Dust is removed only for a dust layer
thickness &i + HQ , where j£0 is the thickness  of the unremoved
layer and £j is the layer thickness needed to provide enough
stress to break the dust layer for a given acceleration.  Thus
the efficiency of dust removal is a function  of dust layer
thickness,  as shown in Figure 1.   In practice the sharp step
is not observed,  as shown in the data obtained by Sproull5
(Figure 2).   The dependence of dust removal efficiency on mass/
unit area (related to dust thickness)  was also observed by
Plato.11

Estimates of  the  tensile strength of  fly  ash were obtained
from Equation (2) and tabulated in Table  1 from Sproull's data.5

-------
   1.0
S
   0.8
v>
0  0.7
o
UJ
   0.6
o  0.5
0.4
o
u_
o
£  0.3
   0.2


   O.I


   0.0
          Lo
                    TOTAL DUST LAYER THICKNESS
        Figure  1.   Theoretical efficiency  of dust removal as a
                    function  of dust layer  thickness.
                              10

-------
    100


    90


    80


 sS  70
 UJ
 o
    60
 u.  50
 UJ
 s:  40
 a.
    30


    20


     10
                                I
I
I
               0.04       0.08       0.12       0.18       0.22

                         FLY ASH MASS PER UNIT AREA , g /cm*
                        0.26
                         0.30
Figure 2.   Rapping efficiency as  a  function of mass per unit  area for a
            precipitated fly ash at  150°C,  rapped with a maximum normal
            acceleration of 1.3 x  101*  cm/sec2 (13G) ;  measured  by Sproull.5

-------
                               Table  1

                       ESTIMATES AND  MEASUREMENTS
           TENSILE STRENGTHS OF PRECIPITATED  FLY ASH  LAYERS
Percent   Mass/Unit                   Tensile
Removal	Area	Acceleration    Strength	Source	

            g/cm2     cm/sec2   (G's)  dynes/cm2

  90         0.17   1.3 x 101*  (13)    2.2 x  103  Calculated  using eq.  2,
                                                 Sproull,5 Fly  ash 150°C,
                                                 Power on

  90         Q.03   3.1 x 101*  (32)   0.97 x  103  Calculated  using eq.  2,
                                                 Sproull,5 Fly  ash 150°C,
                                                 Power on
  89         0.04   6.3 x 10"  (64)    2.5 x  103  Measured with  precipi-
                                                 fuge, Tass i eker*

             0.21   4.2 x 103  (4.3)  0.89 x  103  Measured with  centrifuge,
                                                 Penney and Klingler1
             0.04   3.8 x 1014  (39)     1.7 x 103 Measured with  centrifuge,
                                                 Penney  and  Klingler13
             0.08   3.4 x 101*  (35)     2.7 x 103 Measured with  centrifuge,
                                                 Penney  and  Klingler13
             0.01   1.9 x 101*  (19.5)   2.2 x 103 Measured with  centrifuge,
                                                 Penney  and  Klingler13
             0.004  1.2 x 101*  (12.0)  0.53 x 103 Measured with  centrifuge,
                                                 Penney  and  Klingler13
             	    	             Ixl02-3xl03* Measured with  power
                                                    tensiometer,  Dalmon
                                                    and Tidy11*
*Bulk density 0.2 g/cm3 - 0.9 g/cm3
   no electric field, mechanically deposited.
                                    12

-------
Also included in the table are the results of laboratory mea-
surements by Tassicker,12 Penney and Klingler,13 and Dalmon
and Tidy.11*  These data indicate that the tensile strength of
precipitated fly ash layers range from O.SxlO3 dynes/cm2 to
3x103 dynes/cm2.  As will be seen later, the tensile strength
of the ash used during the FluiDyne experiments appears to also
fall in this range of values.
Quantity of Material Reentrained
The intent of this research project was not only to study the
mechanics of dust removal from the precipitator plates but also
to determine quantitatively the amount of material reentrained
under conditions representative of those found in full-scale
precipitators.  In an earlier study of reentrainment in a
full-scale precipitator, Sproull2 observed that emissions for
a single rapping puff increased with time between raps and con-
cluded that overall rapping emissions could be reduced by
lessening the rap intervals, although he recognized that the
interval should not be less than the minimum required for
dust removal.  It appears that he did not consider the cumula-
tive effect of more frequent rapping.  In addition, Sproull
concluded that the rapping intensity should be kept to a mini-
mum.  Sproull observed a vertical stratification of the emis-
sions during rapping with higher concentrations in the lower
portion of the precipitator.  Sproull*s results on the fre-
quency of rapping are somewhat contrary to the results obtained
in the laboratory by Plato,11 and by Sanayev and Reshidov,15
who observed a reduction in the fraction of collected dust
reentrained for larger accumulations on the plates.  By in-
creasing the time interval between raps, Plato,11 Schwartz
and Lieberstein,l6 and Nichols, Spencer, and McCain17 have all
observed improvements in performance of full-scale precipitators,
                              13

-------
The time interval between raps and the intensity of the rap
are the two major variables for the rapping study.  The lack
of quantitative measurements of all the parameters affecting
rapping reentrainment probably accounts for the differences in
the conclusions drawn from different investigations and for
the large range of approaches to rapping used by various manu-
facturers.

Although the previous studies have added to our understanding
of rapping reentrainment, they have not provided enough quanti-
tative data to aid in calculations of the effect of rapping
reentrainment losses on overall efficiency.

Penetration Including Rapping Reentrainment
Francis18 has developed an expression for the penetration dust
through a precipitator which includes losses due to rapping
reentrainment.

      PT  =  [R + (i-R)  (i-n)1/N]N                        (4)

where
      P   =  penetration corrected for reentrainment
       N  =  number of stages over which reentrainment is
             assumed to occur
       R  =  fraction of collected material reentrained per
             section
       n  =  fractional collection efficiency of a given particle
             size obtained with no reentrainment.
Francis assumed that a fixed fraction of the collected material
of a given particle size is reentrained, and that the fraction
does not vary with the length through the precipitator.  A
modification of Francis1 development gives the fractional con-
tribution of rapping reentrainment to emissions ( fR)  for a
given particle size to total emissions for a particle size as

      fR  =       R                                       (5)
       R    -^	 1)  + R
                              14

-------
 in terms of  the  sectional  efficiency  (n^)  for each particle
 size or as
              (R-i)  [i-(i-n

in terms of the total collection efficiency of a given particle
size with no reentrainment.  These equations indicate that the
percent contribution can be decreased by increasing the
sectionalization of the precipitator and by decreasing the
fraction of reentrained material (R) .  Quantitative measure-
ments of fR (fractional contribution rapping reentrainment to
emissions for a given particle size) and r\- (sectional effi-
ciency for a particle size) were obtained during the experi-
ments discussed in this report for a limited set of parameters.

Other Methods of Dust Removal
There are two methods to dust removal from precipitator plates
other than rapping.  One method is to wash the plates, as in
the so-called wet precipitator.  This type of equipment is
commercially available and is used in processes where rapping
has not been a satisfactory method of removing the dust from
the plates.  Wet precipitators are sometimes also used for
high resistivity dust.  A second method is to use scrapers.
However, the applications of scrapers are limited by expense
and by the difficult mechanical problems involved in operating
them in the harsh environments found inside industrial pre-
cipitators.

TEST PROGRAM
A test program was developed and carried out at FluiDyne to
meet the objectives of this rapping reentrainment study.  The
following test program was completed during two weeks of testing.
                              15

-------
The  first week consisted of setting up equipment and preliminary
tests and was followed by a second week of extensive testing
that generated a substantial quantity of data.

Test Program During First Week at FluiDyne
June 16      Unloaded and set up equipment.
June 17
June 18
June 19
June 20
Set up equipment; turned burners on; measured
clean plate rapping accelerations.

Began Testing.
Conditioned impactor blanks, tested for weight
gains, measured clean plate V-I characteristics,
measured gas velocity at sampling locations, ad-
justed for desired flow, measured gas velocities
at entrance and exit planes of the precipitator,
tested dust feed system, tested real time sampling
system.

Measured inlet particle size distribution, measured
inlet grain loading, tested load cells, determined
if rotating mass trains and impactors could detect
rapping puffs, determined amount of dust picked
up by rapping puff impactors and mass sampling
train when pointed 110° downstream from the flow
direction, measured dust resistivity, adjusted
moisture and temperature for 101 ° S7-cm resistivity
of fly ash.

Ran efficiency test with following conditions:
Dust feed
Current density - 23 nA/cm2
Gas velocity between plates - ^0.91 m/sec
Rapper intensity - 80% of maximum
Rapper interval - 30 minutes inlet (upstream rapper)
                - 60 minutes outlet (downstream rapper)

                 16

-------
Test Program Second Week at FTuiDyne
June 23      Start intensive test program.
             All variables except rapping intensity and time
             interval between raps were held constant.  The
             quantity of dust reentrained and the variables
             affecting reentrainment were measured.

             Test 1   150 minutes between raps, rapping
                      intensity 100% of maximum, 1 rap

             Test 2   120 minutes between raps, rapping
                      intensity 100% of maximum, 1 rap
June 24
June 25
June 26
June 27
Test 3   12 minutes between raps, rapping in-
         tensity 80% of maximum, 6 raps

Test 4   32 minutes between raps, rapping in-
         tensity 80% of maximum, 3 raps

Test 5   12 minutes between raps, rapping in-
         tensity 100% of maximum, 6 raps

Test 6   32 minutes between raps, rapping in-
         tensity 100% of maximum, 3 raps

Test 7   52 minutes between raps, rapping in-
         tensity 80% of maximum, 3 raps

Test 8   52 minutes between raps, rapping in-
         tensity 100% of maximum, 2 raps (rain
         and wind knocked out electrical power
         and burners, test was terminated 15
         minutes after second rap)

Test 9   Deleted due to internal electrical short
         in the precipitator.

                 17

-------
The following test procedure was utilized for each of the tests
during the second week of testing.

Before the start of each test, sampling probes and the pilot
precipitator were warmed up to the desired operating tempera-
ture.  The gas velocity at the outlet sampling ports was then
measured.  The precipitator hoppers were emptied and the dust
feed hoppers were filled.   The dust feed was then turn on,
checked with continuous monitors and allowed to stabilize.
The plates were then rapped.  Once the rapping puff was over
as indicated by the real time sampling systems, a test was
started.  At the end of the between-rap time interval, mass
trains and impactor probes were shut down and rotated out of
the gas flow.  The dust feed was then turned off.  When the
outlet real time system could no longer detect particulate,
a separate set of mass sampling probes and impactors were
rotated into the gas stream to measure emissions from rapping.
The plates were then rapped, first with an upstream rapper
and then with a downstream rapper.  After the rapping puffs
had passed the rapping puff mass sampling trains, probes and
impactor probes were rotated out of the gas stream.  The dust
feed was then turned on again and sampling with the between-
rap systems was resumed.  A semi-traverse of one outlet lane
with the mass train system was obtained for the rapping puffs
by sampling at three separate vertical locations with two probes,
The rapping tests were set up so that for each test condition,
rapping puffs were sampled in multiples of three, except for
the long interval tests.

The experimental arrangement, test conditions, results of the
experiments and analysis of the data are discussed in the
following sections.
                              18

-------
                        SECTION III
                  EXPERIMENTAL ARRANGEMENT

This study of rapping reentrainment was conducted in a nearly
full-scale pilot precipitator owned and operated by FluiDyne
Engineering.  The precipitator is equipped with extensive
instrumentation.  The FluiDyne Engineering facility, the test
apparatus, and the operating procedures used during this
investigation are discussed in this section.

FLUIDYNE ENGINEERING TEST FACILITY
The test facility at FluiDyne is illustrated in Figures 3 and
4.  The test facility was selected for this study because it
closely represented operation of a full-scale unit while at
the same time providing a facility for studying a variety of
rapping variables.  The ability to turn the dust feed on and
off during the tests simplified the measurements of emissions
due to rapping reentrainment.  A detailed description of the
facility follows.

Precipitator Design
The design parameters for the precipitator are tabulated in
Table 2.

This pilot-scale precipitator effectively represents one
electrical section in a full-scale precipitator.  The six
meter plate height is high enough that the reentrainment loss
should be approximately representative of those in a full-scale
unit.  The precipitator was originally constructed to study
operation of a full-scale unit.

In the original design, the plates in the precipitator were
constructed from expanded metal.   For this rapping reentrainment
                             19

-------
                        NOMINAL 48 IN. WIDTH
                        FOR (5) PASSAGES
          EXPANSION TO
          TEST PRECIPITATOR
                                            STEAM
                                            TO MATCH
                                           CONDITIONS
.FLOW RATE TO
 45,000 SCFM
 TEMP-AMBIENT- 4OO°F
    HUMIDITY TO 8 %
             ASH FROM                 H2S04 + HOT AIR (650°F)
             CLIENT'S STATION
             (FROM PRECIP. HOPPERS)
TO MATCH CLIENTS CONDITIONS
Figure  3.   Near  full scale pilot precipitator at FluiDyne Engineering.

-------
       TYPICAL FLOW
       9 FT/SEC
       300°F
       350OOACFM
       SCA 51.4
* 9 ACCELEROMETERS
 ARE MOUNTED ON PLATE 4
                                                          NORTH
                                                             EAST
                                                                       SECTION A-A
                                                                                     WEST
                                                                                      CHANNEL NO

                                                                                      PLATE ROW NUMBER
PLATE ROWS 1, 2, & 6 EXPANDED METAL
PLATES ROWS 3, 4, & 5 SOLID
PLATES 2-1 and 2-3 WERE
SHORTENED .5 METERS
Figure 4.    Various views of FluiDyne  pilot  precipitator.

-------
                           Table 2
        DESIGN PARAMETERS, FLUIDYNE PILOT PRECIPITATOR
                Design Parameters
  Units
Number of electrical sections in direction
  of gas flow
Length of electrical sections
Plate area of each electrical section
Total length of corona wire per electrical
  section
Corona wire
Wire-to-plate spacing
Number of wires per linear section
One-half wire-to-wire spacing
   2.7 m
   167 m2
   457 m
   0.48 cm
(square wire)
     11 cm
     75
    7.6 cm
                               22

-------
study three of these plates were replaced to provide two lanes
with solid plates on each side of the lane.  Outlet sampling
was confined to the lanes with solid plates.

Plate Construction -
There are 18 (0.9 m x 6 m) plates in the precipitator.   The
six rows of plates are numbered from one to six as viewed
from the outlet looking towards the inlet and numbering from
left to right.   Rows 1, 2, and 6 have expanded metal plates,
and rows 3, 4,  and 5 have solid plates.  Plates in each row
are numbered from 1 to 3 from the inlet to the outlet.   Lanes
are number 1',  2', 3', 4", and 5' from left to right as viewed
looking from the outlet towards the inlet.

The solid plates are constructed from 18 ga. (1.2 mm)  hot
rolled steel with 4.4 cm x 3.2 cm x 0.32 cm angle iron on
the sides.  Design details are shown in Figure 5.  No addi-
tional internal baffling or supports were provided.  Plates
2-1 and 2-3 were shortened 0.5 meters to permit viewing of
dust on the adjacent solid plates.

The plates are suspended from four I-beams by 1.3 cm threaded
rods welded to the plates as shown in the Section A-A view
in Figure 4.  The four I-beams are supported by eight rods
(4.4 cm diameter)  which are floated from eight springs
(McMaster Carr 9622K25).

Plate Rappers -
There are two plate rappers; one raps the upstream section of
the plates (I-beam support between first and second section of
plates)  and the other the downstream section (I-beam support
                              23

-------
                   1/2 IN. DIA RODTYP (2)
             18 GA.
             (0.0478) IN.
  1/2 IN. DIATHDROD
/TYP(2)
 20 FT.
  1-3/4-IN.-
                                                   ANGLE IRON
                                                   1-3/4x1-1/4x1/8 IN.
                                                   TYP(4)
Figure  5.   Design details for pilot  precipitator  plates,
                             24

-------
between the second and third section of plates—see Figure 4).
A third rapper is used to rap the high voltage section; this
rapper is insulated electrically by a glass rod.

The plate rappers are the single-shot type of pneumatic rapper.
The rapper weight, a cold rolled steel rod, 12.7 cm long and
7.46 cm in diameter, is housed in an air cylinder.  Between
raps, the rapper weight is supported in the cylinder by low-
pressure  [L = l.OxlO5 Pa (M5 psi gauge)] compressed air.  When
a rap is desired, a signal to a solenoid valve pressurizes
the other side of the cylinder and forces the weight down on
top of a rod that transmits the force to a plate support beam.
The compressed air used was under pressure up to 7.6 x 105 Pa
(110 psi gauge)  and the cylinder stroke was approximately
12.7 cm in length.

Hopper and Baffle Design -
The hopper and the baffles of the FluiDyne pilot precipitator
are shown in Figure 4.  The baffles that extend down from
between the sections of plates come to within about 0.4 meters
of the bottom surface of the hopper.  There are two such
baffles, one between the first and second section of plates,
and one between the second and third section of plates.  There
is also a central baffle that extends up 0.8 meters from the
floor of the hopper directly underneath the second set of plates,

Downstream of the third section of plates a baffle extends
down from the top of the precipitator box to about 0.76 meters
below the top of the plates.   Similarly, a baffle extends up
from the bottom to the level of the inlet ducting floor.  At
this same streamwise location lateral baffles block the
space between the precipitator box walls and the outside plates.
                              25

-------
Precipitator Power Supply -
The FluiDyne test precipitator is powered by a General Electric
transformer rectifier and control.  The transformer is a 480
volt primary, 93 kVA transformer with a secondary voltage of
77,200 volts.  The rectifier is mounted outside the building
in the transformer  yard and the high voltage is transmitted
to the top of the precipitator through a Belden 8871 coaxial
cable.  This cable runs approximately 200 feet in a 3.8 cm
conduit up to the top of the precipitator.  The precipitator
is currently energized with full wave rectification, although
the transformer rectifier is set up to provide double half
wave operation if desired.  The precipitator power supply con-
trol panel is mounted inside the building near the transformer
rectifier.  The instrumentation on the front panel of this
control includes primary current and voltage and secondary
high voltage and secondary direct current for both halves of
the double half wave.

Other Components of the FluiDyne Facility
In addition to a nearly full-scale precipitator, the FluiDyne
facility includes the systems necessary for producing simulated
flue gas conditions at the inlet of the pilot precipitator.

Feed Hopper Design and Dust Dispersion System -
The dust dispersion system consists of a feed hopper and a
primary blower which provides a forced draft to the precipi-
tator.  The feed hopper is a bin made out of sheet metal
in the form of a cone with a cylindrical extension on top of
it.   The 1.8 meter high feed hopper is filled by pneumatically
transferring fly ash through a 5 cm pipe from fly ash storage
tanks which are located outside of the building wall.   The
hopper holds about 680 kg of fly ash and has a 8.9 cm outlet
                             26

-------
at the bottom of the cone.  A dry chemical screw feeder has
been adapted to the bottom outlet of the large feed hopper.
This screw feeder consists of a flexible vinyl trough having
a 10 cm diameter feed helix at the bottom which is rotated
by a variable speed motor which can be set from 0 to 80 rpm.
Dust flow rates up to 362 kg (800 pounds) per hour can be
maintained.  The dust is kept flowing through the 8.9 cm
outlet into the small hopper of the dry chemical screw
feeder by vibrating the large feed hopper with a pneumatically
powered vibrator.  The dust that is fed out of the screw feeder
falls into a funnel which is connected to a 7.6 cm line
running over to the inlet plenum of the primary blower that
supplies the gas for the entire precipitator test system.
The blower acts to mix the dust well into the air flow.  The
weight of the entire dust system is measured by a FluiDyne
Flex-Cell.  The load cell is powered by a low voltage direct
current power supply and the signal is amplified and displayed
on a pair of pen recorders.  One of these recorders is
calibrated to display the total capacity in pounds of the feed
hopper system.   The other recorder has a much higher sensitivity
and also an adjustable bias so that only a small portion of
the hopper load is displayed on the range of the chart.  The
slope of the load trace on this second recorder is used in
setting the dust flow rate out of the hopper system.

Oil Burner System -
Three oil burners, each of which is rated at approximately
2.9 mw are used to heat the test gas stream to the desired
temperature level.  These burners are mounted in a large tank
in the gas flow path downstream of the primary blower and fired
horizontally (see Figure 3).  Any or all of these oil burners
can be fired simultaneously.  Temperature control is achieved
by adjusting individual burner outputs and/or changing the
number of energized burners.  During most of the test runs
only two burners were used.
                              27

-------
Number 2 fuel oil is used in these burners.

Water Injection System -
The gas flow through the precipitator test system is conditioned
by injecting and vaporizing a known amount of water into the
gas flow.  This is accomplished utilizing three atomization
nozzles, each of which has a capacity of about eleven liters
of water per minute.  The water is atomized using compressed
air and is vaporized by the same burners that heat the system
gas flow up to the design temperature.  The amount of water
injected in this manner is measured by a pair of flow meters
in the water line, and the amount of water is determined by
taking into consideration the over-all volumetric content of
water desired, the atmospheric humidity, and the amount of
water in the gas flow due to the combustion of oil.  Typically,
the amount of moisture due to atmospheric humidity runs about
1% or as much as 2% during very humid days.  The amount of
water from the combustion of oil can amount to about 0.5 to
1% and the rest of the water up to the amount desired is
then added through the water vaporization nozzles.

TEST APPARATUS AND OPERATING PROCEDURES
The instrumentation for this study of rapping reentrainment
characteristics in the FluiDyne Engineering pilot precipitator
provided quantitative data about the parameters that affect
rapping reentrainment and the effect of rapping reentrainment
on overall precipitator operation.

A block diagram (Figure 6)  of the experimental layout for
the rapping reentrainment study shows the instrumentation
that was put together for this study.  In Figure 7, the sampling
locations for the real-time particle measurement system, inlet
and outlet impactors,  outlet mass trains, and transmissometer
and resistivity probe are indicated.

                              28

-------
                                  ELECTRICAL
                                  CHARACTERISTICS
         RAPPING
         VARIABLES
DUST LOAD
ON PLATES
PLATE ACCELERA-
TION
           GAS ANALYSIS
           SO3, SO2, H2O
           TEMPERATURE
N)
          RESISTIVITY
          MEASUREMENT
                           MASS LOADING
TIME INTEGRATED
PARTICLE SIZE
MEASUREMENTS
VELOCITY
DISTRIBUTION
                                                               PRECIPITATOR
           * TWO SETS AT 3 LOCATIONS: ONE TO MEASURE DURING
            RAPS AND ONE TO MEASURE BETWEEN RAPS.

          *» TWO SEPARATE UNITS: ONE TO LOOK AT LOWER HALF OF THE
            PRECIPITATOR OUTLET AND ONE TO LOOK AT UPPER HALF OF
            THE PRECIPITATOR OUTLET.
                                                                              OBSCURATION
                                                                              METER
                                                                                TIME INTEGRATED
                                                                                PARTICLE SIZE
                                                                                MEASUREMENTS
                        MASS LOADING
                        UPPER HALF
                                                                              CAMERA AND
                                                                              LIGHTING
                                                    MASS LOADING
                                                    LOWER HALF
HOPPER
SAMPLES

** REALTIME
PARTICLE SIZE
MEASUREMENTS
                      Figure  6.   Block  diagram of experimental  layout for rapping
                                    reentrainment study.

-------
OJ
o
                                            o
                                          n<9
                                         O
o
o
o
o
                                                          7.5m
                                                                             O ROYCO
                                                                             O MASS TRAIN
                                                                             D IMPACTOR
                                                                             ^ TRANSMISSOMETER
                                                                            00
                                                                            o
COLLECTING
SURFACE OF
               Figure  7.   Sampling  locations during the rapping reerttrainment study.

-------
The gas analysis apparatus, IKOR mass monitor, and inlet mass
sampling trains were located in the inlet ducting at point A
in Figure  3.  The Konitest mass monitor was located 6 feet
downstream of the precipitator in the rectangular outlet duct.
Motion pictures of the rapping puffs leaving the precipitator
were taken from points B, C, and D in Figure 7.

Apparatus  and Operating Procedures
Mass Trains -
Three .separate mass trains were used to measure inlet and out-
let particulate concentrations during the rapping reentrain-
ment test  at FluiDyne Engineering.  The inlet mass train con-
sisted of  an SF-Carborundum filter with a 0.95-cm nozzle on a
12-meter probe, a stainless steel condenser, a calcium sulfate
drying tube, and a Research Appliance Corporation metering
box.  The  gas velocity at the sampling point averaged approxi-
mately 8.35 m/sec in a 1.1 m  diameter duct.  A 16-point
sampling traverse was made along two paths 90° to each other.

Two outlet sampling trains were used:   (1)  the upper sampling
train for the upper 68% of the precipitator outlet,  and (2)  the
lower sampling train for the lower 32% of the precipitator.
As shown in Figure 7, the outlet sampling locations were
approximately 1 meter from the plane of the outlet baffles.
Only one lane of the precipitator was sampled.  The outlet
mass trains were modified so that each train consisted of
two systems.  One system was used to measure emissions between
raps and the other system was used to measure emissions during
the raps.  Each outlet sampling probe consisted of a 2.5 cm
pipe, to the end of which two 47 mm Gelman filters with 1.25 cm
nozzles pointed 110° apart were attached.  Separate 0.95 cm
copper tubes were run to each filter from a three-way valve.
The valve was used to connect the appropriate filter to the
                              31

-------
condenser and modified metering box.  The metering boxes at
the outlet were modified to accommodate the low sampling rates
at the outlet.  The velocities at the sampling locations were
only on the order of 1 m/sec to 0.15 m/sec.  The standard
metering orifices were changed to a 2.1 mm orifice to give
reasonable pressure drops.

The operating procedure for the outlet mass trains was as
described below.   Before the start of each test a velocity
traverse of the five upper sampling points and the four lower
sampling points was made.   The velocities obtained at these
points were used for setting the sampling rates.   The velocity
at the inlet was periodically monitored during the day to insure
that the same volume flow rate was maintained throughout the test.

One of the two filters on each of the two outlet probes was
designated the between-rap sampler and the other the rapping-
puff sampler.  The between-rap systems were started after
stable conditions were obtained, as indicated by the real-time
particle size system, the transmissometer,  and the Konitest
meter.  Before rapping the plates, sampling was discontinued and
the probes were rotated so that both nozzles on each probe
pointed downstream.  The dust feed was then turned off.  After
a clear flue was obtained, the second filter was rotated into
the gas stream.  Sampling was resumed and the plates were
rapped.  When a clean stack was again obtained, sampling with
this second set of filters was discontinued and the nozzles to the
filters were again pointed downstream.  The dust  feed was  then
turned on and sampling was resumed again with the between-rap
system.

A test was made during the early stage of the test program to
insure that the filter with  the  nozzle pointing downstream
did not pick up dust.  One mass train was pointed downstream
                              32

-------
 during an  entire  day  of  testing without  sampling  any  gas.
 The weight gain of the filter and nozzle washing  was  insig-
 nificant compared to  the weight gains of approximately 0.1 g
 bv the filters that sampled dust.

 The  data obtained between raps  were handled in the usual
 manner.  These  data gave the  particulate mass loading  between
 raps and, combined with the volume flow rate, gave the mass
 emission rates on an emission per hour basis.  The data from
 the second set of filters, or rap filters, was used to calcu-
 late hourly mass emission rates from the rapping puffs, in-
 dependent of the between-rap emissions.  These emission rates
were calculated using the following equation:
       Ep
              ANNS
where
       E   =   emission  rate  from rapping puffs
       M   =   mass  collected by the  filter while sampling the
       P      flue  gas  during rapping
       A   =   cross-sectional area of the precipitator sampled
       s      by the probe
       A^  =   cross-sectional area of the probe nozzle
       Nu  =   number of raps per hour
       rl
       N   =   number of raps sampled
       S

The emission  rates between  raps and from raps were combined
to obtain the overall  hourly emission rate.  The effective
particulate grain  loading,  including the rapping loss, was
calculated from the overall emission rate and the average
volume flow rate through the precipitator.
                              33

-------
 Impactors -
 The  impactors used  for particle-size distribution measurements
 were modified Brink Cascade Impactors  for the inlet and
 Andersen Stack  Samplers  for the outlet.

 At the inlet, the runs consisted of traverses and single-point
 samples.  The run times were  30 minutes each  at a flow  rate
 of about 1.9x10  5m3/sec  (0.04 acfm).  The modified  Brink im-
 pactor consisted of  seven stages with a precollector cyclone
 and  a back-up filter.  Its  sizing range was approximately 0.5 ym
 to about 10 ym particle diameter.
 At the outlet, four  Andersen  impactors sampled from two ports;
 each port contained  two impactors.  One impactor per port
 sampled between  raps while  the  other sampled  the rapping puffs.
 Before the discontinuance of  the dust feed and the  start of
 rapping, the  two between-rap  impactors were shut off and
 rotated downstream;  the two rapping puff  impactors  were rotated
 upstream and  sampled during the rapping sequence.   After the
 rapping puffs ceased, these impactors were shut off and turned
 downstream; when the dust feed  was  resumed, the first two
 impactors were turned upstream, and sampling  was continued.
 This procedure was repeated throughout each test.   The  Ander-
 sen  impactors consisted of  eight stages and a back-up filter
 and  sampled at about 1.4x10 1*m3/sec  (0.3  acfm).  They also had
 a sizing range of approximately 0.5 ym to about 10  ym particle
 diameter.
Real-Time Particle Size System -
The  contribution of  reentrained particulate to the  flue gas,
in the form of rapping puffs, was measured in real  time by
two  modified Royco Model 225 Particle Counters.  The particle
counters utilize five-channel analog ratemeter modules, which
provide parallel monitoring of the instantaneous concentration
of particles in five pre-selected diameter intervals.   The
diameter interval thresholds in this case were 1.5  ym,  3 ym,
6 ym, 12 ym,  and 24 ym optical diameters.

                              34

-------
The analog ratemeters, designed by Southern Research Institute
for this application, have approximately a half-second response
time.  This resolution is good enough to show concentration
changes throughout a rapping puff.  The information from the
analog ratemeters'  five channels of output was recorded on
a six-channel chart recorder.  The sixth channel was shared
alternately by a PILLS instrument and a transmissometer.  Two
eventmarkers on the recorders were used to individually indicate
when the upstream and downstream rappers were energized.

The temperature and aerosol concentration limits of the
particle counters require that the flue gas be diluted before
measurement.  Because of the very steep gradient in the number
size distribution,  the diluter was constructed as a size
segregating device to dilute the smaller particles more ex-
tensively than the larger particles.  The diluter is shown
schematically in Figure 8.

The sampling probes were configured and installed in such a
manner as to permit a vertical traverse to be made along the
center line of one lane at the exit plane of the precipitator.
Probe losses were minimized by installing the particle
sampling train (probe, diluter, particle counter) underneath
the outlet duct and extracting the aerosol sample through a
vertical probe with a single 90° bend between the sampling
point and the particle sensor.  The probe and nozzle were con-
structed from a continuous length of 4 mm I.D. stainless steel
tubing.  As used in these tests, the probe flow was laminar
with a Reynolds number of 100.  A block diagram  of the  sampling
train  is illustrated  in Figure  9.

Although the large particle sampling train had been calibrated,
the data on the particulate transfer efficiency of the nozzle
were badly scattered.  Also, at FluiDyne a problem arose with
water condensation in the sample probe tubing inside the

                              35

-------
     SAMPLE
     FLOWRATE
     MANOMETER
  PROBE
_ HEATER
 VERTICAL
                        VIEW
                        VOLUME
          OPTICAL
          HEAD
            ELECTRICAL
            LEADS, ETC.
                                          GAS  FLOW
"PROCESS EXHAUST
 LINE
                                  DILUTION AIR
                                  MANOMETER
                          0.02um
                          F|LTER
         BLEED
         VALVE
                                                              PUMP
I// / J


B
V
                                                            BLEED
                                                            VALVE
                 CONDENSOR
                 POT
            LARGE  PARTICLE
            COUNTER
            MAIN FRAME
Figure  8.   Schematic  of diluter  for  large particle sizing system.
                               36

-------
PROBE


ROYCO
OPTICAL
HEAD


ROYCO SIGNAL
CONDITIONER


SRI RATE
METERS
(5 channels)
                                                         BRUSH
                                                       OSCILLOGRAPH
                                                         RECORDER
                                                         (6 channels)
Figure  9.   Block diagram of large particle sizing system.
                             37

-------
diluter  (which was unheated and in the path of the dilution air)
Increasing the flow rate and adding insulation in appropriate
places eliminated the condensation, but because of the required
high flow rate to eliminate the condensation, isokinetic
sampling was not possible.  The nonisokinetic sampling and
uncertain particle losses in the nozzles made it impossible to
reliably determine the particulate concentrations on an
absolute basis.  The nozzles and sampling rates in both setups
were identical, however, allowing meaningful relative compari-
sons of the data obtained with the two trains.

The two large-particle sampling trains had different probe
lengths  (2.1 meters and 5.2 meters), because the sampling
location allowed only a 76 cm traverse.  The 2.1m train
sampled the lower 0.76 m of the precipitator and the 5.2 m
train sampled a region in the upper third of the precipitator
exit.  These sampling locations are shown in Figure 7.  Vertical
traverses were made with each large particle train while the
dust feed was on.  Traverses of the puffs were approximated
for each rapping interval by positioning the sampling probes
at different points after the dust feed was turned off and
before the inlet-outlet rapping combination was started.

Load Cells -
Beam type load cells were used to monitor collection plate
weights of plates 4-1,  4-2,  and 4-3.   The cells (Figure 10)
are mounted on the plate support beam and the plate support
bolt is attached to the cell beam instead of directly to the
support beam.   Two cells support each plate (one on each
support bolt).   Stops are provided to prevent excessive de-
flection during rapping.

The load cells  were read out on a FluiDyne-Vidar System.  The
system has six channels of signal conditioning and the Vidar
                              38

-------
                                        10-1/2 IN.
Co
                                        STRAIN
                                        GAGES
                                  PLATE SUPPORT BEAM
                                                          i
                                                          I
                                                          I
                                                          I
                                                         . J
	i
                                                                       fl  fTT3
                                           PLATE SUPPORT BOLT

                                           LOAD CELL
                                                                   tr
             ITT
                                    Figure  10.   Load cell.

-------
and its associated printer were capable of reading and printing
at a rate of approximately 4 channels per second.  A block
diagram of the complete measurement system is shown in Figure 11.

Accelerometers -
Plate accelerations were measured during the test with piezo-
electric accelerometers mounted on plates 5-1,  5-2, and 5-3.
Three accelerometers were mounted along the centerline of
each plate at locations 0.91 m, 3.05 m, and 5.17 m from the
top of the plates.  The accelerometers were Gulton Industries
Model AHT-40TM,  which have a frequency range from 3 Hz to
9 kHz and a dynamic range from 0.1 g to 250 g.   The weight of
each unit, exclusive of cable, is 52 grams.

The nine accelerometers were mounted to the plates with four
screws and the cables were run up the plates in 1.3 cm conduit
which was strapped all the way up the centerline of the plates.
The accelerometers were mounted so that the Z-channel connector
was perpendicular to the plate surface and the X and Y-connectors
were on the two side surfaces of the block.  Figure 12 shows
the accelerometer system block diagram.

It should be noted that the recording oscillograph limited
the system frequency range to 2 kHz and below.

Camera and Lighting -
Motion pictures were taken of the emissions during rapping;
the camera used was a Bell & Howell Model 70-DL.  This is a
16mm camera with a spring drive.   The film used for black and
white pictures was Kodak Tri-X reversal film #7278, and the
film used for color motion pictures was Kodak Ectrachrome
type #7242.   The lighting was furnished by movie lights of
                              40

-------
 LOAD CELL
 FLUIDYNE
 BEAM
 ELEMENT
6 CHANNEL
                   POWER
                   SUPPLY
                   10V DC
SIGNAL
CONDITIONER
FLUIDYNE
BRIDGE
BALANCE
6 CHANNEL
SCANNER
VIDAR 604
                                     10 CHANNEL
A/D
CONVERTER
VIDAR 260
DIGITAL
COUNTER
B EC KM AN
6230
DIGITAL
PRINTER
B EC KM AN
1455
                                                                                             5 LINES/SEC.
                 Figure  11.
             Precipitator plate weight measurements system
             block  diagram.

-------
Piezoelectric
Accelerometer
GultonAHT-40TM


Charge
Amplifier
Columbia 4102R


Galvanometer
Driver Amplifier
CEC 1-162A
"— —

Recording
Oscillograph
CEC5-124
27 Channels
9 Channels
14 Channels
10 Channels
          Figure  12.   Accelerometer system block diagram.
                              42

-------
250 watt and 650 watt intensities.  These lights were aimed
at the area of interest from portholes adjacent to that through
which the pictures were made.

Transmissometer -
A Lear-Siegler RM41p portable optical transmissometer was
used to monitor the particulate emissions at one point at the
outlet of the FluiDyne pilot precipitator.  A similar on-stack
transmissometer has been described by Beutner.19  The portable
unit that was used at FluiDyne has an optical path length of
two meters and compensation circuitry for determining opacity
in terms of the stack exit diameter.  For simplicity, all of
the opacity and optical density measurements in this report
are given in terms of a two-meter optical path length.

Konitest -
A vortex type Konitest with a Keithley Multi-Range Model 610-CK
electrometer was used to continuously monitor particulate
loading in the rectangular exit duct 3 m downstream of the
precipitator box.  The sample probe nozzle was located near
the duct vertical centerline, about 2 m up from the duct
floor.  The Konitest uses electrical charge generation by the
particle-surface contact as a measure of particulate concen-
tration.   The current-versus-particulate mass concentration
is usually linear20 for constant test conditions.

IKOR -
An IKOR Model 206 portable particulate monitor was used to
monitor inlet grain loadings.  In this system, sampled particles
are continuously drawn through an electronic sensing head,
generating an electric current by charge transfer.  According
to the manufacturer, the current output is precisely related
to mass flow.   Some problems were encountered with dust buildup
in the sampling line.

                               43

-------
 Resistivity  Probe  -
 A point-plane  in situ  resistivity probe was  used  to measure
 the  electrical resistivity of dust  entering  the FluiDyne
 pilot  precipitator.  The operation  and description of  this
 probe  are  given by Nichols and  Spencer.21
 Anemometer  -
 A Thermo-Systems model  1610 velocity transducer with a special
 high  temperature ceramic probe was  used to make velocity
 measurements  in the pilot precipitator.  The two-needle hot
 film  sensor .015 mm  (6  mil) with  special epoxy resin was
 designed for  temperatures up  to 162°C  (325°F).  A  special  7.6
 meter,  high temperature cable connecting the sensor to the
 system  electronics allows vertical  scanning of the entire
 precipitator  test section.  This  temperature compensated anemo-
 meter was factory calibrated  for  the velocity range 0 to 4.6 m/sec
 (0.3-3.0 volts output).   The manufacturers  estimated accuracy
is 2%.
 Gas Analysis  -
 The gas entering the pilot precipitator was analyzed to de-
 termine the concentrations of 02 , H20, S02, and SO3.  The
 concentrations of SO2 and S03 were  expected to be  low.

 The apparatus for determination of  S02 and SO3 consisted of
 the following items in  sequence:  a heated pyrex probe, a
 heated  (60-90°C) glass  spiral condenser having a fritted glass
 plate at the outlet, a  250-ml gas washing bottle (coarse
 fritted glass bubbler)  containing dilute hydrogen  peroxide,
 a drying tube containing Drierite,  a dry gas meter, and a
 vacuum pump.

A measured volume of gas was  drawn  through the system; SO3
was captured in the condenser and S02 was oxidized to SO3  in

                              44

-------
the washing bottle.  The resulting sulfate solutions were
titrated with 0.005 N barium perchlorate, with thorin
indicator.

The C02 and Oz concentrations were determined with Fyrite
gas absorbing solutions.  The moisture contents were determined
by passing a known volume of gas through a column of Drierite
and measuring its weight gain.
                              45

-------
                         SECTION  IV
                      TEST CONDITIONS

During this study of rapping reentrainment the  following
variables were controlled:
  1.  Dust properties
        resistivity
        adhesion and cohesion
        particle size distribution
        chemical composition
        bulk density of  collected ash layer
  2.  Gas composition
        H 20 concentration
        SO 3 concentration
        S02 concentration
  3.  Corona current density
  4.  Gas temperature
  5.  Gas velocity
  6.  Dust loading
  7.  Plate construction and suspension system
  8.  Rapping variables
        Type
          impulse
          vibration
        Location and orientation
        Intensity
        Time interval between raps (amount of dust collected
          on a plate prior to rapping)
On the basis of previous rapping studies discussed in the in-
troduction, these variables were considered to be the ones
that would affect the removal of the dust from the plates
and the amount of emissions due to plate rapping-

                             46

-------
In this section, the results of the measurements of most of
these variables are reported.  The dust loading on the pre-
cipitator plates is discussed in a later section.  The
plate construction and suspension system were discussed in the
previous section.

DUST PROPERTIES
Fly ash was obtained from the Black Dog Power Plant in
Minneapolis, Minnesota for this study.  The ash was selected
to give a representative particle size distribution of dust
entering a precipitator and to have typical properties.
Studies with a dust with a size distribution representing
dust entering an outlet section were also contemplated.  The
difficulties in obtaining such an ash at the test location
and the time limits for the investigation of rapping reen-
trainment at the test location prevented such a study.

The results of a Bahco analysis of the fly ash used in the
study are shown in Figure 13.  The mass median diameter
of the sample was 16um with a geometric standard deviation of
5 and a mode of 4ym.  This distribution is typical of fly ash
from pulverized coal-fired boilers.

The resistivity of the fly ash measured with the in situ
probe is tabulated in Table 3.   Dust resistivity varied over
a larger range than desired during the test, due to changes
in gas temperature and moisture content.  However, by operat-
ing the precipitators at a reduced spark rate, less than one
per minute, the effects of these changes on precipitator
operation were minimized.   Laboratory measurements of the
fly ash resistivity are shown in Figure 14.

The true density of the fly ash was determined with a helium
pycnometer to be 2.65 g/cm3.   The porosity of a dust layer
                             47

-------
          100
       E
       o

       I-
       cc
           10
          1.0
Ill    I
                                              I
            0.1
              10    20     40     60     80



          % LESS THAN INDICATED SIZE
Figure  13.   Cumulative particle size distribution  of the feed

             dust,  Bahco analysis.
                             48

-------
             TABLE 3
     IN SITU RESISTIVITY DATA
                 Electric Field
     Resistivity at
Resistivity  from
Date
6/23/75
6/23/75
6/24/75
6/24/75
6/25/75
6/25/75
6/26/75
6/26/75
6/26/75
Time
10:45
15:25
12:38
15:52
15:25
16:20
13:35
15:15
16:30
Location % H2O
Inlet 9.3
Inlet 10.6
Inlet
Inlet 10.6
Inlet
Inlet 12.3
Inlet
Inlet 13.0
Inlet
Temp °c
130
150
124
126
104
104
121
116
121
at Sparkover
2000
3300
2200
1700
1000

700
1300

V
V
V
V
V

V
V

0
1
6
1
2

1
3

Sparkover
. 70x10 1 l
.02x10"
. 16x10 10
.42X101 l
. 64x10 10

.47xlOl°
. 15x10 10

£2-cm
fi-cm
£2- cm
JJ-cm
n-cm

fj-cm
f2-cm

V-I Data
1.
1.
1.
3.
5.
5.
5.
7.
9.
02x10 l l
38x10 ll
48x10 J '
17x10 l l
41x10 10
09x10 10
OOxlO10
50x10 10
00x10 lo
£2- cm
(2-cm
£2-cm
fi-cm
ft-cm
fl-cm
fi-cm
fi-cm
Si-cm
. Avg.
       122°C
Avg. 6. 40x10 10 S2-cm

-------
E
.c
o
>

>
LU
cc
    10
     12
    10
      11
    10
     10
     10-
    10e
    10'
      DATA TAKEN AT 8% H2O

      O FEED HOPPER SAMPLE
      & OUTLET PLATE SAMPLE
           3.2
           39
           102
2.8
84
183
2.4
144
291
2.0
227
440
1.6
352
666
1.2
560
1040
1000/T(°K)
°C
°F
                                 TEMPERATURE
Figure  14.   Resistivities  measured  in Southern  Research Institute
              laboratory for fly  ash  used during  the test.
                                 50

-------
obtained in a Power Test Code 28 resistivity cell was 56.3%
for a sample from the feed hopper and 49.4% for a sample from
an outlet plate of the pilot precipitator.  These values are
used for estimating dust layer thicknesses from the load cell
measurements.

Instrumentation for determining the dust cohesion characteris-
tics during the test was not available.

A chemical analysis of the fly ash used in this series of
experiments is given in Table 4.  The composition of the fly
ash appeared to be typical.

GAS COMPOSITION
The compositions of the flue gas samples are shown in Table 5.
These measurements, along with the finding that there was no
SO3 and only 16 ppm S02 present, indicated that the gas was
mostly air with a small contribution of C02 from the oil-fired
burner that was used to heat the gas.  Previous research had
indicated that the injection of S03 can increase dust cohesion
characteristics and decrease rapping reentrainment emissions.
An investigation of this effect was considered for this study.
However, to fit the test into the available time, a study of
the effect was omitted.

ELECTRICAL CONDITIONS AND GAS TEMPERATURE
Operational current densities and voltages significantly affect
the collection efficiency of an electrostatic precipitator.
A decrease in current density can result in an increase in
emissions due to plate rapping by decreasing the recollection
of the reentrained particulate.  A decrease in current density
can also lower the tensile strength of the collected particu-
late layer.  The maximum value of current density for any
specific precipitator is set by the resistivity of ash that

                             51

-------
            TABLE 4
CHEMICAL ANALYSIS OF BLACK DOG
  FLY ASH, PERCENT BY WEIGHT
Li20
Na20
K20
MgO
CaO
Fe203
A1203
Si02
Ti02
P205
SO 3
Loss on ignition

Inlet Plate
0.03
1.75
1.23
3.08
12.04
8.94
20.71
46.93
1.04
0.26
1.78
3.40
Sample

Feed
Outlet Plate Hopper
0.04
1.60
1.50
3.69
13.33
7.58
19.33
44.27
1.11
0.26
2.74
4.84
0.03
1.74
1.19
3.08
13.69
11.51
18.64
44.27
1.04
0.25
1.98
3.75
                52

-------
 Date

6/23/75

6/23/75

6/24/75

6/25/75

6/26/75

6/27/75
                      TABLE 5.

            COMPOSITION OF GAS ENTERING
          THE FLUIDYNE PILOT PRECIPITATOR
 Time

11:10

 4:35

 4:12

 4:00

 3:50

 8:40
   CO 2
% by vol.
   3

   3

   4

   3
   02
% by vol.
   20

   20.5

   21

   20
 H20
by vol

 9.3

10.6

10.6

12.3

13.0

10.7
                         53

-------
is collected.  In turn, the resistivity of any given ash is
basically controlled by gas temperature and partly controlled
by the flue gas moisture content.  A continuous log of the
current and voltage to the precipitator and the corresponding
average inlet temperature and water feed rates was maintained
during the test.  (This log is tabulated in Appendix I).

In an attempt to avoid the effects of changes in gas tempera-
ture and moisture content, the precipitator was operated at a
current density and voltage for which sparking rates were low.
However, during the first day of continuous testing, the gas
temperature increased during the day and the current density
decreased.  (See Appendix I, June 23, 1975).  During the sub-
sequent test the water feed rate was increased to lower the
gas temperature and to increase the moisture content, thus
lowering the ash resistivity.  These changes are reflected in
the data tabulated in Tables 3 and 5.  Following these changes,
the electrical conditions remained reasonably stable during
the test, as shown by the data tabulated in Table 6.  Typical
V-I characteristics of clean plates and ash covered plates are
shown in Figure 15.   During the test the average operating
voltage was 40.4 kV and the average current density was
23.3 nA/cm2.  When sparking took place it usually occurred in
the lanes having expanded metal plates.

The average temperature of the flue gas during  each test is
tabulated in Table 6.

GAS VELOCITY DISTRIBUTIONS AND VOLUME FLOW RATE
Rapping reentrainment was shown to be a function of gas
velocity by Sanayev and Reshidov.15  The test plans called
for measuring the rapping emissions with gas velocities of
                              54

-------
Table 6.  AVERAGE ELECTRICAL CONDITIONS

Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Test 7
Test 8
Average
voltages ,
kV
38.25
32.60
40.25
43.50
41.00
42.25
41.00
44.00
Average
Current Resistivity,
nA/cm2
21.0
14.5
23.9
23.6
23.9
22.7
24.9
24.2
Average
Temperature
°C
141
141
125
128
127
130
128
130
                55

-------
           180
           160
           140
           120
           100
       z
       UJ
       ct
       IT

       O
            80
            60
            40
            20
                  150°C AIR CLEAN ELECTRODES

                  130°C ASH ON PLATE
                                            •
             10       20        30        40        50


                          RMS VOLTAGE, kV








Figure 15.   Precipitator electrical characteristics
                          56

-------
0.9 m/sec and 1.8 m/sec between the precipitator plates.  These
two velocities cover the range found inside most modern
precipitators.

During the second week of testing, the average velocity between
the precipitator plates was estimated to be 0.86 m/sec
(2.8 ft/sec).  The estimate was based on measurements 76 cm
(30 in.)  downstream of  the  exit baffles.   The vertical veloc-
ity distribution at this location is shown in Figure 16.  The
average velocity and standard deviation of the measurements
for each test day are tabulated in Table 7.  Considering the
design of the precipitator, the vertical distribution of the
gas flow between the precipitator plates was probably similar
to that 76 cm downstream of the exit baffles.  The distribu-
tion indicates that the velocity was slightly higher in the
lower half of the precipitator than in the upper half and that
the velocities in the baffled regions were reasonably low.
The velocity distributions at the entrance and at the exit
of the precipitator are shown in Figures 17 and 18, respec-
tively.

The vertical velocity distribution at the inlet was reason-
ably good except for a perturbation caused by a structural
I-beam.  The velocity distribution across the precipitator
indicated a lower velocity in the middle than on the two
sides of the precipitator.

The total volume flow rates of gas passing through the pre-
cipitator were determined from a velocity traverse of the
1.1 m inlet duct.  The volume flow rates for each test are
tabulated in Table 8.  The average flow was 5.04 m3/sec
(dry standard T=21°C, P=760 Torr) with a standard deviation
of 0.21 m3/sec.
                              57

-------
Q.   4
til   **
Q
      LL
      LL

      <
      GQ
                                            i         r
                                     I        I
            0.25      0.50
0.75       1.0      1.25


 VELOCITY, m/sec
                                                     1.5     1.75
   Figure 16.   Precipitator outlet  velocity profile.
                               58

-------
             Table 7.  Average Velocities  76  cm
               Downstream of the Exit  Baffles
  Date
Test
Average Velocity
 m/sec  (ft/sec)
  Standard
  Deviation
m/sec  (ft/sec)
Percent
Standard
Deviation
June 18            1.00    (3.27)     0.30    (0.99)

June 19            0.86    (2.82)     0.46    (1.51)

June 20            0.69    (2.27)     0.38    (1.24)

June 23  1 and 2   0.79    (2.60)     0.43    (1.40)

June 24  3 and 4   0.95    (3.13)     0.46    (1.51)

June 25  5 and 6   0.88    (2.90)     0.46    (1.52)

June 26  7 and 8   0.85    (2.79)     0.50    (1.63)
                                               30

                                               54

                                               54

                                               54

                                               48

                                               53

                                               58
                              59

-------




1


2
E
CO
HI
_J
LL
U.
CO
z
UJ
g
P 3
UJ
00
I
r™
0.
UJ
O






5


1 1
O


PORT
^ NO. 2
O O NO. K i xNO. 3
	 a o

a o A
O A
o a £
O A


a o ^
o &

A 0


O A

a o £
o a A
o a A

n — L
5.2m

^ 	 »_ 1.22m




I
-BEAM 	
LOCATION







O O & PORT NO. 1
O PORT NO. 2
a A a PORT NO. 3

00 &
I I
0 0.5 1.0 1.5 2.0 2.5
VELOCITY, m/s
	


3.0

Figure 17.  Precipitator inlet velocity profile.
                   60

-------
                          ft/sec
SYMBOL LANE NO.
Ov
1
0 2'
O 3'
A 4- 1
O 5'

QQQ 2
ALL VELOCITIES
MEASURED AT
CENTER OF LANE
3

E
i"
CL A
tu
Q



5


6


<
7


C





_^
FLOW

N»^
^

0 2 4 £

I I
I I




— O o a
O o
O £ DO O
	 O 0 0 	
O o&o o
0 00
O o A a
OoO
Oo & a 	

O o o
O o o& a
OOO
o o AO a
— O o

O 6 
-------
        Table 8.   VOLUME FLOW RATES DURING THE TEST
Date Test
June
June
June
June
June
June
June
June
23
23
24
24
25
25
26
26
Average flow rate
1
2
3
4
5
6
7
8
5.04
Flow rate
m3/sec
5
5
5
5
5
5
4
4
±0.21
.17
.08
.27
.27
.04
.04
.73
.73
m3/sec
(dry standard) *
(ft3/sec)
(1.
(1.
(1.
(1.
(1.
(1.
(1.
(1.
(1.78
53
79
86
86
78
78
67
67
±0
x 102)
x 102)
x 102)
x 102)
x 102)
x 102)
x 102)
x 102)
.61 x 102 ft3/sec)
kDry  standard  P  =  760  Torr,  T = 21°C.
                             62

-------
Tests with a gas velocity of 1.8 m/sec scheduled for the last
day of test were cancelled due to an electrical short in the
precipitator.

Inlet Grain Loading
An average grain  loading of  7.62  g/Nm3  (3.09  gr/dscf) was  ob-
tained in the inlet ducting for the eight tests for which
rapping data are compared.  The measurements for each test
are tabulated in Table 9.  A comparison with Brink impactor
measurements made within one meter of the inlet of the pre-
cipitator indicated that the particulate loading had decreased
to  6.44 g/Nm3  (2.61 gr/dscf) at the  inlet to  the precipitator.

The correlation of the inlet feed rates determined from mass
train measurements and the hopper feed rates is shown in
Figure 19.  These data indicate that 33% of dust leaving the
feed hoppers did not reach the precipitator.  This was not
surprising considering the amount of ducting the gas passed
through before reaching the precipitator (See Figure 7).

PLATE ACCELERATIONS
To remove a precipitated dust layer from a surface by rapping,
enough acceleration-has to be given to the surface to create
a force between the dust layer and the surface sufficient to
cause the two to separate.  If the acceleration is too small,
the dust layer is not removed and may even be compacted.  If
the acceleration is too large,  energy can be imparted to the
dust layer, reentraining it in the gas stream.

The peak X, Y, and Z axis plate accelerations measured during
the test with 1 to 2 mm thick dust layers collected on the
plates were averaged for each accelerometer location.  These
accelerations are tabulated in Tables 10 and 11 for pressure
                             63

-------
              Table 9.   INLET PARTICULATE LOAD
                 AND PARTICULATE FEED RATES
Date
June
June
June
June
June
June
June
June
June
June
Avg.
19
20
23
23
24
24
25
25
26
26
Test
Test
0-1
0-2
1
2
3
4
5
6
7
8
Hopper
Grain Load Feed rate Feed rate
g/Nm3 (gr/dscf) kg/hr (Ib/hr) kg/hr (Ib/hr)
9
6
5
8
4
7
7
6
8
8
l-8b 7
±0
Avg. collection
(load cells)
Feed
Brink
.57
.78
.66
.06
.73
.79
.13
.29
.37
.10
.62
.76
rate
110
rate at entrance
impactor - 108
(3.
(2.
(2.
(3.
(1.
(3.
(2.
(2.
(3.
(3.
(3.
±0.
880)
751)
296)a
268)
920)a
160)
892)
549)
394)
286)
092
316)
142
106
98
137
	
137
120
106
132
127
127
±12
(313)
(234)
(215)
(302)
	
(303)
(265)
(233)
(291)
(280)
(279
±26)


158
180
151
163
156
163
160
176
163


(348)
(397)
(333)
(360)
(344)
(360)
(353)
(389)
(360)
on precipitator plates ,
kg/hr (243 Ib/hr)
to precipitator as determined
kg/hr (239 Ib/hr)
by

 Hole in filter

 Omitted tests 1 and 3 from average

CNm3 - Normal cubic meter T = 21°C,  P = 760 Torr
                             64

-------
              180
              160
           m

           <
           oc
           Q
           UJ
           UJ
           LL
           Q
           Ol
           CC
140
              120
              100
                     100
                                      70
                                                   V
                                                            8O
                 120         140


                  HOPPER FEED RATE, kg/hr
160
                                                               180
Figure 19.   Correlation of measured feed rate  and hopper feed rate.
                                65

-------
 Table 10.  Composite Average of Plate Acclerations During
     Rapping for AP = 6.2 x 105 Pa, Dust Covered Plate
Verticala
location
0.91 m


3.05 m


5.18 m


Acceleration
Axis
X
Y
Z
X
Y
Z
X
Y
Z
Plate #1
22
19 ±1.5
—
13
—
11
—
11
—
Plate #2
—
24 ±6
20 ±4
15 ±2
20 ±9
11
12
14
9
, G

Plate #3
26
19
12
11
18
11
—
—
—

±3


±4




— No available data

 From the top of the plate
                             66

-------
   Table 11.   Composite Average  of  Plate Accelerations  During
       Rapping for AP  = 4.8  x  10s Pa,  Dust  Covered Plate
Vertical
location
0.91 m


3.05 m


5.18 m


Axis
X
Y
Z
X
Y
Z
X
Y
Z
Acceleration,
Plate #1 Plate #2
— —
— 19
— 18
11 14.3
— 11
9 11
— —
— —
— —
G
Plate #3
—
14
13
11
15
11
—
—
—
— No  available  data

aFrom the top of the plate
                               67

-------
drops across the rapper cylinders of 4.8 x 1Q5 Pa and
6.2 x 105 Pa (70 psi and 90 psi).  A problem with the charge
amplifiers occurred during the test, thus some acceleration
data is missing.  The X axis was parallel to the direction
of gas flow.  The Y axis was in the vertical plane of the
plates, and the Z axis was perpendicular to the plates.  The
method of rapping the plates and the accelerometer descrip-
tions and locations were discussed in an earlier section.

The measurements of the accelerations of plate 5-2 from both
the upstream and downstream rappers were averaged.  The
variations in rapping intensity as a function of the differ-
ential pressure across the rappers with no dust load on the
plates are shown in Figures 20 and 21 for the Y and Z axis
of the accelerometers on plate 4-1.  The accelerations obtained
at all nine locations without a collected dust layer (P =
6.2 x 105 Pa)  are given in Table 12.  These accelerations
are significantly higher than the accelerations obtained
with the dust layers on the plates.

The range of accelerations that were available with dust
layers on the plates of the precipitator was small.  The
change in rapping intensity from 80% to 100% of maximum pro-
duced only a small change in plate accelerations (see
Tables 10 and 11).  A modification of the air reserve for
the pneumatic rappers after completion of the test appears
to have significantly increased the rapping accelerations
(see Table 13).   The effects of changing plate accelerations
could be studied during any subsequent test with the FluiDyne
pilot precipitator.

Typical time traces  of plate accelerations with a dust layer
on the plates are shown in Figure 22.   Time traces of plate
                              68

-------
        50
    C3
    m
    o
    o
        40
        30
        20
        10
                      NO DUST ON PLATES


                      DUST COVERED PLATE
                 100      200      300      400      500



                         DIFFERENTIAL PRESSURE ON RAPPER, kPa
600
                                                                  700
Figure  20.   Plate 5-1 accelerations  as a functional  differential

              pressure  on rapper, upstream rapper.
                                   69

-------
    30
CO
z
O
cc
LJ
UJ
o
    20
    IO
                                        Y AXIS UPPER
             100      200      300      400      500


                     DIFFERENTIAL PRESSURE ON RAPPER, kPa
                                                          600
                                                             700
Figure  21.
        Plate 5-1 accelerations  as a functional differential
        pressure on  rapper, upstream rapper, no dust on plates.
                            70

-------
   Table 12.  Composite Average of Plate Accelerations
   During Rapping for AP = 6.2 x 10s Pa, Clean Plates
Vertical a
location Axis
0.91 m X
Y
Z
3.05 m X
Y
Z
5.18 m X
Y
Z
Acceleration,
Plate
49
24
29.
34.
26
37.
28
20.
13.
#1


5
5

5

5
5
Plate
—
38
34
36.
24.
26.
42.
64.
17.
#2



5
5
0
5
5
0
G
Plate
45
53
27
29.
23.
33
32
31
13

#3



5
0




— No available data

 From the top of the plate
                              71

-------
           Table 13.  Composite Average of Plate
        Accelerations During Rapping with Modified
     Rappers for AP = 6.2 x 10^ Pa, Light Dust Layer
Vertical
location Axis
0.91 m X
Y
Z
3.05 m X
Y
Z
5.18 m X
Y
Z
Acceleration,
Plate #1
73
—
51
41
80
—
75
50
33
Plate #2
IV-
70/107
41/47
40/68
33/52
-/-
69/87
51/113
35/64
G
Plate #3
151
99
35
105
33
—
37
34
34
— No available data

 From the  top  of  the  plate
                             72

-------
          (/}

          b


          o
          cc
          LU
          _J
          LU

          8
          i-

          _i
          Q.
              25
              25
              25
 25
              25
              25
                       •VtaA.
                                       X AXIS
                                      Y AXIS
                                       Z AXIS
                     0.03    0.06   0.09   0.12   0.15   0,18   0.21



                                  TIME, sec
Figure 22.
Typical traces of plate accelerations  for a dust

covered plate  [plate  5-2, 3m  from top  of plate,

AP =  6.2 x 105 Pa  (90 psi)].
                                73

-------
accelerations without dust layers are given in Figure 23.
The plate accelerations of plate 5-1 were small when the down-
stream rappers were activated and the acceleration of plate
5-3 were small when the upstream rappers were activated.

The accelerations of plate 4-3 suspended from load cells
were compared with the accelerations of plate 5-3.  The
results of the comparison of peak accelerations for the lower
accelerometers is as follows:
                          Plate 4-3     Plate 5-3
          Y axis             20G           25G
          X axis             19G           31G
          Z axis             13G           12G
This data indicates that the load cells dampened the accelera-
tions , but that in the important Z axis the accelerations
were nearly the same.
                             74

-------
               CO
               o


               a
               CC
               UJ
               UJ
               o
               UJ
      50


      25


       0


      25





      25


       0


      25






      25.


       0


      25



      50


      25


       0


      25


      50.


      25


       0


      25
                                 1     I     I     T
                                               AXIS

                                               y|ftA4MMlM
                                               AXIS
AXIS
                                            Y2AXIS
                          0.03   0.06    0.09   0.12   0.15  0.18   0.21


                                      TIME, sec
Figure 23.
Typical traces of plate accelerations for  a clean
plate [plate  5-2; Xi,  Yi,  Zi,  0.9 m from top of plate;
X2, Y2  3m  from top of  plate;  AP = 6.2 x 10s  Pa

(90 psi)].
                                   75

-------
                          SECTION V
                  RESULTS AND DISCUSSION

 The measurements discussed in this section were aimed at
 quantification of rapping reentrainment in terms of the per-
 centage  and the particle size distribution of the reentrained
 particulate and at determining the conditions for removal of
 dry dust by rapping.  The results of the measurements indicate
 that  efficiency of dust removal increases as a function of mass
 per unit area collected on the plates between raps while the
 percent  of material reentrained decreases as a function of mass
 per unit area.  These results are in agreement with those pre-
 viously  obtained by Plato11 and by Sanayev and Reshidov.15
 Measurements of particle size distribution showed that the mass
 median diameters of the particles emitted during the raps in-
 creased  with increased time between raps or equivalently with
 mass  per unit area collected on the plates between raps.

 MASS  EMISSIONS
 The results of the mass emission measurements provide data for
 determining the effect of changes in rapping variables, for
 determining the vertical stratification of the dust at the
 outlet,  and for determining the percentage of total emissions
 due to reentrainment for various test conditions.  Independent
 measurements of the mass emissions of the dust that passed
 through  the electrified region without being collected and
 the mass emissions due to rapping reentrainment were made using
 the apparatus and procedures described in an earlier section.
 Separate measurements  were made of emissions from the upper
68% and the lower 32% of the precipitator.
                              76

-------
Separate sets of mass train data were obtained for the fol-
lowing rapping conditions:

                   Rapping Intensity      Rapping Interval
     Test 1           100% of max           150 minutes
     Test 2           100% of max           120 minutes
     Test 3            80% of max            12 minutes
     Test 4            80% of max            32 minutes
     Test 5           100% of max            12 minutes
     Test 6           100% of max            32 minutes
     Test 7            80% of max            52 minutes
     Test 8           100% of max            52 minutes

Since the difference observed in plate acceleration  (for the
two different rapping intensities) was small (see previous
section),  data from tests having the same rapping intervals
were combined to generate an average set of data.  Data for
the two tests with rapping intervals of 120 and 150  minutes
were also combined and averaged.  The resulting set  of mass
train data showed that hourly emission rates for the emis-
sion due to rapping reentrainment decreased as the time in-
terval between raps increased.  Emissions due to rapping were
reduced by collecting thicker dust layers on the plates be-
fore they were rapped.

Precipitator efficiency eventually decreases if the  collected
dust layer becomes too thick.  However, the between  rap data
showed that the emission rates and particulate loadings
between raps remained constant during these tests as the time
interval between raps was increased, for time intervals less
than 155 minutes.  The above effects are illustrated in Fig-
ures 24 through 29.  The effective particulate loading for
the rap contribution given in Figures 27 and 29 were calculated
by dividing the rate of emission due to rapping by the

                              77

-------
00
      5 <
      LU cc
                 10    20
                                  40    50
60    70    80    90




TIME BETWEEN RAPS, minutes
100    110    120    130   140    150
                  Figure  24.   Mass emission rates  for lower  32%  of the precipitator

                               cross sectional area as a function of the time interval

                               between  raps.

-------
-J

vo
        5 <
        \u a:   2
                                          TOTAL UPPER 68%
                                                       BETWEEN RAPS UPPER
                                                                              RAP UPPER
10     20    30    40    50    60     70    80    90





                             TIME BETWEEN RAPS, minutes
                                                                     100
                                                                           110
                                                                                 120   130
                                                                                            140
                                                                                                  150
                   Figure 25.  Mass emission rates  for upper 68% of  the precipitator

                                cross sectional area as a  function  of the time interval

                                between raps.

-------
O)
UJ

<
CC


O
13




12




11



10




 9



 8




 7




 6




 5



 4



 3



 2




 1
             20
COMBINED BETWEEN RAPS
                    40
                       60
                                   80
                                          100
               120
                                                        140
                            TIME BETWEEN RAPS, minutes
 Figure 26.   Combined  upper  and lower  mass emission rates as a
               function  of the  time interval between raps.
                           80

-------
CO
 E
 Z
 ~oi
 Q"

 O
 _J
 UJ

 5
  o

  en
                                                                       u_

                                                                       §
                                                                       fe
                      40      60      80     100     120


                         TIME INTERVAL BETWEEN RAPS, minutes
                                                        140
Figure  27.  Particulate  loading - lower 32% of  the precipitator as
             a function of  time interval between raps.
                                  81

-------
 E
 2
 o>

 Q"

 O

 LLI
 I-
 D
 O

 IT
2.4


2.2


2.0


1.8


1.6


1.4


1.2


1.0


0.8

0.6


0.4


0.2
        -  '    |
Figure  28.
1.0


0.90


0.80


0.70


0.60  u_
                                                                        0.50  ^
0.40


0.30


0.20


0.10
                       40      60      80      100      120


                           TIME INTERVAL BETWEEN RAPS, minutes
                                                               140
         Particulate  loading -  upper 68%  of  the precipitator as
         a  function of time interval between raps.
                                      82

-------
00
                  n

                   a>
                   d
                   O
                   K
                   tr
2.4
2.2
2.0
1.8

1.6

1.4

1.2

1.0

0.8
0.6
0.4


0.2
~ ' | ' | ' I ' I ' | ' I ' I '
— —
I — —
— • STANDARD EFFICIENCY
TEST —
—
__
—

	 —

— —
a
~~ ^^v^ —
~~ S~---— _ COMBINED TOTAL _^_
"V^ ° COMBINED BETWEEN RAP
^^^^^ ^""^
— **"^^_ COMBINERAP, UPPER AND LO\NER — A«—
1 .U
0.90
0.80

0.70

0:60


0.50

0.40

0.30
0:20

0.10








u_
O
a
O)







                                20      40      60      80      100     120


                                         TIME INTERVAL BETWEEN RAPS, minutes
140
                Figure  29.   Combined particulate load as a function of  time interval
                             between raps.

-------
appropriate volume flow rates.  The total emission rates of
the combined rap and between rap periods, for the upper and
lower portions of the precipitates  were obtained by adding
the rap and between rap emission rates.  The total upper and
lower emission rates were added to give total emission rates.
The total effective particulate loading was calculated from
the total emission rates.

Good agreement was obtained between the total effective out-
let grain loading determined by the above procedure and an
outlet grain loading measured when the precipitator was
operated in a normal manner.  During the normal test, the
precipitator was operated with a continuous dust feed.  The
upstream rapper was activated every 30 minutes and the down-
stream rapper every 60 minutes.  During this test, mass
emissions were determined using standard in-stack filter
techniques.  The procedures used during the intensive test
program  gave an effective total grain  loading of  0.58 g/Nm3
 (0.235 gr/dscf) for the  32  minute interval  test  and 0.49  g/Nm3
 (0.19 gr/dscf) for the 52 minute interval test.   The more  nearly
normal procedure gave a  grain  loading  of 0.53 g/Nm3 (0.215 gr/
dscf) for  the  30-60 minute  rap interval test.

Based on the data given  in  Figure 29 and the average inlet
grain loading, the overall  precipitator efficiencies were  de-
determined to  be:
          12 minute interval tests          88.6%
          32 minute interval tests          92.4%
          52 minute interval tests          93.9%
     120-150 minute interval tests          93.1%

-------
For the normal test  (30-60 minute rap interval) the measured
efficiency was 93.0%.  The average efficiency of collection
without including the rapping loss was 94.8%.  These values
show that significant changes in emission were obtained by
changing rapping intervals.  However, it should be noted that
the pilot scale precipitator is basically only one section of
a full scale unit.  The effect of rapping reentrainment for
larger, more efficient units is discussed later in this section.

The data given in Figures 24 through 29 show that the emissions
between raps and during the raps in the lower 32% of the
precipitator were significantly higher than the emissions
in the upper 68% of the precipitator.  For the tests with
32-minutes rapping intervals, the emission due to rapping in
the lower 32% of the precipitator accounted for 82% of the
total emissions due to rapping and the emissions between raps
in the lower 32% of the precipitator accounted for 60% of the
between rap emissions.  The between-rap vertical gradient is
probably due to gravitational settling, sneakage through the
hoppers, and reentrainment in the hoppers.  The vertical
stratification of the rapping puffs was visually observed and
was photographically recorded.  The major sources of vertical
stratification during the rapping puff appeared to be
gravitational settling and hopper "boil-up".  The dust re-
moved from the plates  appeared  to drop  into the hoppers  and
then to boil up over the hopper baffles and be carried out of
the precipitator.  The vertical stratification in particulate
concentration was clearly evident with the real time system
and is discussed in more detail later in the report.

As mentioned in the introduction, an improvement in overall
efficiency with increasing time between raps had been pre-
viously observed during a Southern Research Institute study
                              85

-------
of rapping reentrainment at the Bull Run Steam Plant in the
TVA system,17 by Plato,11 and by Sanayev and Reshidov.15  In
Figure 30 the percentage of particulate removed from the
plates that escaped the precipitator is shown as a function
of mass per unit area of dust collected on the precipitator
plates between raps.  Data obtained from the Southern Research
Institute-FluiDyne experiment and from the laboratory studies
of Sanayev and Reshidov are shown.  The Southern Research
Institute percentages were calculated in two ways.  First,
the amount of emissions per rap was divided by the average
mass collected between raps.  The average mass collected was
determined for the entire length of the precipitator from
load cell measurements which are discussed later in this sec-
tion.  Second, the percentages were calculated in terms of
mass collected only in the first 0.9 meters of the precipita-
tor neglecting the smaller quantity of dust collected in the
last 1.8 meters.  According to this data, during the Southern
Research Institute-FluiDyne experiments only 2.7% to 7.2%
of the dust collected on the plates was reentrained and
emitted from the precipitator.  The data of Sanayev and
Reshidov show that these percentages depend not only on the
dust layer mass per unit area, but also on flow velocity.
The data in Figure 30 indicate that the most efficient re-
moval of the dust from the plates and transfer to the hoppers
for the ash used in these experiments was obtained with a
dust surface density of 0.5 kg/m2 or greater.  Sanayev and
Reshidov15  explained the dependence of reentrainment on dust
surface density as follows.  They claimed that thin dust layers
are more intensely disaggregated during rapping than thicker
layers.  The thin layers produce fine aggregates that are easily
reentrained, while thicker layers lead to an increase in
size and number of large aggregates reaching the hopper.
                              86

-------
                        20
cr>
                   (3
                 z ?
                 0 fc
                 o <
                 UJ rr
LU  I—
-I  LU

8  °

3*
Q  t
                 <-> T
                 S5 f
                 Q. (/)
                   LU
       15
                        10
SOUTHERN RESEARCH INSTITUTE DATA
BASED ON MASS FIRST. .91m
SOUTHERN RESEARCH INSTITUTE DATA
BASED ON TOTAL MASS ON PLATES
SANAYEV & RESHIDOV5 GAS VELOCITY 2m/sec

SANAYEV & RESHIDOV15 GAS VELOCITY 3m/sec
                 V
                                                                  1.0
                                                                                       1.5
                                                       DUST SURFACE DENSITY, kg/m2
                       Figure  30.
                     Percent of  dust  collected on precipitator plates  that
                     is  emitted  due to rapping.

-------
 These effects  were  also observed by  Plato.11   Plato found
 that with dust surface densities greater than 1  kg/m2,  dust
 was  removed in cakes,  and for values less than this it  formed
 clouds.   Impactor measurements made  during the Southern
 Research Institute-FluiDyne experiments showed that the
 thicker layers produced larger aggregates than the thin
 layers.   We expect  the percentage of reentrained dust to
 decrease faster when large agglomerates are produced than
 when small agglomerates or discrete  particles are entrained
 due  to rapping.  The large agglomerates will  be  recollected
 faster by gravitational settling and by electrostatic pre-
 cipitation than the small ones.

 It should  be noted,  as mentioned by  Sanayev and  Reshidov,  that
 the  data  shown  in Figure  30 characterize  the  optimum con-
 ditions  for the rapping process  and  do  not consider the
 precipitation process.  At FluiDyne  the effect of  the col-
 lected dust on  the precipitation process  appeared  to be
 minimal.   As is discussed later, this is  not  the case for  higher
resistivity dusts.

 RESULTS OF IMPACTOR  MEASUREMENTS
 Precipitator performance, in addition to  being a function
 of many other variables,  is a  function  of the particle size
 distribution of  the  dust  being collected  in the precipitator.
 Reentrainment when the plates  are rapped  significantly changes
 the particle size distribution of the dust in the  precipitator.
 To obtain data  on this change, we measured the particle  size
 distribution of  the  particulate  in the  gas stream  at the inlet,
 at the outlet between raps and at the outlet  during raps.
 The measurements were made with  impactors.  The resultant  data
 are  used  to indicate changes in  the  particle  size  distribution
 of the suspended particulate when the plates  are rapped, to
 calculate fractional collection  efficiencies with  and without
                              88

-------
rapping reentrainment and to provide input data for a com-
puter model of electrostatic precipitation of dust including
rapping reentrainment.

The particle size distribution of the particulate at the
inlet of the precipitator is shown in Figures 31 and 32.  A
procedure by Davies22 was used to correct the distribution for
an isokinetic error.  The impactor samples at the inlet were
collected with a sample velocity of  1.5 m/sec while  the flue
velocity at the sampling point was  1.2  m/sec.   The  corrected
distribution is shown by the dashed curve in Figure 31.  The
measurements shown in this graph were made within a half meter
of the entrance screen of the precipitator.

The average cumulative percent distribution of the inlet
particulate distribution is given in Figure 32.   This data
represents a composite of data obtained from separate loca-
tions at the precipitator entrance.  This data indicates that
the inlet distribution was log normal with a mass median
diameter of 13 ym and with a geometric standard deviation of
3.3.

Outlet Between Rap Particle Size Distribution
The average outlet particle size distributions between raps
for all of runs is shown in Figure 33.   The distributions for
both the lower and the upper impactor are shown.  In Figure 31,
the weighted average of these two distributions is given.  The
distributions were weighted according to the area of the pre-
cipitator sampled by the lower and upper mass trains.  The im-
pactors were situated near the middle of each of these areas.
The lower impactor showed a higher loading in the 1 to 8 ym
range.  This was probably due to gravitational settling and
sneakage through the hopper area.  The finding of the same
                              89

-------
       1000
   01
   O)
   o
   13
   E
        100
                                   CORRECT FOR AN ISOKINETIC ERROR    —I
                                                32% LOWER
                                                68% UPPER
                            1                 10

                              PARTICLE DIAMETER, um
100
Figure 31.   Inlet and weighted outlet particle size distributions,
              dm/d log D.
                                90

-------
               100
DC
HI

HI
5


O

LU
_l
O


a:
               10
              0.5
        I   Tl  I    T
                                 I    I  '  I  '  I   '  I
                     j I   .   .I   I   .  I ,  I  i   I
                0.1     1         10   20    40   60    80   90


                      PERCENT LESS THAN INDICATED SIZE, by mass
Figure 32.   Inlet  cumulative percent particle  size distribution.
                             91

-------
       104
   5
   8
   O
   ^i
   a"
   S
   TJ
       10"
       10'
         0.1
                                   • INLET
                                   • OUTLET LOWER
                                   A OUTLET UPPER
                          1.0                10
                             PARTICLE DIAMETER, urn
100
Figure  33.  Inlet and outlet (lower 32% and  upper 68%  of the
             precipitator)  particle  size distribution,  dm/d log  D.
                             92

-------
concentrations for particles greater than 10 ym was unexpected
and may be an artifact of the measurement methods  resulting
from the lack of a complete exit plane traverse.   The outlet
emission rates between raps are given in Figure 34 for various
size intervals.

Outlet Particle Size Distribution During Raps
The outlet emissions measurements with the impactors agreed
with the results obtained with the mass trains and the real
time system.  All three systems measured a decrease in emis-
sions as the time interval between raps was increased.  The
results of the impactor measurements are shown in Figure 35.

The particulate loadings during the rapping puffs determined
from the impactor data are shown in Figure 36.  These data
have considerable scatter.  The real time system also showed
large variations in rapping puff concentrations.  The particu-
late loading during the rap was as much as 10 to 200 times
the particulate loading between raps.  The particulate load-
ing during the rapping puffs increased as the dust layer
thickness increased.  Data given in Figure 35 and 36 also
illustrates the vertical stratification of the rapping puffs.

The particle size distribution of the rapping puffs measured
with the lower impactor are shown in Figures 37 and 38.  The
data in Figure 38 indicate that as the time interval between
raps was increased the mass median diameter of the particles
increased.  These data show that a smaller percentage of fine
particles and a larger percentage of large particle agglomerates
are produced with thick dust layers than thin dust layers.   An
inspection of the upper impactor stages for impactors sampling
rapping puffs indicated that a substantial number of particles
smaller than the lower cut off for the stages were present.
Particles on these stages appeared to be agglomerates made
up of fine particles.

                              93

-------
                     10'
  g
  I

  §
  W
                     100
                 oc
                 z
                 o
                 HI
                     10-1
                    10-2
                                 UPPER 68% OF THE
                                  PRECIPITATOR
                                 LOWER 32% OF THE
                                  PRECIPITATOR
                         SB
               SI
                                    se
                                   Stage

                                   S1
                                   S2
                                   S3
                                   S4
                                   S5
                                   S6
                                   S7
                                   S8
S5
                                              S4
                                                   S3
                                                         S2
                                                             S1
                                         STAGE
                                                Size Interval
                              8.5
                              5.3
                              3.7
                              2.4
                              1.1
                              0.67
                              0.50
           12.2 urn
           12.2pm
            8.5pm
            5.3pm
            3.7 pm
            2.4pm
            1.1 pm
           0.67 pm
Figure  34.
Outlet emission  rates between  raps  for various size
intervals.
                                     94

-------
U1
                   01
                   5
                   DC
                                20
                                                           LOWER IMP ACTOR

                                                           UPPER IMPACTOR
40       60       80       100


     TIME INTERVAL BETWEEN RAPS, minutes
                                                                          120
                                                                                   140
                                                                                           160
                      Figure 35.   Outlet emission  rates measured with  the impactors  as
                                    a function of the  time interval between raps.

-------
CTi
               n
               E
               1C
               o
               z
               IT

               Q
               Q
               uu
               O
               <
               DC
               LLI
               OJ
               ID
               O
               tr
               <
               0.
96

90

84


78


72

66

60

54


48


42

36


30

24


18

15


 6
                              20
                   40        60       80       100       120


                         TIME INTERVAL BETWEEN RAPS, minutes
                                                                                   140
160
                          Figure 36.   Average  particulate  load  during  a rap  as a
                                        function of the time interval between  raps.

-------
     105
     10*
  o>
  1
     103
                                    AVG.
                 • TESTS

                 * TESTS 4 & 6

                 A TESTS7&8

                 • TESTS 4, 5,6, 7, & 8
     102
       0.1
1.0                10


 PARTICLE DIAMETER, urn
100
Figure 37.   Particle size distribution for rapping puffs,
             dm/d log D's  (average volume of gas  sampled during
             the rapping puffs was determined from the sampling
             rate and the average  persistence time of the rapping
             puffs)  tests 5,  4  &  6,  7 & 8 had 12,  32, and 52
             minutes, respectively,  between rap intervals.
                           97

-------
                100
                 10
            oc
            LU
            5
            LU
            o

            OL
                 1.0
                0.1
                        I  '   '  I
                          I    I  '   I  '  I
                                    BARS REPRESENT SCATTER
                                    FOR TWO TEST EACH FOR
                                    THE 32MIN AND52MIN
                                    RAP INTERVAL TEST
                                    I
                                       I    I.I.I.
                  0.01   0.1     1        10   20    40   60

                        PERCENT LESS THAN INDICATED SIZE, by mass
                                            80
Figure 38.
Cumulative percent distribution for rapping puffs, lower
32% of the precipitator cross  sectional  area,  rapping
intervals of  12,  32 and 52 minutes.
                                  98

-------
The emission rates for various particle size intervals  (each
irapactor stage) are shown in Figures 39 and 40 for the  upper
and the lower impactors.  The lower area emission rates were
calculated on the basis of a sampled cross sectional area
of 5.6 m2 (60 ft2) and the upper area emissions were calculated
on the basis of a sampled cross sectional area of 8.9 m2  (96 ft2)
By summing the emission rate measured by the upper and  lower
impactor in a particle size band, the total emission rate in
each size band can be obtained.  A comparison of the data
shown in Figures 39 and 40 shows that the emissions in the lower
portion of the precipitator dominated the rapping emissions.
As already mentioned previously, the rapping emissions  in the
lower portion of the precipitator appeared to be due to gravi-
tational settling and hopper boil-up.  The particle size
distributions measured in the upper portion of the precipitator
have large fluctuations.  The fluctuations in the data  are
possibly due to real variations in the quantity and distri-
bution of the reentrained particulate.

The percent contributions of rapping puffs to total emissions
for each particle size are shown in Figure 41.  This data
shows that rapping puffs contributed up to 60% of the emis-
sion of particles with diameters greater than 1 ym.  The
data also indicate that the percent contribution for the parti-
cle sizes less than 3 um decreases as the time interval be-
tween raps is increased.  These percentages are equivalent to
the percentages defined by equation 5 in the introduction.
Their application for computing rapping emissions is discussed
later in the report.
                              99

-------
        LU
        I-
        <
        CC
        Z
        g
        55
10'1
         i LU
         . o
        IS
         r
        < LU
        CC 0.
              10'3
                      	12MINTEST
                      — X	 32MIN TEST
                      	52 WIN TEST
                   S8
                         S7
                             S6
                                   S5
                                        S4
                                             S3
                                                 S2
                                                       S1
                                   STAGE

                              Stage      Size Interval
                              S1
                              S2
                              S3
                              S4
                              S5
                              S6
                              S7
                              S8
                       8.5
                       5.3
                       3.7
                       2.4
                       1.1
                       0.67
                       0.50
12.2 urn
12.2 urn
 8.5 wm
 5.3 urn
 3.7 «m
 2.4 jjm
 1.1 urn
0.67 «m
Figure 39.
Mass emission  rates for  rapping reentrainment,  lower
32%  of precipitator exit for  various particle size
intervals•
                                 100

-------
                    1x10°
              HI
              cc
              2
              01
              
-------
   Ill
   N
   53
   LU
   DC
   <
   o.
   cr
   o
   LL
   a.
   O
   z
   X


   o
   CC
   CO
   c/J



   LU
       70
60
50
40
       30
       20
        10
                                    TIME BETWEEN

                                    RAPS


                                    12min

                                    30 min

                                    50 min
                                6       8      10



                                 PARTICLE DIAMETER, urn
                                               12
                                                             14
                                                              16
Figure 41.   Percent emissions due to rapping  reentrainment  as a
              function of particle  size.
                                102

-------
RESULTS OF REAL TIME MEASUREMENTS OF RELATIVE CONCENTRATIONS
IN FIVE SEPARATE PARTICLE SIZE INTERVALS
Dust removal from the precipitator plates by rapping produces
temporal and spatial variations in the concentration and parti-
cle size distribution of the dust leaving the precipitator.
The rapping puffs can account for a significant portion of the
emissions for some particle sizes.  Two modified Royco Model
225 particle counters with five-channel analog ratemeters were
used to monitor rapping emissions in time and space in the
following five diameter intervals:  1.5 to 3 ym, 3 to 6 ym,
6 to 12 ym, 12 to 24 ym, and greater than 24 ym.

Relative emissions between raps for each size interval were
determined by multiplying average  count rates between  raps
by the time interval between raps.  The relative emissions from
rapping puffs for each size interval were calculated by inte-
grating the ratemeter output over the time interval for which
the rapping puff persisted.  The dust feed was turned off
when the plates were rapped;  thus rapping puff emissions
were observed independently of the emissions not due to
rapping.

Although a size-selective dilution system was used, concentra-
tions in the 1.5 to 3 ym and 3 to 6 ym diameter intervals were
too high for the Royco system during some raps for accurate
analysis, while concentrations in the 12 to 24 ym and greater
than 24 ym intervals were so low that a statistical analysis
based on a Poisson distribution of counts had to be used to
determine total relative emissions in the large particle size
intervals.
Vertical concentration gradients between raps were explored
with the real time system.  Significant increases in concen-
tration toward the bottom of the precipitator were observed.
Between raps, the vertical distribution of particles in the
1.5 to 3 ym size interval was found to be relatively constant
                             103

-------
while the vertical distribution in the other particle diameter
ranges decreased with increasing elevation.  An example of
the vertical stratification is shown in Figure 42.  These data
indicate that there was some change in the particle size dis-
tribution of the suspended particulate with increased elevation,

The vertical stratification of the particulate is attributed
to several causes.  In the lower portion of the precipitator
the increased emissions are due to hopper sneakage and to
gravitational settling of the particulate, especially for
the larger particle sizes.  Sneakage through the corona wire
and plate support structure appears to explain the increase in
emissions observed at the top of the precipitator.

The relative total emissions between raps in the five particle
size intervals studied with the real time system were compared
for time intervals of 12, 32, and 52 minutes between raps.
The data indicated that the between-rap emission rates did
not change with the increase from 12 minutes to 32 minutes
between raps.  However, there were increased emission rates
at the sampling point just above the lower baffle for the 52
minute interval test.  Emissions at this point are probably
dominated by sneakage and may be signficant only for this
particular precipitator-hopper geometry.  The increase may
have resulted from voluntary rapping or "sloughing".  Relative
emissions with respect to the 12-minute interval between rap
tests are tabulated in Table 14.  In the absence of changes in
collection efficiency with increases in dust layer thicknesses
on the plates, the ratios in Table 14 would be expected to be
the same as the ratios of the time intervals between raps.

The vertical stratification of the rapping puffs that was ob-
served with the impactors and mass sampling trains is
clearly shown by the real time particle size data given in

                              104

-------
         z
         o
         I

         i
         Q
         LU
         Z
         3
         o
         o
         ta
         in
         _l
         O

         IT
         d
         z
             10'
             106
10*
            103
            102
             10
                                          i     i    i    r
                              PARTICLE DIAMETER BAND

                                 • 1.5-3/jm
                                 • 3 -6pm

                                 ^ 6- 12 urn
                                 A  >-24/jm
                   I     I     I
                B  20
                A
                F
                F
                L
                E
           40    60   80   100  330  350   370   390  410

                DISTANCE FROM BOTTOM BAFFLE, cm
Figure 42.
  Spatial distribution of  between rap  emissions  for
  various particle  sizes.
                                 105

-------
       Table 14.   RATIO OF RELATIVE TOTAL BETWEEN RAP
       EMISSIONS  IN FIVE PARTICLE SIZE INTERVALS FOR
        RAPPING INTERVALS OF 12, 32, AND 52 MINUTES
Particle Diameter Interval
(microns)
Time Between
Raps
12 min
32 min
52 min

12 min
32 min
52 min
Time
Ratio
1
2.7
4.3

1
2.7
4.3
1.5-3
3-6
6-12
12-24
>24
Emission Ratio Lower Sampling Points
1
2.7
6.0
(5.9)
1
2.9
10.0
(12.7)
1
1.1
3.7
(8.0)
1
2.5
6.0
(14.3)
1
.8
2.9
(5.3)*
Emission Ratio Upper Sampling Points
1
2.7
5.5
1
2.5
7.8
1
2.5
4.4
1
3.1
2.8
1
4.6
2.3
*Includes lowest sampling point
                              106

-------
Figures 43 and 44.  The data in Figure 44 shows that stratifi-
cation occurs for both the upstream and the downstream raps.
The emissions at 23 cm and 42 cm above the lower baffle were
nearly the same.  For higher elevations above the baffle the
dust concentration in the puffs decreased with increasing height.

The puffs in the lower portion of the precipitator were ob-
served to last approximately 28 seconds and 24 seconds respec-
tively for the upstream and downstream raps.  The rapping puffs
observed in the upper portion of the precipitator lasted about
9 seconds for the upstream rap and 7 seconds for the downstream
rap.  Typical traces of upstream and downstream rapping puffs
for both the upper and lower sampling locations are shown in
Figures 45 through 48.  On the basis of the flow velocity
(^ .9 m/sec) through the precipitator, it was estimated that
the rapping puffs should last only 2 seconds for both the up-
stream and downstream rap.  This estimate was based on a rap
of a 1.8 meter length of the precipitator and the assumption
that the reentrained particulate moves with the velocity of
the gas through the precipitator.  The raps observed with the
real time system persisted for 3 to 19 times the 2 second
estimate.  In the lower half of the precipitator two separate
puffs were observed.  The first puff appeared to last for
approximately 2 seconds, and the second puff approximately 25
seconds for the 1.5-3.0 um diameter particles. The persistance for
large particles was somewhat less.  Examples of rapping puffs
in the lower portion of the precipitator are displayed in
Figure 45 for both upstream and downstream raps.  The two
separate rapping puffs from both the upstream and downstream
raps were recorded on film.  This data indicates that some of
the dust removed from the plates when they are rapped is
reentrained directly into the gas stream and is carried out
of the precipitator at the gas velocity.  The remainder of the
dust drops into the hoppers.  Of the dust that drops into the
                              107

-------
IV



105


LL
LL
Q.
O
? 104
Q.
CC
z
03
HI
I
0
i io3
Q.
d
z


102


ml
I I I I I "|
DOWNSTREAM
0 UPSTREAM I PARTICLE D.AMETER
RAP BAND
2 Q O*1.5- 3jjm
O O • 3- 6tim
O •O^6-12um
^ ^ _ D O •l2-24jum
« D
^ 0 0
£ «
B 5^
— - 0^0 •—
• B ^^
o o
* , 8 §

^— ^ • *— -«

^
CJ ^^
6
A A 0
— A A 0 —
A &
A i
I I I I I 7, I
                  20   40    60   80   100

                          POSITION ABOVE BAFFLE, cm
                                                 406
Figure  43.
Spatial distribution of particles in upstream  and
downstream rapping puffs for  12  minute rap interval
test
                            108

-------
      Q
      UJ
      Z
      < I-
      1- UJ
      Z W

      o*1
      UJ <
      o«
      WOE

      <|
      o 5

      II
      oc z
      u. <
      z-
      ? IE
      1- £0 Q.
              10b
10B
103
              102
                           U
                     I    I     I     I     I "I
    I     I     I
PARTICLE DIAMETER

    • 1.5- 3.0 jam

    • 3.0- 6.0 ijm
   ^ 6.0-12.0 jam
    • 12.0-24.0 jam
    A  >  24 jam
                 -•   •
   8  20
   A
   F
   F
   L
   E
                                        -U
                                                 \     \     \
                         40   60   80    100  330   350   370   390  410
                              DISTANCE FROM BOTTOM BAFFLE, cm
Figure  44.  Spatial distribution  of particles in combined  upstream
              and downstream  rapping puff.
                                     109

-------
                   -H
                   o
                   c
                   o)
                   I-
                   z
                   D
                   O
                   o
50

40


30


20


10
                      500

                      400
                   «  300
                   z
                   O  200
                   o

                      100
                                 > 24 urn DIAMETER PARTICLES
                          I  I  I  I
               I.  I  I  I I
                                                Ll
                                 12 - 24 urn DIAMETER PARTICLES
                    5x10J
                    I  2
                    O
                    o
                       1
                                 6 - 12 urn DIAMETER PARTICLES
                    5x103
                    o
                    c.  3
                         _  UPSTREAM
z

u
                     5x103|	
                                  3 - 6 urn DIAMETER PARTICLES
                                          DOWNSTREAM 	
                                 1.5 - 3 wm DIAMETER PARTICLES
                          I  I  I  Lf I  I  IVI  I
                        I I JI KL_
                                   TIME. 12sec/div
Figure 45.   Rapping puffs lower portion  of the  precipitator,  32
              minutes between  raps.   First set of puffs,  upstream
              rap.   Second set of puffs, downstream rap.
                              110

-------
                                 24 «m DIAMETER PARTICLES
o
-?
6
c
(/>
1-
Z
O
u

ou
40


30


20

10
I I I I I I 1 I 1 I I I I I
__ __


^_ ^^


— 	

— • L —
k il A .
                               12 - 24 «m DIAMETER PARTICLES
                  8
                  i
                  o
                  o

40
30
20
10

—
—
_
—


7 i
I -
-

6.0 - 12.0 «m DIAMETER PARTICLES
40
30
20
10

—
—




,
J
I-
~~
-
-
I
                              3.0 - 6.0 urn DIAMETER PARTICLES
auu
•8 400
o
c
OT 300
g 200
O
100

__ __

	 UPSTREAM 	
k,
DOWNSTREAM 	
, n-
                   5x103
                  i
                  o
                  o
                              1.5 - 3.0 urn DIAMETER PARTICLES
                         I  I  I
                               TIME, 12sec/div

Figure 46.  Rapping puffs upper  portion  of the precipitator,
             32  minutes  between raps.  First set of puffs,
             upstream rap.  Second set of puffs, downstream rap.
                              Ill

-------
                               > 24 urn DIAMETER PARTICLES
                        z

                        o
                        u
                        ill

                        O
                        I-
                        i
                        D
                        O
                        U
                        o

                        tc

                        a.
50
40
30
20
10

-I I
—
—
—
__
A . 1 ,





j
1 1 1 1 1 1 1-
	



... 1
—
—
	

                               12-24 Mm DIAMETER PARTICLES
        JL
                       g 5x103


                       i   4



                       I   3
                       O
                       u   2
                               6 - 12 Wm DIAMETER PARTICLES
                       IT

                       a.
g 5x1 03
o
c. 4
§ 3
O
2 ^
_i
u
p 1
tc
a.
J - o Mm UlAivie i tn r«n i li^uca

UPSTREAM
~ i
-A
M
— 1 1 DOWNSTREAM —
1 L
- IV L -
                       z

                       o
                       o
                       ui
5x103




  4



  3



  2



  1
                               1.5 - 3 urn DIAMETER PARTICLES


                                            •Lr
                                   TIME, 12sec/div


Figure  47.   Rapping puffs,  lower portion of the  precipitator.

             12 minutes between raps.   First set  of puffs,  upstream

             rap.   Second set of puffs,  downstream rap.
                                    112

-------
                             24 Mm DIAMETER PARTICLES
-2
§°


en"
1-
z
3
O
o
UJ
o
H
oc
<
0.

1
o
C
w"
1-
z
3
O
o
UJ
_i
PARTIC

i
.
o
y>
r~
3
0
O
Ul
_j
y
DC
<
Q.
U
-H
O
c
to"
H-
3
O
O
UJ
o
1-
oc
<
a.
u
-2
d
c
P
z
3
O
O
UJ
^ml
O
1—
It
<
a.
3U

40


30


20

10



| |
i i i i i i i i

i^ ._
•
1
1
1
1
1
- -
L
	 •
• ^^««
n.
fl i i
A II 1 I
12 - 24 urn DIAMETER PARTICLES
50


40


30


20
10
	


—


^~


—
	
—


—


^M


—
il.r
6 - 12 urn DIAMETER PARTICLES
50

40
30


20

10


—
	


	

"
	

— —
.^


	

•>ii . . .•!.«
3 - 6 urn DIAMETER PARTICLES

400

300


200

100


•~- ~^

-
1
1
Ln^ !• _
1
— 1 —
11
, 1.5 - 3 urn DIAMETER PARTICLES

4







1



_ ^_

^^ ^^





•» — ^
y»
-urH-u-uuw-k.
                                TIME, 12sec/div

Figure 48.   Rapping puffs, upper portion of the precipitator.
             12  minutes between raps.   First set of puffs,
             upstream rap.  Second set  of puffs, downstream rap
                               113

-------
 hoppers  a  portion  rebounds, breaks  up,  and  is carried out of
 the precipitator.   At  FluiDyne  the  time separation of the
 maxima in  concentration  for the two puffs were approximately
 7 seconds  for  the  upstream rap  and  10 seconds for the down-
 stream rap.  It was also noted  that the puff from the upstream
 rap for  the  lower  portion of  the precipitator was first ob-
 served approximately 3 seconds  after the rapper was activated
 and the  downstream puff  was observed approximately 1 second
 after the  downstream rapper was activated.  The puffs in the
 upper portion  of the precipitator lasted for approximately
 4 seconds.   Multiple puffs were also observed in the upper
 sampling area  with each  puff  persisting approximately 2 seconds.
 There are  several  possible mechanisms for the secondary puffs
 in the upper region of the precipitator.  The most likely
 mechanism  is a secondary breakaway  of the dust initially dis-
 lodged by  the  rap  but  not immediately removed from the plate.
 Some  of  the  upper  secondary puffs were  observed to occur as
 much  as  17 seconds  after the  primary puff.  These delayed puffs
 may have been  due  to hopper boil-up and turbulent transport
 of the hopper  boil-up.   Examples of rapping puffs in the upper
 portion  of the precipitator are shown in Figures 46 and 48.

 The data displayed  in  Figure  43 and the rapping puffs shown in
 Figures  45 and 47  show that in  the  lower portion of the pre-
 cipitator  the  upstream rap always produced the largest contri-
 bution to  the  rapping  emissions  in  the  lower half of the pre-
 cipitator.   This phenomenon occurs  because precipitators col-
 lect  dust  in an exponential fashion.  The first plate of the
 FluiDyne precipitator  collected the thicker dust layer.  Since
 the dust surface density was  higher than in other parts of the
 precipitator and the particle size  larger, the dust was more
 easily removed from the  first plate.  Thus, on rapping, a
 larger quantity of  dust  is removed  from the first two plates and
dropped into the hoppers  than from the third plate.   Some  of this
dust rebounds and passes  through the hoppers without  being

                              114

-------
recollected.  On the downstream plate, the dust surface density
was much smaller thus did not contribute as much to rapping
emissions.  In the upper portion of the precipitator, the rela-
tive size of the rapping puffs from the upstream and downstream
rap did not depend on the amount of hopper boil-up.  When the
dust surface density in the downstream section was too small
for dust removal, the upstream rap contributed the larger
portion of the emissions in the upper portion of the precipita-
tor, as indicated by the rapping puffs shown in Figure 48,
which were obtained with a 12 minute interval between raps.
However, when the dust surface density on the downstream plates
was high enough that dust was removed, the downstream rap pro-
duced the largest quantity of dust in the upper portion, as
shown by the rapping puffs displayed in Figure 46 which were
recorded with a 32 minute interval between raps.  Although a
larger quantity of material was initially reentrained in the
inlet of the precipitator, some of this particulate was re-
precipitated.   Reprecipitation also occurs for material re-
entrained in the outlet section,  but at a much lower efficiency.
It is for this reason that outlet field raps in full scale pre-
cipitators play a major part in rapping reentrainment,  while
reentrainment from the inlet section is usually not a severe
factor in performance unless hopper boilup can sneak through
the hoppers.   The data from FluiDyne indicates that the effect
of inlet raps on outlet emissions will depend on what happens
to the hopper boilup.  Hopper boilup in the outlet section of
a full scale precipitator can play a major part in rapping
reentrainment according to these  experiments.

The total percent contribution of the raps to total emission
in each size band determined with the Royco system for dif-
ferent time intervals between raps are given in Figure 49.
The percentages obtained with the real-time system are
                           115

-------
            100
         HI
         N
         O
        
-------
comparable to the percentages calculated from the impactor
data.  The real-time system data show a decrease in the net
total emission rates as the time interval between raps is
increased, in agreement with the impactor and mass train data.
The decrease was most significant for particles with optical
diameters less than 6 ym.  As mentioned previously, this is
most likely due to agglomeration of the particles with the
thicker dust layers.  The data given in Figures 41 and 49
shows that the percent increase in the concentration of the
large particles during rapping is significantly larger than
the percentage increase for the small particles.  This occurs
because the rapping puffs emissions have a larger percentage
of large particles than the between rap emissions.  For a
high efficiency precipitator which removes nearly 100% of
the large particles entering the precipitator, rapping can
contribute up to 100% of the particle emissions for particles
with large diameters.  Total percent emissions were not de-
termined from the real time system, since comparison could
only be made for data from the same size interval without
correcting the data for size dependent probe losses, size
selective dilution and anisokinetic sampling.  Correction
factors for probe losses, size selective dilution, and aniso-
kinetic sampling were not considered accurate enough to
justify the effort, especially since the percent of total
emissions due to rapping were available from mass train and
impactor data.  However, one comparison between the real
time data and the impactor data for a between rap case was
attempted.

Expected counting rates for various particle sizes were gen-
erated from the impactor data.  The expected counting rates
included the correction factors for probe losses, size selec-
tive dilution, and anisokinetic sampling in the real time
system.  The results of the comparison are given in Table 15.
The sampling flow rate used with the real time system was ap-
proximately 3 times greater than that required for isokinetic
                             117

-------
  Table 15.  COMPARISON OF REAL-TIME SYSTEM
DATA AND IMPACTION DATA OBTAINED BETWEEN RAPS
                  Predicted
                  Count Rate
    Particle         from           Measured
    Diameter     Impactor Data     Count Rate
       ym         number/sec       number/sec
     1.5-3            4240             4500
       3-6             590              475
       6-12             12                4
      12-24              2.3              0.4
      24-48              0.83             0.2
                      118

-------
sampling.  This high flow rate was used to solve a condensa-
tion problem in the diluter.  The real time test results ob-
tained during the test with 120 minute and 150 minute intervals
between raps were confused by condensation forming in the
diluter.  Increasing the sample flow velocity raised the tem-
perature in the diluter enough that the condensation did not
occur during the other tests.

The real time particle size system also proved useful for
studying variations in between-rap outlet concentrations as a
function of precipitator currents and voltages.

CONTINUOUS MONITORS, KONITEST AND TRANSMISSOMETER
The dust concentrations at the outlet were continuously
monitored with a Lear Siegler RM41p transmissometer and with
a Konitest meter.  These instruments provided fast real time
response for determining changes in precipitator operation
during the test.  The effects of sparking and changes in
current density were immediately observed with these instru-
ments.

The amplitude of the Konitest current tended to correlate with
the amount of dust removed from the plates, as shown in Figure
50.  Unfortunately, neither the transmissometer nor the Konitest
were monitored closely enough during the test and the majority
of the rapping puffs produced signals that were off-scale on
the recording instruments.   Thus, these instruments were not
used to make quantitative measurements of rapping emissions.
An additional reason for not using the transmissometer data
is that the transmissometer output is a function of the
particle size distribution of the particulate in the view
volume and this changed when rapping puffs occurred.
                              119

-------
                0.25
                0.20
           Z
           LU
           CC
           tr
           D
           O
           111
           j-

           o
                0.15
   0.10
               0.05
                               234


                             MASS OF DUST REMOVAL, kg
Figure 50.
Correlation  of mass of dust  removed from plate  4-1
and amplitude  of Konitest meter current output  for
the upstream rapping puff for  test 5 (12 minute  rap
interval).
                           120

-------
Between-rap data obtained with the Konitest during tests 5
and 6 indicated an increase in the between-rap loading during
the tests.  However this was not observed during the previous
tests 3 and 4 with the Konitest nor during any of these tests
with the transmissometer.

DUST COLLECTION RATES ON THE PLATES AND DUST REMOVAL FROM THE
PLATES
In this section, the removal of the collected particulate from
the plates is discussed.  The data provide a background for
modelling the process of removal of the collected particulate
from the precipitator plates by rapping and for determining
optimum rapping conditions.

In the FluiDyne test facility the suspended particulate is
collected on large plates by electrostatic precipitation and
removed from the plates by rapping them with a pneumatic
rapper.  The design of the rapping system was described in an
earlier section.  In place of the pneumatic system, mechanical
drop hammers, magnetic drop hammers, or vibrators might have
been used with similar results.

For proper rapping, these systems must supply sufficient
force to remove dust from the plates and at the same time not
directly reentrain the dust into the gas stream.  Ideally, the
collected layer should be freed just enough to slide down the
precipitator plates and into the hoppers.  Even under this
ideal condition dust can still be reentrained when it falls
into the hoppers.

The rapping system at FluiDyne was successful in removing
dust from the plates of the first meter (3 ft)  of the preci-
pitator but unsuccessful in removing the collected particulate
efficiently from the last 1.8m(6 ft) of the precipitator.  The
mass of particulate collected on plates 4-1, 4-2, and 4-3 as
                             121

-------
a function time for tests with rap intervals of 12 and 32
minutes are shown in Figures 51 through 54.

From this load cell data and similar load cell data for the
various test conditions, we calculated the collection rates/
removal rates, and percentage removal rates as functions of
the time between raps.

The data given in Figure 51 shows that no dust was removed
from plate 4-1 until more than 15 kg of particulate had col-
lected on the plate.  The data also show that the residual
layer remaining after a rap increased throughout the 12-minute
interval rap test.  The data in Figures 52 and 54 showed that
very little dust was removed from the outlet plate when it
was rapped.

It was also observed that during the night about 10 kg of dust
fell off plate 4-1.  According to the load cell data, dust fell
off the plates every time the precipitator was turned off
after several hours of testing.

The collection rates determined from load cell data are dis-
played in Figure 55.  The collection rates show the expected
theoretical exponential dependence on precipitator length and
a collection rate that is independent of the time interval be-
tween raps.  A decrease in collection rate with increasing
time between raps was expected.  This was not observed.  From
data, it was estimated that 76% of the particulate was collected
in the first meter of the unit.

Particulate removal from the plates was observed to vary as a
function of the time interval between raps and as a function of
location in the precipitator.  The dependence of particulate
                              122

-------
         30
         25
         20
    O
    Q
    O
    UJ
    O
    O
    OT
         15
         10
                           O DUST FEED ON

                           D UPSTREAM RAP

                           O DOWNSTREAM RAP

                            X MASS ON PLATE AT
                             END OF PREVIOUS TEST
                             >v14 HRS BETWEEN TEST
                             PRECIPITATOR OFF
              60
               120
                                            180
                                                           240
                                   TIME, minutes
Figure 51.
Load cell  data,  mass  collected on  plate  4-1,  12 minute
interval between raps.
                                123

-------
           45
           40
°?
v
Q
Z
       UJ
       H
       Q.

       Z
       O

       O
       UJ

       O
       UJ
      O
      O
      CO
      CO
      <
           35
    15
O DUST FEED ON

O UPSTREAM RAP


O DOWNSTREAM RAP

X MASS ON PLATES AT
  END OF PREVIOUS TEST
  -v14 MRS BETWEEN TEST,
  PRECIPITATOR OFF
                 60
                                120
                                               180
                                                              240
                                   TIME, minutes
Figure 52.
       Load cell data mass collected  on plates  4-2 and 4-3,
       12  minute intervals between raps.
                                  124

-------
            35
        UJ
        a.
        Z
        O
        a
        UJ
        o
        LU
        o
        o
        oo
        CO
             30
             25
             20
15
             10
                  60
                          PLATE NO. 4-1
             O DUST FEED ON

             a UPSTREAM RAP

             O DOWNSTREAM RAP
                                         I
                                        O PRECIP. OFF
                    120             180


                        TIME, minutes
240
Figure 53.   Load  cell data, mass collected  on plate 4-1,  32 minute
              intervals between  raps.
                                125

-------
       O)


       n
<
IN
       V)
       Ul
       O
       Q
       UJ

       u
       UJ
       O
       U
           45
           40
           35
           20
           15
           10
                                              PRECIPITATOR OFF
                    O DUST FEED ON


                    D UPSTREAM RAP


                    O DOWNSTREAM RAP
                                        ,
                 60
                        120             180


                            TIME, minutes
                                                             240
Figure 54.   Load cell data,  mass  collected on plates  4-2 and  4-3,
              32 minute intervals between  raps.
                                   126

-------
u.tu

0.35

c 0.30
OJ
£ 0.25
DC
Z
° 0.20
O
LLJ
_l
_J
8 0.15
to
-o
0.10
0.05

1 | I | I | 1 | 1 | 1 | 1 | 1 |
& PLATE NO. 4-1
O PLATE NO. 4-2
D PLATE NO. 4-3
— —
A A
— A —

— —
— . —


— ' —
00 ° _
~~ D 0 0
n n
I ,D I , I , I , I , V , I , I
20 40 60 80 100 120 140 160
                     TIME INTERVAL BETWEEN RAPS, minutes
Figure 55.   Collection rate  of particulate  as  a function  of time
             interval between raps.

-------
 removal on these parameters can be explained by the effect
 that dust surface density (mass per unit ares) plays in removal
 of the dust from the plates and the effects of consolidation
 time on the cohesion characteristics of the collected dust
 layer.  In Table 16, the removal efficiencies, the percent of
 freshly collected dust that is removed by a rap, and the per-
 cent of total mass on the plate that is removed by a rap are
 given for time intervals of 12, 32, 52, and 120 minutes be-
 tween raps.  The removal efficiencies versus time between raps
 are displayed in Figure 56.   These data were computed from
 load cell data similar to that in Figures 51 through 54.  The
 data show that efficient removal (>75%) of the dust collected
 between raps occurred only when the mass per unit area collected
 between raps exceeded .8 kg/m2.  It should be noted that the
 percentage dust removed from the plates varied from rap to rap
 even  for the same conditions.  There are also  indications that
 no  removal  occurred  unless the mass per unit area on the plates
 was greater than 2.5 kg/m2.  The dust that is  not removed with
 the first rap after  it  is collected appears to develop higher
 tensile and shear strengths than the freshly collected layer.
The load cell data for plates 4-1, 4-2 and 4-3 indicated that
dust was not removed even when the mass per unit area of the
dust collected on these plates finally exceeded the levels for
which freshly collected dust was removed from plates. Raps with
insufficient intensity to remove the dust layer may have com-
pacted the dust layer making it even harder to remove.

In addition to the consolidation that appears to have occurred
during these experiments, it is suspected that the collected
layer is harder to remove from the second and third plates
due to changes in the physical properties of the layer.
                               128

-------
                     Table 16.  MASS  PER UNIT AREA ON PLATES WHEN RAPPED,
                        MASS PER UNIT AREA COLLECTED BETWEEN RAPS AND
                              MASS PER UNIT AREA REMOVED BY RAP
NJ

Rap
Interval
Test minutes Plate
3 12 4-1
4-2
4-3
4 32 4-1
4-2
4-3
7 52 4-1
4-2
4-3
2 120 4-1
4-2
4-3
Mass per Unit
Area of Dust
on Plate When
Rapped
kg/m2
2.54
3. 30
0.78
2.78
3.98
1.08
3.03
2.60
1.65
4.54
	 *
— *
Mass per Unit
Area of Dust
Collected Be-
tween Raps
kg/m2
0.40
0.08
0.05
0.82
0. 13
0.09
1. 35
0.23
0.12
2.91
1.07
0.25
Mass per Unit
Area Removed
By Raps
kg/m2
0.27
0.00
0.00
0.77
0.02
0.03
1.07
0.01
0.00
3.16
0.77
0.04
Efficiency of
Removal of
Particulate
Collected
Between Raps
%
68
00
00
94
15
33
79
4
0
109
72
16
      *Load cell zero shifted.

-------
OJ
o
      Ol
      O.
           100
            90
            80
            70
            60
            50
           40
            30
            20
            10
                  .26
                        20
                                  40
                             % OF PARTICULATE
                             COLLECTED BETWEEN RAPS
                           O % OF TOTAL PARTICULATE
                             COLLECTED ON THE PLATES
                                            60
                                                      80
                                                                 100
                                                                           120
.78
   TIME INTERVAL BETWEEN RAPS, minutes

1.3        1.8        2.3        2.9

  MASS/AREA GAINED BETWEEN RAPS, kg/m2
                                                                                     140
3.4
                                                                   160
                 Figure 56.   Dust  removal efficiency  for dust  on plate  4-1  versus
                               the time interval between raps.

-------
The cohesive strength of the collected dust layer theoretically
increases along the length of the precipitator, since the parti-
culate is fractionated into particle size distributions with
decreasing mass median diameters from the inlet to the outlet.
The feed dust at FluiDyne had a mass median diameter of 16 ym
and the dust collected in the last 20 cm length of the precipi-
tator had a mass median diameter of ^3 ym.

The simple theory of dust removal discussed in the introduction
(see equations 2 and 3) indicates that, for a given accelera-
tion perpendicular to the collection plates, the collected
layer will be removed if the mass per unit area exceeds some
minimum value.  The average acceleration perpendicular to the
plates at FluiDyne was approximately l.lxlO1* cm/sec2 (11 G) .
An estimated value for the tensile strength of the collected
dust layer is ^2x103 dynes/cm2 based on the tensile strength
tabulated in Table 1.

The application of this tensile strength to other precipitated
fly ash layers is highly uncertain, since cohesion and adhesion
of dust layers depend on many parameters, including the physi-
cal properties of the dust, the electrical conditions at the
collection plate, and the chemical composition of the gas from
which the dust is precipitated.  If we ignore the above objections
to the use of this tensile strength, a minimum mass per unit
area of 1.8 kg/m2 was calculated for dust removal (from the
ratio of the above tensile strength and plate acceleration).
This minimum dust surface density agrees nicely with the data
from the FluiDyne experiment, which indicates that 100% re-
moval of the freshly collected layer occurred at a mass per unit
area between 1 and 3 kg/m2.
                              131

-------
These load cell data show that if plate accelerations and
time intervals between raps are not properly chosen to match
the dust layer tensile strength, 100% removal of the dust
from the precipitator plates will not occur.  The data indi-
cate that larger percentages of dust were removed from the
plates as the time interval between the raps was increased.
The data also indicate that, if the residual dust layers
collected during the test were to be removed, plate accelera-
tions needed to be increased.  The amount of increase needed
was not determined, but estimates could possibly be obtained
by measuring dust removal with increased plate accelerations or
by measuring the effects of consolidation time, particle size,
and vibrational compaction on the tensile strength of the
precipitated dust layer.

 The FluiDyne experiments show the buildup of a residual dust
 layer.   Precipitator performance is often observed to degrade
 after a week or two weeks of operation after starting with clean
 plates.   This degradation is usually ascribed to the buildup
 of a steady-state layer of collected particulate on the pre-
 cipitator plates.   One solution to the residual dust layer may
 be the  use of a set of auxiliary rappers that would rap the
 plates  with enough force to remove the layers on a daily or
 weekly  basis.   An increase in plate acceleration to continually
 prevent any buildup would in many cases produce undesirable
 reentrainment characteristics.

 The data indicate that for efficient dust removal from the
 plates,  the plates should not be rapped at all until there is
 sufficient dust buildup to give a reasonable chance for
 efficient removal with the available plate acceleration.
                              132

-------
The upper limit on the time interval between  raps  is  set by
the effect the dust layer has on precipitator operation.
Obviously, if the dust layers become too thick  they interfere
with gas flow through the precipitator and inhibit performance.
The collected dust layers also affect the electrical  operation
of the precipitator.  For low resistivity dust  the electrical
effects are small but for high resistivity dust they  can be
significant.  The resistivity of the dust layer affects
precipitator operation in two ways:  (1) by setting the current
density at which the precipitator operates, if  the resistivity
is high enough that the current density is limited by back
corona to a current density below the maximum clean plate cur-
rent density, and (2) by adding a resistivity element to the
circuit.  The first effect is relatively independent of the
thickness of the dust layer and depends only  on the dust
resistivity- The second effect depends on the thickness of
the dust layer, dust resistivity, and current density.  To
prevent an excessive voltage drop in the dust layer,  the
thickness of the layer must be kept to a reasonable value.
Another electrical effect of collecting too thick a layer be-
fore rapping is a reduction of collection efficiency during
rapping due to corona quenching produced by the space charge
that develops with a large particulate load in  the gas stream.
The impactor data indicate that particulate loads range from
4.9 to 49 g/Nm3 (2 to 20 gr/dscf) during the  rapping puffs.
However, the data from these experiments also indicate that
the collected mass per unit area would need to  exceed ^4 kg/m2
for the particulate load produced by rapping  to have a large
effect on overall emissions.

This data on dust removal from precipitator plates does not supply
all the information desired for modelling dust  removal in any
full scale precipitator.  The effects of variations in dust
properties and in plate acceleration were not determined.  The

                              133

-------
 FluiDyne data  showed  substantial variations  in the removal
 efficiency as  a  function of  location in  the  precipitator and
 as  a  function  of the  time  interval between raps.  It  is sus-
 pected  that the variations will be even  larger in full scale
 precipitators.  These variations present problems for the
 modelling of rapping  reentrainment.  The data show that only
 under ideal conditions can it be assumed that all the dust
 collected at a particular  location on a  plate is subject to
 reentrainment  every time the plates are  rapped.

 COLLECTION RATES, FRACTIONAL EFFICIENCIES, AND CONTRIBUTION
 OF  RAPPING REENTRAINMENT TO TOTAL EMISSIONS
 During  this study of rapping reentrainment,  the performance
 of  the  precipitator excluding rapping reentrainment losses
 was determined by measuring the fractional efficiency between
 raps and by measuirng the collection rate of the particulate
 on  the  precipitator plates between raps.  The collection rate
 data and the fractional efficiency data  obtained between
 raps were compared with predictions made with the Southern
 Research Institute computer model of electrostatic precipita-
 tion. 1  The comparison of the collection rates measured with
 the load cells and the collection rates  calculated with the
 computer model is shown in Figure 57.  Reasonable agreement
 between measured and calculated rates was obtained.   However,
 the computer model calculation was low for the last meter of
 the precipitator length.   This discrepancy could be due either
 to  a load cell error or to an unmodelled collection effect.

 The measured fractional efficiency between raps and the
measured fractional efficiency including rapping losses
were compared with theoretical fractional efficiencies
 (Figure 58).   The theoretical fractional efficiencies were
 calculated using the Southern Research Institute computer
model, the design parameters for the FluiDyne pilot

                             134

-------
           0.30
 c

 E
UJ
oc

2
g
i-
o
111
o
o
u.
O

'E
       IT
       O
       u.
       tr

       g

       a
       o
       o
           0.25
           0.20
           0.15
    O.IO
            0.5
                                           —— THEORY, SRI Model1


                                           	EXPERIMENTAL
                                                            O

                                                            i
                       O.5        1.0        1.5        2.0



                            LOCATION INTHEPRECIPITATOR, m
                                                          2.5
Figure 57.   Collection  rate  versus horizontal location in  the

              precipitator.
                            135

-------
u>
           0.01
            0.1

         cc

         UJ
         z
         UJ
         o_
10
20
            40
            60
             0.1
                                      1.0
                                                   10
99.99





99.9







99


98



95



90
                                                                                             o
    o
    u.

80   tu





60




40
                                                                                       100
                                            PARTICLE DIAMETER, >un
                Figure 58,
                 Fractional collection  efficiency versus particle

                 diameter,  solid line theory, circles  experimental

                 between  rap,  diamonds  experimental, include rapping

                 reentrainment.

-------
precipitator, and the measured electrical conditions, volume
flow rate, and inlet particle size distribution for the
FluiDyne test.  The fractional efficiencies are compared
for the case in which nonidealities are excluded from the
computer calculations.  The shapes of the computed and
measured fractional efficiency curves did not agree either
when rapping losses were included or when they were excluded.
For the no rap condition, the computer model underpredicted
the fractional collection efficiencies for particle sizes
between .7 ym and 4.5 ym and overpredicted the fractional
collection efficiencies for particle sizes greater than 4.5 ym.
The discrepancy for the large particles has been previously
reported.23  It was speculated that the discrepancy was due
to rapping reentrainment.  The measured fractional efficiency
curve without rapping reentrainment losses indicates that
rapping reentrainment is not the only cause for the discrepancy.
A comparison of the computed and measured fractional effici-
ency curves indicates a shift in the particle size distribution
of the particulate inside the precipitator that is not accounted
for by the different migration velocities calculated for each
particle size.  A portion of the discrepancy is undoubtedly
due to reentrainment between raps.  The reentrainment between
raps of agglomerates from hoppers, collection plates, chamber
walls, corona wires, plate support structures, etc., was not
included in the computer calculations.  Sneakage of particulate
through hoppers, and corona wire and plate support areas which
permits large particles to escape collection can also cause the
discrepancy.  Coagulation of the particulate in the gas stream
may be another possibility.  Strong electric fields in electro-
static precipitators produce coagulation and increase the
efficiency of precipitation.21*   This effect is not included in
the present computer model and possibly could produce the
effects observed, explaining the load cell measurements and
the fractional efficiency curve without rapping losses.

                              137

-------
The fractional efficiency curve including rapping given in
Figure 58 is for the tests with 12 minute intervals between
raps.  The largest rapping emissions were observed during these
tests.  The fractional efficiency curve including rapping was
computed from the measured inlet feed rates in each size band
and the calculated total outlet emission rates in each size
band.  The total outlet emissions were obtained by summing out-
let emissions for each size band for the raps and for the
between rap intervals.

As expected, the inclusion of the rapping losses increased the
discrepancy between the computed fractional efficiencies and
the measured fractional efficiencies for the large particles.

Contribution of Rapping Reentrainment to Total Emissions
In the FluiDyne tests, rapping losses were observed to account
for a significant portion of the total emissions.  The per-
centage of total emissions resulting from rapping reentrain-
ment ranged from 18% to 53%, depending on the amount of dust
on the plates when they were rapped.

A plot showing the percentage contribution of the rapping puffs
to total emissions for all size bands is given in Figure 59.
This data indicates that minimum rapping losses would have
been obtained with a time interval between raps somewhere
between 52 and 150 minutes.

Since the FluiDyne test was conducted under conditions
representative of the inlet sections of many precipitators
attached to coal-fired boilers, the data indicate that im-
provements in efficiency can be obtained by increasing the
time interval between raps.   Many field units operate with
time intervals between raps for inlet sections on the order
of 2 to 6 minutes; the FluiDyne data indicate that these times

                             138

-------
U)
               LU



               Q

               l/J
               UJ 5
               II
               u_ ill
               O UJ
                     50
40
                     30
                     20
                     10
                               20
                                                 I	I
                                             I	I
                                       40       60        80       100       120




                                               TIME INTERVAL BETWEEN RAPS, min
                                                                                  140
                                                                                           160
                  Figure 59.   Percentage of total emission  due to  rapping  reentrainment.

-------
should be increased by a factor of 6 to 15.  The exact factor
depends on many parameters/ including the effect on between-
rap operation.

EFFECT OF RAPPING REENTRAINMENT ON EFFICIENCY, PENETRATION,
AND SIZING OF ELECTROSTATIC PRECIPITATORS
Generally, if rapping reentrainment were entirely prevented
the percentage increase in efficiency would be small, but the
percentage decrease in penetration would be large.  Examples
using the percentages obtained during the FluiDyne experiments
are given in Tables 17 and 18.  These examples also show that
the percentage increase in efficiency decreases with increas-
ing efficiency of the precipitator.  Due to the exponential
dependence of precipitator efficiency on particle migration
velocity, plate collection area, and gas volume flow rate, the
effect of rapping reentrainment on the size of a precipitator
required for a given efficiency is substantial.  Estimates of
the effect of rapping reentrainment on precipitator size are
also tabulated in Tables 17 and 18 for various conditions.
Two approaches were used to generate the tabulated values.
One approach was to assume that the percentage penetration due
to rapping was independent of the size of the unit.  This
assumption, although not strictly true, may fairly represent
real situations.  According to the data obtained during the
experiments at FluiDyne, the rapping puff concentration has a
significant vertical gradient and the particle size distri-
butation of the rapping puff consists of large particles.
Thus, in a multi-section precipitator the rapping puffs from
upstream sections may be recollected in the lower portion of
a succeeding section and later transferred from this lower
position into the hoppers without making any significant con-
tibution to outlet emissions.  In this case, outlet fields
would contribute the largest portion of the rapping emission
                              140

-------
                                          Table 17.  EFFECT OF RAPPING  REENTRAINMENT ON  EFFICIENCY  FOR RAPPING LOSSES
                                       ACCOUNTING FOR 20%, 30%, 40% AND 50% OF EMISSIONS FOR VARIOUS OVERALL EFFICIENCIES
                                                                        (first approach)
SCA* required
Assumed Penetration
percent of Due to
total pene- rapping
tration due reentrain-
to rapping ment,
reentrainment %
3.75
2.25
50 0.75
0.25
0.05
3.00
1.80
40 0.60
0.20
0.04
2.25
1.35
30 0.45
0.15
0.03
1.50
0.90
20 0.30
0.40
0.08
Without
rapping
reentrain-
ment. Total
3.75
2.25
0.75
0.25
0.05
4.50
2.70
0.90
0.60
0.06
5.25
3.15
1.05
0.35
0.07
6.00
4.60
1.20
0.10
0.02
7.50
4.50
1.50
0.50
0.10
7.50
4.50
1.50
0.50
0.10
7.50
4.50
1.50
0.50
0.10
7.50
4.50
1.50
0.50
0.10
Assumed
Overall
Efficiency
n
92.50
95.50
98.50
99.50
99.90
92.50
95.50
98.50
99.50
99.90
92.50
95.50
98.50
99.50
99.90
92.50
95.50
98.50
99.50
99.90
Efficiency
without
rapping re-
entrainment ,
riK, %
96.25
97.75
99.25
99.75
99.95
95.50
97.30
99.10
99.70
99.94
94.75
96.85
98.95
99.65
99.93
94.00
96.40
98.80
99.90
99.92
For n without
Increase in rapping re-
r| without entrainment,
rapping re- (ft2/
entrainment, m2/ 1000 ft3/
% m3/sec sec)
3.90
2.31
0.76
0.25
0.05
3.24
1.88
0.61
0.20
0.04
2.43
1.41
0.46
0.15
0.03
1.62
0.94
0.30
0.10
0.02


26.2
41.8
68.9


26.2
41.8
68.9


26.2
41.8
68.9


26.2
41.8
68.9


(133)
(212)
(350)


(133)
(212)
(350)


(133)
(212)
(350)


- (133)
(212)
(350)
For n,R or for n Additional SCA
with rapping required because
reentrainment, of reentrainment Addi-
(ft2/ (ft2/ tional
' m2/ 1000 ft3/ m2/ 1000 ft3/ SCA,
m3/sec sec) m3/sec sec) %
16.2
21.3
35.5
53.2
81.7

18.7
32.9
50.2
78.8


30.5
47.3
75.8


29.2
68.9
73.3
( 82)
(108)
(180)
(270)
(415)

( 95)
(167)
(255)
(400)


(155)
(240)
(385)


(148)
(350)
(372)


9.3
11.4
12.8


6.7
8.5
9.8

•••
4.5
5.5
6.9


3.0
3.5
4.3


(47)
(58)
(65)


(34)
(43)
(50)


(23)
(28)
(35)


(15)
(18)
(22)


353
27.4
18.6


25.6
20.3
14.3


16.5
13.2
10.0


11.3
8.5
6.3
•Specific Collection Area (SCA)  based on data given by Gooch, et al.
   (Figure 26 computed performance curves at
nA/cm2).

-------
                                        Table 18.   EFFECT OF RAPPING REENTRAINMENT ON PERFORMANCE  FOR VARIOUS REENTRAINMENT
                                  PERCENTAGES AS A FUNCTION OF NUMBER OF SECTIONS AND A CONSTANT COLLECTION EFFICIENCY PER SECTION.
Assumed
percent of
collected
material
per sec- Number
tion re- of
entrained Sections
1
5 9 2
3
4
1
2 3 2
3
4
1
0.97 2
3
4
SCA* Needed
Penetration
Due to
rapping
reentrain-
ment ,
%
6.2
1.02
0.136
0.0193
2.23
0.28
0.027
0.0023
0.90
0.10
0.009
0.0007
Without
rapping

reentrain-
ment.
%
5.2
0.27
0.014
0.0007
5.20
0.27
0.014
0.0007
5.20
0.27
0.014
0.0007
Total
%
11.4
1.29
0.150
0.020
7.43
0.55
0.041
0.0030
6.1
0.37
0.023
0.0014
Percent of
penetration
due to rap-
ping reen-
trainment.
%
52
79
91
97
30
51
66
77
15
27
39
50


Efficiency
n.
%
88.6
98.7
99.85
99.98
92.57
99.45
99.959
99.997
93.9
99.63
99.97
99.9986
Assumed
efficiency
HR without
rapping re-
entrainment.
%
94.8
99.73
99.986
99.9993
94.80
99.73
99.986
99.9993
94.8
99.73
99.986
99.9993
Increase
Without rapping
in n with reentrainment to
no rapping obtain n
reentrain-
ment.
%
6.54
1.03
0.14
0.02
2.35
0.28
0.03
0.002
0.95
0.10
O.O16
0.001
m2/
m3/sec

28.0
62.1
98.5

40.4
83.7
108.5

46.3
91.0
108.3
(ft2/
1000 ft3/
sec)

(142)
(315)
(500)

(205)
(425)
(550)

(235)
(462)
(550)
With
rapping
reentrainment

m2/
m3/sec

51.6
105.4
108.3

51.6
105.4
108.3

51.6
105.4
108.3
(ft2/
1000 ft3/
sec)

(262)
(535)
(550)

(262)
(535)
(550)

(262)
(535)
(550)


Additional SCA

mV
ra3/sec

23.6
43.3


11.2
21.7


5.3
14.4

(ft2/
1000 ft3/
sec)

(120)
(220)


( 57)
(110)


( 27)
( 73)


Addi-
tional
SCA,
%

84
69


28
25


11
16

*Specific Collection Area (SCA) based on data given by Gooch, et al.
   (Figure 26 computed performance curves at 20 nA/cm).

-------
and the percent emission due to rapping should be independent
of the size of the unit.  Estimates based on this approach are
tabulated in table 17.

                                                             1 8
The second procedure was based on the development by Francis,
which was discussed in the introduction.  This approach assumes
that the fraction of collected dust per section reentrained
and emitted is the same for every section, that the between-
rap sectional efficiency is constant from one section to
another, and that the recollection efficiency for the rapping
puffs is the same as the between rap sectional efficiency.
According to Francis1 procedure, the penetration of any down-
stream section is equal to the penetration of the first
section (Pi) raised to the power of the number of preceding
sections plus one.  Thus, the penetration of the N section (PN)
is given by:

          PN = (Pi)N                                       (8)

Using this equation, the efficiencies of a precipitator having
1, 2, 3, or 4 sections with a between-rap sectional efficiency
of 94.8% were calculated for percent penetrations of the first
section due to rapping reentrainments of 52%, 30%, and 15%.
These percentages correspond to 5.9%, 2.3%, and 0.97% of the
collected material per section being reentrained and emitted
from a section.  These values were chosen to correspond with
those obtained during the FluiDyne experiments.   The results
of the calculations based on this procedure are tabulated in
Table 18.   According to this second procedure for the con-
ditions corresponding to the worst rapping reentrainment case
at FluiDyne, the contribution of rapping reentrainment to
total emissions would be greater than 90% for a three or four
section precipitator with 94.8% sectional efficiency.  For
                              143

-------
a case in which reentrainment accounts for only 15% of the
emissions for the first section, a situation slightly better
than the best obtained during the FluiDyne experiments, rapping
reentrainment according to the Francis procedure would account
for 39% of emissions for a 3-section precipitator and 50% of
emissions for 4-section precipitator, if the precipitator had
a sectional efficiency of 94.8%.  These calculations indicate
that, as the size of the precipitator increases, the percentage
contribution of rapping reentrainment to total emissions in-
creases.  As previously discussed, this may not happen.  Further
investigations with a multi-section precipitator are needed.

To obtain an indication of the effect of rapping reentrainment
on the sizing of a precipitator, estimates of the theoretical
specific collection area needed to obtain a given collection
efficiency for a particular set of conditions were obtained
from the model developed by Gooch, et al.1  For the example
case, a typical particle size distribution for an ash from a
coal-fired boiler plant was selected and a precipitator operat-
ing current density of 20 nA/cm2 was assumed (see the report
of Gooch, et al,1 for further details).

The theoretical specific collection area (SCA)  required to
obtain a desired collection efficiency with rapping reentrain-
ment was compared with a theoretical SCA that would give the
same efficiency if there were no rapping emissions.  This gave an
estimate of the increase in SCA needed to counterbalance the
rapping reentrainment emissions.  The theoretical SCA needed
to obtain the desired efficiency with rapping reentrainment
was determined by first computing the efficiency that would be
obtained if there were no rapping reentrainment, by assuming
that either the percentage of emissions due to rapping re-
entrainment or the fraction of collected dust reentrained and
emitted from a section due to rapping were known.   Once this

                              144

-------
efficiency was calculated and assuming that there were no
other nonidealities beside rapping reentrainment, the required
theoretical SCA was obtained from data given by Gooch, et al.
Then, again using the data given by Gooch, et al, the SCA was
determined that would give the desired efficiency if there were
no nonidealities such as rapping reentrainment.  For the parti-
cular cases considered, the increase in required SCA due to
the rapping reentrainment ranged from 6% to 84%  (see Tables 17
and 18).  Thus, although the percentage decrease in efficiency
due to rapping reentrainment is relatively small, the effect
on the size required for a given efficiency can be significant.
The data obtained from the FluiDyne experiments indicates that
even moderate changes in rapping parameters can significantly
affect the size of a precipitator for a given efficiency.

EFFECT OF AGGLOMERATION AND FRACTIONAL REENTRAINMENT PERCENT-
AGES OF COLLECTED PARTICIPATE FOR VARIOUS PARTICLE SIZES
Modern electrostatic precipitator technology is concerned not
only with the overall collection efficiency but also with the
fractional collection efficiency,that is, the collection ef-
ficiency for various particle size ranges.  The fraction of
collected material that is emitted due to rapping reentrain-
ment for various particle sizes is in the over simplified case
independent of particle size.

For a homogeneous mixture of discrete particles of various
sizes if one quarters out from the mixture a sample, theoreti-
cally the percentage of particles removed from the mixture
would be the same for all sizes.  Rapping reentrainment does
not produce this simplified case.  The particle size distribu-
tion of the discrete particles reentrained is not the same as
the particle size distribution of the individual particles
collected on the plates.  Many of the discrete particles that
are reentrained are agglomerates of the individual particles

                               145

-------
              Appendix I
Log of Precipitator Data,  June 23, 1975

               Test 1
Power Supply
Time
1040
1050
1100
1110
1120
1130
1140
1150
1200
1210
1220
1230
1240
1250
1300
1310
Water,
gal/min
4.5
4.5
4.5
4.5
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
Dust off
Avg . duct
Temp . @
inlet, °F
283
283
278
278
278
278
278
283
283
283
283
283
288
288
293
and on
Secondary
voltage
kV
36
38
39
39
39
39
39
39
39
40
38
38
39
39
39

Secondary
current
mA
43
43
43
43
42
42
42
42
42
42
24
24
32
32
32

                 152

-------
               Appendix I
Log of Precipitator Data, June 23,  1975

                Test 2
Power Supply
Time
1320
1330
1440
1450
1500
1510
1520
1530
1540
1600
1610
1620
1630
1640
1650
1700
1710
Water,
gal/min
4.6
4.6
4.6
4.6
4.6
4.6
4.1
5.0
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
Avg. duct
Temp . @
inlet, °F
293
293
303
308
308
308
308
293
278
273
273
273
273
263
263
263
263
Secondary
voltage
kV
38
37
37
37
37
37
37
40
40
40
40
38
38
39
39
42
42
Secondary
current
mA
23
24
24
24
24
24
24
34
34
34
34
16
16
17
17
32
32
                   153

-------
              Appendix I
Log of Precipitator Data, June. 24, 1975
                Test 3
Power Supply
Time
1110
1120
1130
1140
1150
1200
1210
1220
1230
1240
1250
1300
1310
1320
1330
Water ,
gal/min
6.0
6.0
6.0
6.0
6.0
5.8
Dust off
Dust on
5.8
Dust Off
5.8
Dust off
5.8
Dust off
5.8
Avg. duct
Temp . @
inlet, °F
250
250
250
258
258
258


263
and on
263
and on
263
and on
266
Secondary
voltage
kV
38
40
41
41
41
41


42

42

42

42
Secondary
current
mA
40
40
40
40
40
40


40

40

40

40
                  154

-------
              Appendix I
Log of Precipitator Data, June 24, 1975

                Test 4
Power Supply
Time
1500
1510
1520
1530
1540
1550
1600
1610
1620
1630
1640
1650
1700
Water,
gal/min
5.8
5.8
5.8
5.8
Dust off
5.8
5.8
5.8
Dust off
5.8
5.8
6.0
Dust off
Avg. duct
Temp . °
inlet, °F
258
263
263
263
and on
263
266
266
and on
266
266
260

Secondary
voltage
kV
42
44
43
43

43
43
43

43
43
44

Secondary
current
mA
31
41
40
41

40
40
39

40
40
40

                  155

-------
              Appendix I
Log of Precipitator Data, June 25, 1975
                Test 5
Power Supply
Time
0950
1000
1010
1020
1030
1040
1050
1100
1110
1120
1130
1140
1150
Water,
gal/min
5.7
5.8
5.8
5.8
5.8
5.5
5.5
5.7
5.7
5.7
5.7
5.7
5.7
Avg . duct
Temp . @
inlet, °F
256
256
263
263
263
268
268
268
258
258
258
258
262
Secondary
voltage
kV
40
40
40
40
40
40
41
41
42
42
42
40
42
Secondary
current
mA
40
40
40
40
40
40
40
40
40
40
40
40
40
                  156

-------
              Appendix I
Log of Precipitator Data,  June 25,  1975

                Test 6
Power Supply
Time
1200
1210
1220
1350
1400
1410
1420
1430
1440
1450
1500
1510
1520
1530
1540
1550
1600
1610
1620
1630
1640
1650
1700
Water,
gal/min
5.7
5.7
5.7
5.7
5.8
5.9
5.9
6.0
6.0
6.0
6.0
6.0
6.0
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.9
Avg. duct
Temp . @
inlet, °F
262
262
262
274
264
264
264
264
264
264
264
264
264
265
266
266
266
266
267
268
268
269
269
Secondary
voltage
kV
42
40
40
43
43
43
43
44
44
42
42
43
44
44
44
44
44
44
44
44
44
44
44
Secondary
current
mA
40
40
40
40
40
40
40
40
40
36
39
37
40
38
40
40
40
40
38
38
38
38
38
                 157

-------
              Appendix I
Log of Precipitator Data,  June 26,  1975

                Test 7
Power Supply
Time
0950
1000
1010
1020
1030
1040
1050
1100
1110
1120
1130
1140
1150
1200
1210
Water,
gal/min
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.9
5.9
6.0
6.0
6.0
6.0
6.0
6.0
Avg. duct
Temp . @
inlet, °F
257
257
257
258
258
258
264
264
267
267
267
261
269
269
269
Secondary
voltage
kV


39
40
40
40
41
41
41
41
42
42
42
43
41
Secondary
current
mA


40
44
44
41
41
41
41
41
40
40
40
41
41
                  158

-------
              Appendix I
Log of Precipitator Data, June 26, 1975

                Test 8
Power Supply
Time
1220
1230
1240
1250
1300
1310
1420
1430
1440
1450
1500
1510
1520
1530
1540
1550
1600
1610
1620
1630
1640
1650
1700
1710
1720
1730
Water,
gal/min
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
6.1
Avg. duct
Temp . @
inlet, °F
269
269
269
269
269
269
264
265
265
265
265
267
267
267
264
264
264
264
264
264
264
264
264
264
264
Secondary
voltage
kV
43
43
43
43
43
43
43
43
43
43
43
43
44
44
44
44
44
44
45
45
45
43
44
45
45
Secondary
current
mA
40
40
41
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
Test stopped
                  159

-------
Precipitator Ash Loading Data
        Tests 1 and 2
Time
Test 1 1050
1108
1126
1139
1151
1155
1210
1230
1247
1300
1310
Test 2 1333
1440
1500
1525
1547
1608
1625
1645
1700
1715
Actual
Feed
Rate
(Ibs) (Ibs/hr) Comments
930 348
850
760
700
900
875
800
675
560
500
440 Dust off (13:04-15:25) -
44Q reload hopper
1300
1300
1300
1200 407
1075
950
825
725
640 Dust off
            160

-------
Precipitator Ash Loading Data
        Tests 3 and 4
Time
Test 3 1116
1133
1144
1155
1203
1216
1232
1250
1308
1327
Test 4 1500
1514
1527
1546
1600
1615
1627
1649
1658
(Ibs)
1300
1300
1280
1225
1160
1140
1075
1000
925
875
1175
1090
1010
950
875
760
740
600

Actual
Feed
Rate
(Ibs/hr)
332






338
335
369
342


380


358




Dust


Dust

Dust
Dust
Dust
Dust
Dust


Dust


Dust

Dust


on


off

off
off
off
off
off


off


off

off
Comments




(12:05-12

(12:25-12
(12:42-12
(13:01-13
(13:18-13
(13:37-14


(15:37-15


(16:17-16







:10)

:30)
:48)
:07)
:25)
:56)


:44)


:24)


             161

-------
Precipitator Ash Loading Data
          Test 5
Time
Test 5 0951
1005
1014
1025
1039
1043
1051
1104
1113
1122
1126
1134
1143
1148
1155
1207
(Ibs)
1225
1130
1080
1035
970
925
925
860
860
775
750
750
700
680
680
620
Actual
Feed
Rate
(Ibs/hr)
358



352

340

326


327


400

Comments




Dust off (10:25-10:31)
Dust off (10:43-10:51)
Dust off (11:04-11:13)
Dust off (11:26-11:34)
Dust off (11:48-11:55)
Fill hopper





Dust off
             162

-------
Precipitator Ash Loading Data
           Test 6
Time
Test 6 1356
1407
1414
1423
1432
1442
1446
1455
1504
1527
1533
1542
1549
1559
1607
1620
1626
1634
1642
1648
1655
1700
(Ibs)
1200
1150
1100
1050
990
930
920
920
870
720
720
670
630
560
520
520
480
430
380
350
310
270
Actual
Feed
Rate
(Ibs/hr) Comments
362 Fill hopper
Dust off (12:07-13:56)





Dust off (14:46-14:55)
357

Dust off (15:27-15:33)
358



Dust off (16:07-16:20)
360





Dust off
             163

-------
Precipitator Ash Loading Data
           Test 7
Time
Test 7 1009
1027
1048
1657
1127
1151
1200
1208
1224
1238
1246
1257
1308
(Ibs)
1100
980
870
810
660
520
470
420
360
280
230
170
100
Actual
Feed
Rate
(Ibs/hr)
341



350 Dust


Dust
356




Comments




off (11:11-11:15)


off (12:08-12:15)





              164

-------
Precipitator Ash Loading Data
           Test 8
Time
Test 8 1426
1447
1457
1506
1515
1523
1535
1544
1552
1613
1617
1624
1634
1643
1651
1707
1716
1724
1730
Actual
Feed
Rate
(Ibs) (Ibs/hr) Comments
1250 382 Fill hopper
Dust off (13:08-14:26)
1120
1060
1000
950
900
820
760
700
590 420 Dust off (15:52-16:00
560
520
450
400
340
280 350 Dust off (16:51-16:59)
220
160
Dust off
             165

-------
 TABLE OF CONVERSION FACTORS
            MASS
   1 grain  =  0.0648 grams
   1 gram   = 15.4324 grains

            AREA
      1 ft2 =  0.0929 m2
      1m2  =10.76 ft2

           VOLUME
      1 ft3 =  0.0283 m3
      1 m3  = 35.31 ft3

          PRESSURE
      1 kPa  = 0.1451 Psi
      1 Pa   =1 N/m2
1 Psi  =  6.894 kPa  (kilopascal)
        ACCELERATION
      1 G = 9.8 m/sec2

          VELOCITY
   1 m/sec  = 0.3048 ft/sec
   1 ft/sec = 3.28 m/sec
             166

-------
                                 TECHNICAL REPORT DATA
                          (Please read Inunctions on the reverse before completing)
 1. REPORT NO.
  EPA-600/2-76-140
                                                       3. RECIPIENT'S ACCESSION-NO.
 4. TITLE AND SUBTITLE
 Rapping Reentrainment in a Nearly Full-Scale
 Pilot Electrostatic Precipitator
            5. REPORT DATE
             May 1976
            6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)

 Herbert W.  Spencer, III
                                                       8. PERFORMING ORGANIZATION REPORT NO.
             SORI-EAS-76-061
             3489-V
 9. PERFORMING OROANIEATION NAME AND ADDRESS
 Southern Research Institute
 2000 Ninth Avenue South
 Birmingham, Alabama 35205
                                                       10. PROGRAM ELEMENT NO.
             1AB012; ROAP 21ADL-027
             11. CONTRACT/GRANT NO.
             68-02-1875
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
             13. TYPE OF REPORT AND PERIOD COVERED
             Final;  3/75-1/76	
             14. SPONSORING AGENCY CODE
             EPA-ORD
 15. SUPPLEMENTARY NOTES ffiRL-RTP Project Officer for this report is Leslie E. Sparks,
 Mail Drop 61, Ext 2925.
 16. ABSTRACT
           The report gives results of an experimental investigation of rapping
 reentrainment in a nearly full-scale pilot electrostatic precipitator.  The study
 included a fundamental examination of the mechanics of removal of dry dust by
 rapping and the quantification of rapping reentrainment in terms of the percentage
 and the particle size distribution of the reentrained dust.  During the study, the
 contribution of rapping reentrainment to total emissions ranged from 53 to 18%,
 depending on rapping conditions.   These percentages corresponded to 5. 4 and 2. 7%
 of the dust collected on the plates being emitted from the precipitator during plate
           A major portion of the  reentrained material was observed to result from
 hopper 'boil-up.'  A decrease in  rapping emissions was obtained by increasing the
 time interval between raps.  The  mass median diameters of the size distributions
 of the particles emitted during the raps were observed to  increase with increased
 time between raps.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                         c.  COSATI Field/Group
 Air Pollution
 Electrostatic Precipitators
 Dust
Air Pollution Control
Stationary Sources
Rapping
Reentrainment
Particulate
13B

11G
 8. DISTRIBUTION STATEMENT

 Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
     178
                                          20. SECURITY CLASS (This page)
                                           Unclassified
                                                                   22. PRICE
EPA Form 2220-1 (9-73)
                                        167

-------