U.S. Environmental Protection Agency Industrial Environmental Research      FPA-fiOO/7-77-112
Office of Research and Development Laboratory
                Research Triangle Park, North Carolina 27711 November 1977
        DEVELOPMENT OF A HIGH-
        TEMPERATURE/HIGH-PRESSURE
        ELECTROSTATIC PRECIPITATOR
        Interagency
        Energy-Environment
        Research and Development
        Program Report

-------
                       RESEARCH REPORTING SERIES
 Research reports of the Office of Research and Development, U.S.
 Environmental Protection Agency, have been grouped  into  seven series.
 These seven 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 seven series
 are:

     1.  Environmental Health Effects Research
     2.  Environmental Protection Technology
     3.  Ecological Research
     4.  Environmental Monitoring
     5.  Socioeconomic Environmental  Studies
     6.  Scientific and Technical Assessment  Reports  (STAR)
     7.  Interagency Energy-Environment Research and  Development

 This report has been assigned to the  INTERAGENCY ENERGY-ENVIRONMENT
 RESEARCH AND DEVELOPMENT series.  Reports  in  this series result from
 the effort funded under the 17-agency Federal Energy/Environment
 Research and Development Program.  These studies relate  to EPA's
 mission to protect the public health  and welfare from adverse effects
 of pollutants associated with energy  systems.   The goal  of the Program
 is to assure the rapid development of domestic energy supplies in an
 environmentally—compatible manner by providing the necessary
 environmental data and control technology.  Investigations include
 analyses of the transport of energy-related pollutants and their health
 and ecological effects; assessments of, and development  of, control
 technologies for energy systems; and  integrated assessments of a wide
 range of energy-related environmental issues.

                            REVIEW NOTICE

 This report has been reviewed by the participating Federal
Agencies, and approved for publication. Approval does not
 signify that the contents necessarily reflect the views and
 policies  of the Government, nor does mention of trade names
 or commercial products constitute endorsement or recommen-
 dation for use.
This document is available to the public through the  National Technical
Information Service, Springfield, Virginia  22161.

-------
                                        EPA-600/7-77-132

                                          November 1977
            DEVELOPMENT OF A
HIGH-TEMPERATURE/HIGH-PRESSURE
    ELECTROSTATIC  PRECIPITATOR
                          by

                  J.R. Bush, P.L Feldman, and M. Robinson

                    Cottrell Environmental Systems
                      Research-Cottrell, Inc.
                   Bound Brook, New Jersey 08805
                     Contract No. 68-02-2104
                    Program Element No. EHE623A
                   EPA Project Officer Leslie E. Sparks

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

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

-------
                             ABSTRACT
     The feasibility of electrostatic precipitation at tempera-
tures and pressures varying from room conditions to 1366 K and
3550 kPa, respectively, has been demonstrated in a laboratory
wire-pipe electrode system.  Subject to appropriate choices of
polarity and pressure, stable corona discharges are obtained at
all temperatures.  Current-voltage characteristics for both
polarities are reported for dry air, a simulated combustion
gas, and a substitute fuel gas.  The effects of temperature,
pressure, and electrode geometry on sparkover voltage, corona-
starting voltage and current are evaluated.  It is recommended
that this study be extended to pilot coal gasifiers and fluidized-
bed combustors to obtain detailed design data, and permit
determination of collection efficiencies and precipitate
characteristics under actual operating conditions.  A pre-
liminary cost estimate is presented for a precipitator intended
for use in connection with a commercial fluidized-bed combustor
and turbine system.
                                11

-------
                         EXECUTIVE SUMMARY


     At the elevated temperatures and pressures characteristic
of many advanced coal-conversion processes, electrostatic
precipitators exhibit potentially higher particle-collecting
efficiencies than prevail under conventional conditions of
precipitator application.  High efficiencies are an anticipated
consequence of the increased voltages that can be applied to the
precipitator over broad pressure/temperature ranges of practical
interest.  This efficiency effect is unique for precipitators.
In other particulate collection devices, efficiency generally
decreases with increasing temperature and pressure.  But even
for precipitators/ it must be remembered that improved performance
is not an unvarying consequence of high pressure and temperature.
Beneficial results depend on suitably combining the two variables.

     Current-voltage characteristics and sparkover data for
both positive and negative polarities are reported here for
three gas mixtures:  dry air, a simulated flue gas, and a substitute
(noncombustible) fuel gas.

     In the present study, no practical upper limit is reached
for stable corona at temperatures and pressures as high as 1366
K (2000 F) and 3550 kPa  (515 psia).  In some instances, however,
limits imposed by the critical pressure are approached.  The
critical pressure is the lowest elevated pressure at which
corona starting and sparkover voltages coincide and beyond which
a stable corona is not possible.

     Increasing the temperature raises the critical pressure and
so results in a broader operating range of pressure.  The negative-
corona critical pressure, attained here only at room temperature,
is found to be significantly higher than the positive value.  In
other words, the pressure range over which negative corona is
possible is greater than the range for positive corona.

     Positive sparkover voltage exceeds negative for conditions
of low gas density and high temperature.  For high density,
negative sparkover is higher than positive sparkover.


                             iii

-------
     The amplitude of discharge-electrode oscillations is found
to be dependent on electrode length, shape, tension and material.
Short, rigid or high stressed electrodes produced minimal oscil-
lation.  The rigors of high-temperature service require discharge
and collecting electrodes of high tensile strength and high
oxidation and carburization resistance.  Oxidation and carburization
result in an expanding scale growth that effects both corona-
starting and sparkover voltages.  Inconel 601 performs satisfac-
torily at temperatures at least to 1366 K (2000°F).

     A preliminary cost estimate is presented for a precipitator
intended for use in connection with a commercial fluidized-bed
combustor and turbine system.

     In view of the encouraging results obtained so far—but
which are limited to a particle-free system—it is recommended
that we next proceed to the design and operation of a pilot
precipitator treating fluidized-bed combustor or gasifier ef-
fluents.
                              IV

-------
                             CONTENTS
Abstract

Executive Summary

Figures

Tables

Nomenclature                                                    x

Acknowledgments                                                 Xil

1.   Introduction                                               *•
          Objectives and Scope                                  1
          Fundamentals of High Temperature/Pressure
            Precipitation                                       2
               Overview                                         2
               Critical Pressure                                ^
               Thermal lonization                               4
               Earlier Studies                                  6

2.   Conclusions                                                 '

3    Recommendations                                            ^

4.   Experimental Apparatus and Procedure                       ^
          Test Precipitator                                     ^
          High-Voltage Power Supply                             ^2
          Temperature Control System                            12
          Gas-Flow System                                       ^-^
          Test Conditions and Procedure                         2^-
          Discharge-Electrode Oscillation Test                  22

5.   Results and Discussion                                     25
          The Experimental Data                                 2^
          Temperature/Pressure Limitations on Corona  Stability  25
               General Remarks                                  2^
               Excessive Current
               Positive Critical Pressure                       2^
               Negative Critical Pressure                       27
               Practical Precipitator-Tube  Diameter            27
               Polarity Effects                                 28

-------
               Background Considerations                    28
               Experimental Observations in Air             29
               Experimental Observations in Combustion
                 and Fuel Gases                             29
          Corona-Starting Voltage                           30
               Data Fluctuations                            30
               Peek's Equation                              30
               Current-Voltage Elbows                       31
          Discharge-Electrode Oscillations                  32
          Material Evaluation                               32

6.   Commercial Cost Estimate                               49
          Precipitator Size for Fluidized-Bed Combustion    49
          Material-Temperature Dependence                   52

7.   References                                             57

8.   Appendices                                             60

          A.   Calculation of Thermal Conductivity          60
          B.   Calculation of Ionic Mobility                62
          C.   Determination of Gas-Mixture
               Composition                                  63
          D.   Components of Power-Supply Circuit           65
          E.   List of Experimental Precipitator fDtawings   66
                                 VI

-------
                            FIGURES
Number                                                  Page


  1     Laboratory Precipitator and Pressure Vessel      H

  2     Power Supply and  Instrumentation                13
        Circuitry

  3     Three Zone Precipitator Heater and Control       14
        System

  4     Internal Tube and Gas Temperature Calibra-       ig
        tion to Control System

  5     Pressurizing and Mixing Flow System to Pre-      17
        cipitator

  6     Humidifier Internal Details                      ig

  7     Overall Test and Control Room Layout
        Schematic                                        19

  8     Test Apparatus for Electrode Oscillation         23
        Studies

  9     Discharge Electrode Geometries for Oscilla-      24
        tion Studies

 10     Current-Voltage Curves in Air at 294 Kf
        533 Kf and 950 K            t                     36

 11     Current-Voltage curves in Air at 811 K           37

 12     Current-Voltage Curves in Air at 1089 K          33

 13     Current-Voltage Curves in Air at 1366 K          39

 14     Current-Voltage Curves for a Simulated
        Combustion Gas Mixture                           40

 15     Current-Voltage Curves for a Substitute
        Fuel Gas Mixture                                 41

 16     Sparking and Corona Starting Voltages in
        Air at 294 K and 533 K                           42
                                VII

-------
17     Sparking and Corona Starting Voltages in         43
       Air at 811 K and 950 K

18     Sparking and Corona Starting Voltages in         44
       Air at 1089 K and 1366 K

19     Sparking and Corona Starting Voltages for        45
       a Simulated Combustion Gas

20     Sparking and Corona Starting Voltages for        46
       a Substitute Fuel Gas

21     Negative Corona Starting Fields in Air           47
       Compared with Peek's Equation

22     Positive Corona Starting Fields in Air           48
       Compared with Peek's Equation
                              Vlll

-------
                              TABLES

                                                       Page

1.   Test Gas Composition                               15

2.  Thermal Conductivity and Mobility of Test Gases     20

3.  Experimental Test Variables                         21

4.  Discharge Electrode Geometries                      22

                                                        34
5.  Discharge Electrode Oscillation Results

6.  Determination of Gas Flow Requirements for
    a Fluidized Bed Combustor Plant                     54

7.  Design Alternatives for a Pressurized Preci-
    oitator                                             54

8.  Cost Estimation for Two Precipitator Designs        55

A-l  Calculated Thermal Conductivity of Test
     Gas Mixture                                        61

A-2  Calculated Thermal Conductivity of a Low
     Btu Fuel Gas                                       61

B-l  Calculated Ionic Mobilities of Gas Mixtures        62

C-l  Combustion Gas Mixture Composition                 63.

C-2  Substitute Fuel Gas Mixture Composition            64

D    Components of Power-Supply Circuit                 65

E    List of Experimental Precipitator Drawings         66
                               ix

-------
                           NOMENCLATURE





b         negative ion mobility at standard conditions

 0         (m2/(sec-V))



b         positive ion   mobility of qas at standard conditions

 0         (m2/(sec-V))



C.         capicitator



CR.       silicon rectifier



D         inside diameter of tube  (m)



d         outer diameter of wire  (m)



E         electric field at wire surface  (V/m)



E         corona-starting field at wire surface  (V/m)



i    '     current  (A)



j         linear current density  (A/m)



k         thermal conductivity of gas  (W/m-K))



L.         inductor



M.         molecular weight of species  i



m         surface roughness factor  (dimensionless)



P         pressure


p
 o         standard pressure


R.         resistor



r.         wire radius  (m)



r         tube radius  (m)
 o


T         temperature  (K)




                                x

-------
T         standard temperature  (294 K)
TC        thermocouple sensor
V         voltage (V)
V         corona-starting voltage  (V)
 c
w         ratio of weight of discharge electrode and tensioning
          weight to weight of discharge electrode alone
          (dimensionless)
y.        mole fraction of component i
6         relative gas density  (dimensionless)
C         permittivity of free space  (8.85 x 10 ~12 p/m)
                              xi

-------
                         ACKNOWLEDGEMENTS
     The authors are grateful to Messrs. Robert F. Brown, Alex
Penyak, III, and Joseph C. Shepard of the staff of Research-
Cottrell, Inc. for their numerous contributions to the successful
outcome of this study.  We are likewise indebted to Or. Leslie
Sparks, EPA Project Officer, for his assistance and helpful
comments.

-------
                             SECTION I

                           INTRODUCTION
OBJECTIVES AND SCOPE

     The objective of this project is to demonstrate the feasi-
bility of electrostatic precipitation at the elevated tempera-
tures and pressures associated with developing coal-conversion
processes.  In particular, it is desired to

     (1)  Explore corona stability over a range of temperatures
and pressures from room levels to 1366 K and 3550 kPa, respec-
tively.

     (2)  Estabish current-voltage characteristics (including
corona-staring and sparkover voltages), in terms of temperature,
pressure, geometry and polarity, in a particle-free test precipi-
tator using dry air, a simulated combustion gas and a substitute
fuel gas.

     (3)  Prepare a preliminary cost estimate for a commercial
precipitator to be used in conjunction with a 300-MW pressurized
fluidized-bed combustor.  Detailed and more accurate estimates
must await the availability of particle-collection data.

     Technological developments in coal-utilization systems are
directed at processes producing (1) heat and electricity through
combustion, (2) synthetic fuel gases for local industrial or
pipeline use, and (3) liquids and tars for use as chemical
feedstocks.  Each process operates under characteristic conditions
of temperatures, pressure, gas composition, and particulate
type.  Pressurized fluidized-bed combustion and low-Btu gasification
require particulate removal ahead of a gas turbine at tempera-
tures ranging from 950 K to 1400 K and pressures above 700 kPa.
In the generation of high-Btu gas, the removal of particulate
tar and char usually occurs at temperatures between 500 and 950
K and pressures above 7000 kPa.  The particulate-cleanup stage
in the gasification system is necessary to protect downstream
catalytic processes.  Liquefaction processes generate tars and
oils that must be separated from a gas phase at temperatures
below 500 K and pressures generally below 700 kPa.

-------
     The eras-cleaning requirements of the foreaoinq processes
govern the selection of the temperature and pressure ranges
and the gas compositions considered in this investigation.

FUNDAMENTALS OF HIGH TEMPERATURE/PRESSURE PRECIPITATION

Overview

     The electrostatic precipitator appears to be unique among
conventional particulate collectors such as fabric filters,
cyclones, and scrubbers, in that elevated gas pressure offers
the promise of increased collecting efficiency.  Such improve-
ment results from the potentially higher sparkover voltages,
and hence greater particle-migration velocities, attainable at
high pressures (20).  The advantage holds for precipitation at
high temperature, as long as this condition occurs in combination
with adequate pressure.  This important consideration for high-
temperature and pressure particulate gas cleaning deserves
special mention in view of its lack of emphasis—if not neglect—
by recent writers on the subject  (4, 5, 18, 37).

     There is a fundamental problem encountered either in de-
signing a precipitator for a given high temperature/pressure
service, or in making a realistic economic assessment of its
merits.  This is our incomplete knowledge of i) the range of
variables  (pressure, temperature, electrode geometry, gas com-
position, polarity) over which a stable corona discharge can be
maintained, and ii) the current-voltage characteristics in that
range.  In particular, there exists for the positive discharge
a critical pressure  (not to be confused with its thermodynamic
namesake) above which sparkover alone, without antecedent
corona, prevails.  When the discharge polarity is negative, the
critical phenomenon is not so precisely defined, and a post-
critical discharge (often unstable) may be found at pressures
extending beyond the critical value.

     Thus, in either polarity, the critical pressure effectively
sets a pressure limit on conventional electrostatic precipitation.
The practical consequences of this limitation are, however, not
likely to be serious, for the critical pressure is susceptible
to control.  Available data indicate this range of control to
be so broad in practice that the elimination of critical pro-
blems, at any presently foreseeable temperature/pressure com-
bination, should not be difficult, but only insofar as the
dependence of the discharge on its governing variables is
known.

          This study represents the first reported attempt, at
any pressure, to establish a corona discharge suitable for
electrostaic precipitation at as high a gas temperature as 1360
K, 1200 K being the earlier maximum  (3).  The highest pressure
at which a precipitator has ever operated is 5500 kPa (at
outdoor temperature) (11), exceeding the present maximum of
3550 kPa.

-------
     If pressure presents no insurmountable barrier to the pre-
cipitation process, at least in terms of currently anticipated
applications, neither does temperature, as long as the two
variables appropriately complement one another.  Again, since
high-temperature precipitation (as opposed to hot precipitation
(35) will usually be accompanied by elevated pressure, loss of
corona stability by reason of temperature is not expected to
constitute a general problem.  However, at temperatures beyond
those herein contemplated, thermal ionization must ultimately be
supposed to set an upper limit to stable electrostatic pre-
cipitation.

Critical Pressure

     It is a well-known fact that an increase in gas pressure in
the neighborhood of atmospheric will raise the breakdown voltage
for electrode systems of both parallel-plate and corona-discharge
geometries.  Less familiar are the contrary effects produced by
much lower or higher pressures.  Thus, the breakdown voltage
drops as the pressure falls below atmospheric, passing through a
minimum at pressures of the order of 100 Pa.  The minimum occurs
because the number of gas molecules in the interelectrode gap is
proportional to the pressure.  Low pressure results in a long
mean free path and, consequently, few electrons, in traversing
the gas, manage to collide with gas molecules and ionize them.
In order to have sufficient ionization to lead to breakdown, the
accelerating electric field and, hence, the applied voltage,
must compensate for the low-density effect.   The less the
pressure below a certain value, the higher, then, is the voltage
level required for breakdown.  However, as the pressure rises
away from its minimum voltage value, the mean free path decreases.
Although electron-molecule collisions are now more frequent, few
electrons are able to acquire sufficient kinetic energy over the
reduced mean free path to ionize gas molecules by collision.  It
follows that in order to generate enough ionization in the gap,
the voltage must be raised together with the gas pressure.  This
relation holds until the neighborhood of the critical pressure
is approached.  (We note, in passing, that low-pressure (to
^300 Pa) electrostatic precipitation has been employed in connection
with stratospheric aerosol sampling  (20)).

     The earliest observations bearing on the critical pressure
were made by Ryan  (24) and his student Ekert  (10) at the be-
ginning of this century.  Later workers showed that for elec-
tronegative gases, as the pressure is raised above atmospheric
for a positive-corona system, a point is reached at which
sparkover is not preceded by corona.  In other words, although
both the corona-starting and sparkover voltages initially
increase with pressure above atmospheric, eventually the difference
between the two voltages begins to narrow, and ultimately they
coincide.  The pressure at this point of intersection, beyond
which a positive corona discharge is impossible, is termed the
critical pressure.

-------
     A negative critical pressure, i.e., a critical pressure for
negative corona (at first though not to exist), was discovered
recently by Robinson (20) using combinations of electrode
geometries and pressures not employed by earlier investigators.
The negative critical pressure differs from the positive in that
the former is not so sharply defined (corona-type discharges,
often erratic, may occur over an uncertain range of still greater
pressures) and, for a given system, it is much higher and,
therefore, more elusive.

     Two opposing effects are responsible for the phenomenon of
the critical pressure.  First, as pointed out, shorter mean-
free paths at elevated pressures impede ionization by collision
and so tend to raise the sparkover level.  Second, the denser
packing of gas molecules renders photoionization more likely and
reduces ion diffusion.  Thus, pressure facilitates streamer
propogation from the anode across the gap and, at the critical
pressure, sparkover results.

     The likely explanation of the relatively low value attained
by the positive sparkover voltage and its concomitant lower
critical density is as follows:  Intense ionization of the gas
is produced in the high-field region in the vicinity of the
discharge wire which attracts and removes the highly mobile
electrons.  The heavy positive ions are repelled from the wire
and move slowly toward the collecting electrodes.  However, on
the far side of the ion cloud  (away from the wire), the field,
suffering from positive-ion space-charge distortion, is increased
and with it, the rate of ionization.  Conditions are now favorable
for the growth of a positive streamer to develop toward the
cathode.  On the other hand, when the wire is negative, the
positive space charge surrounding it in the corona sheath tends
to shield the wire from the anode.  This action reduces the
field on the anode side and so a higher voltage is needed to
promote a spark.

     At pressures above the critical value, the slopes of the
negative current-voltage curves abruptly and significantly de-
crease, the discharge is frequently punctuated by irregular
sparking as the voltage is raised, maximum currents attainable
prior to sustained sparkover are reduced, and the data are less
reproducible.  Precipitator performance in this posteritical
region has never been explored.

Thermal Ionization

     We have so far been concerned with mechanisms of ionization
stemming from i) electronic and molecular collision processes in
the presence of an electric field and ii) photon absorption.  At
very high temperatures an additional mechanism enters the
picture:  thermal ionization.  This introduces the following new
possibilities.

-------
     i)  Electron-ion production by collision of the gas
molecules with each other.  Such ionization may occur even in
the absence of ionized particles and a high-intensity electric
field.  It is the high-temperature condition that provides
velocities and kinetic energies high enough to cause ionization.

     ii)  Photoionization resulting from thermal emission of
the hot gas.  When a high-temperature gas flows through a
precipitator, it will emit quanta in accord with the laws of
black-body radiations.  The electrode surface and the walls of
the pressure vessel will reflect most of this radiation.
Thus, photons capable of ionizing are always available when
the temperature is high enough.  Calculation, however, shows
that this condition is probably not met under present circumstances.

     iii)  Ionization by collision with high-energy electrons
that have been generated by the above two processes.

     Apprehensions arise in running precipitators at very high
temperatures because significant thermal-ionization rates may
lead to catastrophic currents at reduced voltages.  Earlier
predictions that this condition might occur at a temperature
of about 1100 K (6) have been shown experimentally to be
untenable (3).  Revised theoretical work (8) now suggests that
whereas thermal effects may become noticeable below 1100 K,
practical high-temperature precipitation may be limited only
by temperatures exceeding 1400-1600 K.  The lower limit of
this range would apply to gases containing a component of low
ionization potential, say potassium, in quantities as low as
one atom in 10 .  Potassium is singled out because of the
alkali metals (all of which have characteristically low ionization
potentials), it is the most common, constituting 1.4 percent
of the earth's crust.  Hence, the probability of its appearance
in industrial gases is not remote.

     It is worthwhile bearing in mind, however, that normally
unanticipated minority ionic species may effect the corona
process.  According to Loeb (15), the relatively high negative
currents and low accompanying sparkover voltages that several
observers have reported at high temperatures (27, 33, 36) are
probably due to the action of the 0" or 02~ ion.  The point to
consider is the fraction of the time an electron can remain
detached from a molecule or atom, the average detachment
interval increasing with temperature.  Since the negatively
ionized oxygen molecule 0 "sheds its electron at about 600 K,
followed at somewhat higher temperatures by the 0~ ion, the
high temperatures of present interest should be dominated by
heavy free-electron concentrations.  This condition gives rise
to heavy currents and causes a loss of corona stabilization
otherwise resulting from negative-ion space charge.  In Loeb's
view, the 0~ ion with an electron affinity of 1.5 eV, well
below the 4.3 eV ionization potential of potassium, would

-------
probably be the primary culprit for the ultimate high-temperature
failure of the precipitation process in oxygen-bearing gases.
However, if potassium were present in sufficient quantity, its
effect would govern.

     It must be emphasized that details of the thermal-ionization
process described above are, as applied to precipitators  (even
under laboratory conditions), largely hypothetical.  For example,
none of the ionic species has been positively identified in a
high-temperature precipitator test, and the predictions of
limiting temperature are not supported in any consistent
manner by the experimental data.

Earlier Studies

     Previous investigations relating to this study are cited,
as appropriate, in the text.  Other precipitation work worthy
of note is that of Roller and Fremont, dust-free, in air and
methyl chloride, to 800 K and 500 kPa  (14); Thomas and Wong,
dust-free, in air, to 1100 K and 810 kPa  (33); Shale and asso-
ciates, with and without dust, in air and flue gas, to 1100 K
and 650 kPa  (25-32); and most recently, Brown and Walker, with
dust, in flue gas, to 1200 K and 896 kPa  (3).

-------
                             SECTION 2

                            CONCLUSIONS


     At high temperatures and pressures such as are found in
advanced coal-conversion processes, electrostatic precipitators
exhibit potentially higher collecting efficiencies than prevail
under conventional conditions of precipitator application.
High efficiencies are an anticipated consequence of the increased
voltages that can be applied to the precipitator for a broad
range of pressure-temperature combinations.  This efficiency
effect is unique for precipitation; in other particulate
collection methods, efficiency generally decreases with increasing
temperature and pressure.

     In the present study, no practical upper limit is reached
for electrostatic precipitation at temperatures and pressures
as high as 1360 K and 3550 kPa.  In some instances, however,
limits imposed by the critical pressure are approached.  The
critical pressure is the lowest elevated pressure at which
corona-starting and sparkover voltages coincide.

     Increasing the temperature raises the critical pressure
and so results in a broader operating range of pressure.  The
negative-corona critical pressure, attained here only at room
temperature, is found to be significantly higher than the
corresponding positive value.  Positive-corona critical pressure,
reached at all temperatures, is observed to increase with
decreasing wire diameter.

     The higher the pressure, the more slowly does current in-
crease with voltage, i.e., the more stable the discharge.  As
a practical rule, high pressure should accompany high temperature.

     Positive sparkover voltage exceeds negative for conditions
of low gas density and high temperature.  For high density,
negative sparkover voltage is higher than positive.

     On the basis of the stable corona currents obtained for
the three gas compositions examined — air, combustion gas and
a substitute fuel gas — electrostatic precipitation is judged
feasible in each case.  The fuel gas yields the highest corona
currents for a given voltage.

-------
     The amplitude of discharge-electrode oscillations is de-
pendent on electrode length, shape, tension and material.  Short,
rigid or highly stressed electrodes produce minimal oscillation.
The rigors of high-temperature service require discharge and
collecting electrodes of high tensile strength and high oxida-
tion and carburization resistance.  Oxidation and carburization
result in an expanding scale growth that alters the electrode
surface and affects both corona-starting and sparkover voltages.
Inconel 601 performs satisfactorily at temperatures up to 1360 K.

-------
                             SECTION 3

                          RECOMMENDATIONS
     In view of the encouraging results obtained so far  but
which are limited to a particle-free system  it is recommended
that we next proceed to the design and operation of a pilot
precipitator treating fluidized-bed combustor or gasifier ef-
fluents.  Estimates of particle-migration velocity derived from
the current-voltage data of this report make possible the ra-
tional design of an operational precipitator.  Pilot tests with
such a precipitator will reveal the effect of the presence of
particulates on electrical performance, will permit the checking
of predicted values of particle-migration velocity, and will
provide experience in aerosol sampling and precipitate removal
under extreme conditions of temperature and pressure.

     The diameter of the pilot collecting electrode will be
severalfold greater than in the present case.  Beneficial
consequences of the larger electrode include a more stable
corona characteristic and higher critical pressure.

     High-temperature evaluation of materials of construction,
with special attention to electrodes and insulators, will ex-
tend into the pilot phase.

-------
                             SECTION 4

               EXPERIMENTAL APPARATUS AND PROCEDURE
TEST PRECIPITATOR

     The experimental electrostatic precipitator comprises a
concentric wire-pipe electrode system surrounded by an electric
heater.  The entire assembly is enclosed in a pressure vessel
as shown in Figure 1.  The vessel, designed to withstand pressures
upwards of 3550 kPa, is assembled in three sections.  Mounted
on the uppermost part is an alumina bushing through which a
high-potential conductor enters the vessel and connects to the
discharge electrode.  A rupture disc in the top flange protects
against the hazard of overpressurization.  The bottom third of
the vessel has a side access port for making adjustments and
observations, a support insulator for centering the lower end
of the discharge electrode, and a gas inlet.  The precipitator
tube together with a three-zone heater is fixed in the center
segment of the pressure vessel.  Disc-shaped alumina end
pieces serve to center the precipitator tube at its top and
bottom, and by restricting convection currents, help maintain a
more uniform temperature over the tube length.  The heater is
packed in a layer of Kaowool insulation about 5-cm thick; this
material fills the space between the heater and the wall of the
pressure vessel.

     The tubular  (collectincr) electrode is fabricated from
Inconel 601 and is 7.26 cm in internal diameter.

     In order to obtain a corona-current measurement uninfluenced
by end effects, the 76.2-cm long collecting electrode is divided
into three tubular segments, each electrically insulated from
its nearest neighbor(s).  The 61.9-cm long midsection is grounded
through a current measuring device —either a milliammeter or
one axis of an X-Y recorder.  In this manner, only current to
the midsection is measured.  The terminal guard sections, each
7.15-cm long are grounded directly, and currents to them are
excluded from the experiment.  The ends of the guard sections
are flared sufficiently outward to insure that sparkover occurs
not at the edges of the tube but in its long straight portion.

                               10

-------
 ALUMINA
 END PIECES
                                              TO POWER SUPPLY
                                              TOP INSULATOR BUSHING
                                              RUPTURE DISC

                                              HIGH VOLTAGE FEED
                                              THROUGH ROD
                                              DISCHARGE ELECTRODE
TUBE ELECTRODE
                                              HEATER
                                              ACCESS PORT
                                              BOTTOM SUPPORT
                                              INSULATOR
                                              GAS INLET
                     TEST PRECIPITATOR

Figure L.  Laboratory Precipitator and Pressure Vessel for Test
           Program
                             11

-------
     The wire discharge electrode, also of Inconel 601, is
secured at its upper end, via a shroud, to the high-voltage
feedthrough rod.  At its lower end, the wire is connected, again
by means of a shroud, to a 0.98-kg weight.  The weight keeps the
wire taut and suppresses oscillations.  Three wire diameters are
employed in these tests:  1.575 mm, 2.344 mm and 3.175 mm.

HIGH-VOLTAGE POWER SUPPLY

     The high voltage power supply is a silicon-rectified vol-
tage quadrupler  (Figure 2a) on the secondary output of a 105-kV
peak transformer, yielding voltages to 400 kV peak and currents
to 60 mA.  The transformer primary voltage is controlled by
twelve variable transformers driven by a variable-speed motor.
The power-supply output is connected to the experimental precipitator
through a series resistance of 145,000 ohms and an inductance of
0.57 mH to protect against spark transients.  Output-voltage
ripple at maximum precipitator current is less than one percent
of the rms voltage and is, therefore, assumed negligible.

     High voltage is measured by means of the resistive divider
network shown in Figure 2c.  The voltage is read both on a
voltmeter and the remaining axis of the X-Y recorder mentioned
previously.  Thus, provision exists for the automatic plotting
of corona current as a function of applied voltage.  The high-
voltage measuring circuit is calibrated in terms of the accurately
measured currents flowing through the precision resistors of the
divider and, independently, by an electrostatic voltmeter.

     'The current arm of the recorder is protected from transient
surges by a needle spark gap, an inductor, a metal-oxide varistor,
and shunt capacitors  (Figure 2d).

     Polarity may be changed by reversing the silicon rectifiers.

TEMPERATURE CONTROL SYSTEM

     The temperature of the gas and the precipitator is controlled
by a three-zone heater surrounding the tube.  The top and
bottom zones, each 15.2 cm long, control the temperatures at
eachend of the tube.  The end heaters, in combination with the
45.7-cm long center, zone, maintain a uniform temperature along
the entire tube length.  Three proportional controllers employing
thermocouples placed next to the heater elements maintain a set
temperature level using the control circuit shown in Figure 3.
For additional protection from high-voltage transients, each
heater is grounded inside the pressure vessel, with one power
lead emerging from each zone element.  An isolation transformer
applies voltage to each variable transformer and heater through
the relays controlled by the outputs of the proportional controllers.


                                 12

-------
                       Cl*    (*  f*
                     I  l**»    ** 'I *"*J\

                     SLfiW
460V
(u)
                                                      >
            Figure 2  Power Supply and Instrumentation Measuring Circuits:
                     (a)  Power Supply Quadrupler
                     (b)  Total Current Circuit
                     (c)  Voltage Divider Circuit
                     (d)  Tube Current Circuit

-------
                  I  t
Isolation
Transformer
208V
                                                   End
                                                   Heacer
7 	
^ ~
O.SuF
^™
j
^^
                                                           TC
                                                    /77
                                                   End
                                                   Heater
                                               0.
                                                           TC
                                                    777
                   HEATER CONTROL CIRCUIT
   Figure  3-  Three-zone Precipitator Heater and Control Systems
             for Temperatures to 1400 K.
                               14

-------
The temperatures of the interior of the tube is calibrated
against the temperature of the thermocouples located at the
heating elements.  This is done by recording temperatures at six
positions along the inner length of the tube for each controller
setting.  The results form the temperature-calibration curve
shown in Figure 4.  Using this calibration curve with the
three-zone control, a uniform temperature can be established and
held throughout the entire tube for each test.

GAS-FLOW SYSTEM

     The gas flow system is designed to control the final gas
composition within the precipitator.  The process flow system,
shown in Figure 5, has (1) five individual gas cylinders, with
regulators containing N-, He, dry air, and mixtures of N_ -
CO, and CO, - O2; (2) a set of four pressure guages appropriately
spanning tKe required pressure range to 6890 kPa; (3) provision
for passage either through or around the humidifier; and  (4)
heated lines into the precipitator vessel.  Sampling points are
available for analysis of the gas.

     The humififier as shown in Figure 6 is required for adding
moisture to the gas mixture.  By dispersing the gas through a
heated water bath, the gas becomes saturated at the bath tem-
perature.  The gas is then immediately heated to a higher tem-
perature (477 K) in a heated packed bed and sent through heated
lines to the pressure vessel, constantly maintaining the gas
above its dew point.  A water-bath temperature of 388 K will put
5 percent water vapor into the gas phase with a total pressure
of 3550 kPa.  The test facility (Figure 7) is arranged to provide
maximum control and safety during the test runs.

     Three gas compositions were used for testing:  air, a
simulated combustion gas, and a substitute fuel gas with component
concentrations shown in Table 1.

                              Table 1

                       TEST GAS COMPOSITIONS

                                   (Volume %)
                 Fuel Gas                 Combustion Gas
Component	Mixture	Mixture	Air

   C02              23.0                       9.2

   He               18.5

   02                -                         2.8          21

   N2               53.5                      83.0          79

   H20               5.0                       5.0

                            15

-------
1400
1200
1000
 800
 600
 400
 200
                I     l     I
                                    •     •      i     i     i x
           J	L
                                  0  CENTER HEATER

                                  A  END HEATERS


                                  '     I	I	L_
    200        400        500        300       1000      1200      1400
 Figure  4.
       CONTROLLER TEMPERATURE °K

Precipitacor Tube and Internal Gas Temperature Based
on Each Heater Thermocouple Temperature.
                                16

-------
                                                                    TO
                                                                   VENT
                                                      OXYGEN
                                                      ANALY-
                                                      ZER
 TO
VENT
                                 PRESSURIZING  FLOW  SYSTEM
Figure 5.  Gas Pressurising and Mixing Flow System Used to Fill Precipitator
           for Each Test Series.

-------
                                             THERMO COUPLE WELL
                                             GAS OUTLET
                                             HEATER
                                             CERAMIC BEADS
                                             WATER BATH

                                             IMMERSION HEATER





                                             WATER INLET

                                             GAS INLET
                         HUMIDIFIER
Figure  6.  Internal Schematic of Humidifying System Showing
           Water Bath and Heater Used for Gas Saturation and
           Moisture Level Control.
                               18

-------
                                        MOTOR
                                        DRIVE

                                        POWERSTAT
                    POWER
                    SUPPLY

r


i
OP*
\j
a
Ocs
^
                              o
                                   PRESSURE VESSEL
                                     TEST ROOM
                                                         TRANSFORMER
                              PRESSURE CONTROL  PANEL


                              INSTRUMENTATION



                                CONTROL CONSOLE
                                HEATER
                                TRANSFORMER
                   EQUIPMENT FLOOR LAYOUT

Figure  7   Schematic Layout of all Test Equipment, Power Supply
           and Instrumentation Showing Both the Test Room and
           Control Room.
                                 19

-------
     Dry compressed air  (dew point, 216 K) is used for all air
tests.  The humidifier is bypassed when pressurizing the vessel
to 3550 kPa with air.

     To simulate combustion gas, after purging the pressure
vessel with nitrogen, the gas components are passed through the
humidifier in the following manner:

      (1)  Using a 77%-CO- and 23%-02 mixture, the pressure
vessel is filled to 550 KPa.

      (2)  Using 99.99% N2 gas, the pressure vessel is filled
from 550 kPa to a final pressure of 3550 kPa, yielding the
composition given in Table 2.

     The substitute fuel gas is prepared in a similar fashion,
by passing all gases through the humidifier:

      (1)  Using a 30%-C02 and 70%-N2 mixture, the precipitator
is pressurized to 1140 kPa.

      (2)  Using pure helium, the vessel is filled from 1140 kPa
to 1830 kPa.

      (3)  The final step is to finish filling the pressure
vessel to 3550 kPa with the 30%-CO2 and 70%-N2 mixtures. This
results in the composition shown in Table 2.

     •Helium is chosen as a substitute for the hydrogen of
actual fuel gas because the two constituent gases offer the
closest readily attainable match of thermal conductivity and
ion mobility.  Carbon monoxide is replaced with carbon dioxide.
A comparison of conductivity and mobility values for the various
gas mixtures is shown in Table 2.  Calculation procedures are
found in Appendices A and B.

                              Table 2

    THERMAL CONDUCTIVITIES AND IONIC MOBILITIES OF GAS MIXTURE

Real Fuel Gas
Substitute Fuel
Gas
Combustion Gas
Air
(mW/m-K)
38.3
41.1
28.4
29.9
(cm /(s-V))
1.64
1.87
1.91
2.5
(cm'/ (s-V))
1.49
1.58
1.59
1.8
                                 20

-------
TEST CONDITIONS AND PROCEDURE

     The experimental procedure is chosen (1) to meet the
rigors of extended high-temperature and high-pressure testing,
(2) to obtain sufficient data to estabish ranges of corona
stability, and (3) to show effects of composition, electrode
geometry, and polarity.  The ranges for each variable are shown
in Table 3.

                              Table 3

                      EXPERIMENTAL VARIABLES

               Temperature         294, 533, 811, 1089,
                                   1366 K

               Pressure            101 to 3550 kPa

               Wire Diameter       1.575 mm, 2.344 mm,
                                   3.175 mm

               Polarity            positive, negative

               Gas Mixture         air, combustion gas, sub-
                                   stitute fuel gas

     Three wire sizes are used in the air tests; however, only
one wire size is used for the combustion-gas tests and the
substitute-fuel tests.

     In conducting the tests, a steady-state temperature is
attained by maintaining the heaters at the desired temperature
setting for a minimum period of ten hours, with the gas mixture
added four hours prior to testing.

     In moving from one condition of pressure to another, but
at constant temperature, the gas is always released from the
vessel in order to retain constant composition and thermal
stability.  At each pressure, results for both negative and
positive wire polarity are obtained.

     Prior to recording a current-vs-voltage curve, the voltage
is held at sparkover for a period of two minutes to "condition"
the electrodes.  Duplicate curves are then recorded with both
increasing and decreasing voltage.  The minimum voltage at which
repetitive sparking occurs over a 2-min interval is reported
as the sparkover value.  Corona-starting voltages are determined
by two methods:   (1) observation of voltage at which corona pips
disappear on an oscilliscope with decreasing applied voltage and
(2) extrapolating the current-voltage curve to zero current.  Some-
times, a sharp elbow occurs at currents of a few tenths of a
milliampere per meter or less.  See, e.g., Figure 13e.  In such a
case, the knee is disregarded in extrapolating the curve.


                                 21

-------
     In the presence of corona, the gas pressure within the
vessel increases.  Changes of as much as 3 or 4 percent, cor-
responding to temperature increases of 5 or 6 percent, are
observed under some conditions.  Gas density, however, remains
essentially constant (+  2 percent).

DISCHARGE-ELECTRODE OSCILLATION TEST

     Different discharge electrode geometries are evaluated at
elevated wire temperatures to determine the effect of geometry
on  (1) amplitude of oscillation and  (2) strength of material.
Previous experience shows that wire oscillations result in sub-
stantially lower sparking voltages due to reduced clearances.
The equipment designed to test the several geometries is shown
in Figure 8.  A collecting pipe electrode made of type 316
stainless steel with an internal diameter of 0.213 m and an
overall length of 3.658 m is securely mounted on a support
structure that isolates it electrically from ground.  The discharge
electrode is centrally mounted within the pipe, and a weight is
attached to the bottom end.  This tensioning weight is needed to
keep the wire taut and dampen potential oscillations.  A stepdown
transformer, connected to each end of the wire electrode, provides
the current needed to reach the high electrode temperatures
approaching 1366 K.  A mirror and graduated scale are placed
beneath the pipe to permit the oscillations to be observed and
measured.  A 106-kV peak, high-voltage, reversible-polarity
power supply is connected to the pipe.  Current- and voltage-
measuring circuits are shown in Figure 8.

     Electrode temperature is determined by using temperature-
sensitive color indicators on each electrode.  The various
geometries of the electrodes are shown in Figure 9 and are
tabulated below:

                              Table 4
                       ELECTRODE GEOMETRIES
               7.94 mm

               6. 35 mm

               6.35 mm

               6.35 mm

        three  2.77 mm

               2.77 mm

          two  1.59 mm
Twisted Square

Twisted Square

Round Wire

Stranded Cable

Twisted Wire

Round Wire

Barbed Wire with 19-mm
Barb Diameters
                                22

-------
/f?
                n
                                                   CERAMIC INSULATORS
              DISCHARGE-
              ELECTRODE
                                   •COLLECTING
                                    ELECTRODE
                                    WEIGHT
                                       .MIRROR
                                                         STEP-DOWN
                                                         TRANSFORMER
220V
                        OSCILLATION TEST

                            APPARATUS
  Figure  8.  Laboratory Apparatus Set to Test Effect of Electrode
             Size and Geometry on Degree of Oscillation at High
             Voltage.
                                    23

-------
                                                                      y
s
/
                        o
o
 7.94 nun
TWISTED
SQUARE
6.35 mm
TWISTED
SQUARE

6.35 nun
ROUND


6. 35 mm
STRANDED
CABLE

3-
2.77 mm
TWISTED
ROUND
2.77 mm
ROUND


2-
1.59 mm
BARBED

            OSCILLATION ELECTRODE GEOMETRIES
    9.   Cross Suclional and Lateral Views of the Seven  Electrode
        Geometries Used in the Oscillation Study.

-------
                             SECTION 5

                      RESULTS AND DISCUSSION
THE EXPERIMENTAL DATA

     The raw experimental data, consisting of curves of linear
current density (mA/m) vs impressed voltage (kV) are reproduced
in Figures 10 through 13 for air, Figure 14 for simulated com-
bustion gas and Figure 15 for substitute (i.e., noncombustible)
fuel gas.  Corona-starting and sparkover voltages, derived from
these curves and independent measurements, are shown as functions
of relative gas density 5 in Figures 16 to 18 for air, Figure 19
for combustion gas, and Figure 20 for substitute fuel gas.
Figures 21 and 22 compare experimental and calculated corona-
starting fields.

     The relative gas density 6 is taken with respect to atmos-
pheric pressure (P  = 101.3 kPa) and room temperature (T  = 294
K)                °                                     °

                   T
               6 = -2-  - = 2.90 x 10~3 P.                   m
                       m               m                    V -Ly
     In Figures 10 through 15, all current-voltage curves ter-
minate in sustained sparkover indicated by a short broken line
segment.

TEMPERATURE/PRESSURE LIMITATIONS ON CORONA STABILITY

General
     Our first and most important objective is to examine the
data for the purpose of establishing temperature or pressure
limits to a stable corona discharge.  Such limits may be either
of a fundamental nature, that is, inherent in the physics of the
gaseous discharge, or practical, as for example, restrictions
imposed by the strength of materials at high temperature.  Fo-
cussing for the present on the fundamentals, we search the
current-voltage curves of Figures 10 to 15 for i) excessive
currents at low voltages resulting from thermal ionization
(where "excessive" and "low" are taken from the point of view

                                   25

-------
of practical precipitator operation) and ii) the disappearance
of (stable) corona due to the manifestation of the critical
pressure.

Excessive Current

     Catastrophic high-temperature currents are not observed
in this study under any conditions.  This significant point
is evident over the full range of experimental pressures
(atmospheric to 3550 kPa) and temperatures  (294 K to 1366 K)
and both polarities.

     It might be expected that runaway currents at low
voltages are most likely to occur at the lowest gas densities.
We note, however, for the negative corona at 6 less than
about unity, the reverse is generally true:  the presparkover
currents are much less than at higher densities.  The curves
of Figures 12(d), (e) and  (f) for 6 = 0.33 and 0.34 are
typical examples.  For positive polarity, the low-6 presparkover
currents are mixed:  in most cases they are not lower than
at higher 6.  Still considering the positive discharge, we
see that combustion gas  (Figure 14), and fuel gas to a
lesser extent  (Figure 15). but not air, reveal a tendency to
sparkover without corona at the lowest densities.  But, in
any event, a problem of high gas conductivity associated
with low densities does not arise.

     Reduced presparkover currents at low <5 have been reported
elsew*here for both polarities in air and are not a consequence
of high temperature  (20).

     CONCLUSION:  Uncontrollable corona currents, whether due
to high temperature" or low density, do not occur in these te₯ts.

Positive Critical Pressure

     On the basis of earlier critical-pressure studies  (20,
21) it might be supposed, at least at lower temperatures,
that the critical-pressure phenomenon would set an upper-
pressure limit to the positive corona in the present experiments.
That this is indeed the case is borne out by the positive
current-voltage curves of Figures 10 to 13 for air, Figure
14 for flue gas. Figure 15 for combustion gas, and the
associated voltage-density curves of Figures 16 to 20.  It
is also seen from the curves —upon converting density to
pressure — that the positive critical pressure increases
with temperature.  This comes about because the greater
molecular diffusivity prevailing at the higher temperatures
more effectively suppresses the sparkover streamer before it
completely bridges the interelectrode space (12).  The
effect described has not hitherto been demonstrated over so
wide a span of temperatures.

                                 26

-------
     CONCLUSION:  Temperature raises the positive critical
pressure.

Negative Critical Pressure

     The negative critical pressure, as already explained,
should be sought at higher levels than the positive, all other
conditions being held fixed.  The critical pressure, it is seen,
is attained, or at least closely approached, for the positive
corona in every case (Figures 10-15}.  The negative current-
voltage curves of Figures 10 to 15 suggest, especially at the
lower temperatures, that the critical pressure is being approached -
in Figure 10 (d) perhaps even attained —as the pressure nears
its experimental maximum.  I.e., with increasing pressure,
sparkover voltages increase less rapidly or actually drop, and
maximum presparkover currents fall abruptly.  Since the negative-
corona data permit a lower bound to be set to any negative
critical pressure, it is clear, by comparison of families of
curves for each polarity, that the negative critical pressure
exceeds the positive.  It is further apparent that the negative
critical pressure, like the positive, increases with temperature.
In other words, the higher the temperature, the greater the
range of pressures of the negative corona.  The highest temperatures
of this study, instead of rendering the corona unstable by
reason of thermal ionization, actually enhance corona stability
by moving the critical pressure outside the pressure range of
interest.

     CONCLUSION:  The negative critical pressure exceeds the
positive.

     CONCLUSION:  Temperature raises the negative critical
pressure.
Practical Precipitator-Tube Diameter
     The diameter of the outer electrode used in this study is
7.62 cm, about one-third of the diameter likely to be selected
in a practical precipitator.  Since the negative critical pressure
for wire-tube configuration in air is proportional to the tube
diameter (20) a realistic practical tube diameter of three times
that used here should eliminate the critical-density restriction
even at the lowest temperatures employed.

     Another matter for concern are the steeply rising currents
typifying most of the current-voltage curves, and especially
those for negative corona in substitute fuel  (Figure 15(f},  (g),
and  (h).  In these cases, rough extrapolation to a threefold
greater tube diameter may be made by means of the quadratic
corona approximation
                    •

                                  27

-------
                                        V(V-Vc)
                        ro  ln  {ro/ri}-

where :

     j is current density  (A/m)

     eQ is permittivity of free space  (8.85 x 10~12 F/m)

     b is ion mobility  (m /sec-v)

     rQ is tube radius  (m)

     ri is wire radius  (m)

     V is applied voltage  (V)

     VG is corona-starting voltage  (V)

It is clear that a threefold increase  in r  will reduce j by
about an order of magnitude at a given vol€age.  The slopes of
the current-voltage curves will be much shallower, and the
currents correspondingly more susceptible to control.

     CONCLUS ION :  No pressure limitations to negative-corona
stability over the range explored are  expected in a precipitator
tube of practical diameter.

     The effect of temperature in reducing gas density or causing
thermal ionization tends to raise the  current.  Pressure, on the
other hand, compensates by reducing current.

     CONCLUS ION :  Generally speaking,  high pressure should
accompany high temperature as long as  the critical pressure
is not too closely approached.

POLARITY EFFECTS

Background Considerations

     It is virtually axiomatic in conventional precipitation en-
gineering, and has been since its early recognition by Cottrell,
that the negative corona yields higher sparkover voltages than
the positive and, consequently, higher collecting efficiencies.
The negative advantage, however, is demonstrably not universal.
Sometimes, at least when the electrodes are clean, there is
little apparent difference between the polarities (20) , and
sometimes, as for certain combinations of high- resistivity
precipitate and pressure, the positive sparkover voltage may
exceed the negative by many tens of kilovolts (22) .   Additional
actual and theoretical cases are cited by Cooperman (7, 9).

                               28

-------
     The foregoing higher positive sparkover voltages apply at
room temperature.  At elevated temperatures, several inves-
tigators (27, 33, 36) have also reported higher positive spark-
over voltages, but here the high-temperature condition is an
essential feature of the phenomenon.  In the high-temperature
negative corona, the equilibrium attachment-detachment conditions
that seem to obtain give rise to heavier free-electron concentration!
Two things now happen that superimpose their effects, in some
degree, on the usual corona processes.  The high-mobility free
electrons are rapidly swept out of the interelectrode gap by the
field, an action that yields higher negative than positive
corona currents at a given voltage.  In addition, the resultant
loss of negative-ion space charge will fail to provide for
stabilization of the negative discharge by suitable space-charge
modification of the interelectrode field as is normally the case
at ordinary temperatures.  Negative-corona sparkover, therefore,
occurs at lower levels than positive.

Experimental observations in Air

     Comparison of the sparkover-voltage vs gas-density data of
Figures 16-18 reveals a tendency for the positive sparkover
voltage to exceed the negative at temperatures of 533 K and
higher and for air densities 5 less than about 1 or 2.  Now the
data are not unequivical on this score in each case, a not
unusual development in view of the uncertainties of the measure-
ments.  The trend, nevertheless, seems clear, particularly when
supported by experimental results of earlier workers.  The
significantly higher negative than positive currents prevailing
at a given voltage at the higher temperatures are, however,
unmistakable  (Figures 10-15).

     The somewhat higher negative than positive currents that
may be observed at the lower temperatures are, in part, to be
attributed to the significant free-electron component of the
current present for relatively long mean free paths  (low 6) and
narrow interelectrode spacing.

     Again, it may be generally (though not invariably) seen
from Figures 16-18 that, above an air density of 1 or 2, the
negative sparkover voltage is higher than the positive.  Since
increased density reduces the mean free paths and mobilities of
the charge carriers, enhanced electron attachment an increased
negative-ion space-charge density might be expected to lead to
higher negative sparkover voltages.  That is, high pressure in
combination with high temperature restores, in a sense, the low-
temperature situation.

Experimental Observations in Combustion and Fuel Gases

     In the case of  substitute fuel gas  (Figure 20) the positive
sparkover voltage exceeds the negative, over the full temperature
range  shown, up to a density of 6 or 7.  For combustion gas
 (Figure 19) the transition occurs at about a density of 4 for
temperatures of, or  greater than, 1089 K.
                              29

-------
     CONCLUSIONS:  At a given density and voltage, the negative
corona current exceeds positive, the difference increasing with
temperature.

     CONCLUSION:  High pressure favors higher negative spark-
over and low pressure higher positive sparkover.

CORONA-STARTING VOLTAGE

Data Fluctuations

     Each frame of the corona-starting voltage vs gas-density
data (Figures 16-20) consists of a series of corona-starting
measurements taken at a fixed temperature, with descending pres-
sures, over relatively brief intervals of time.  Longer-term
fluctuations of the data, caused by temperature cycling and other
factors, are not expecially  apparent in any single frame.  Such
fluctuations in corona starting voltage measurements do, nevertheless,
exist, expecially for the negative corona  (19).  Since, in this
study, thermal ionization is not primarily responsible for the
discharge in the negative corona, it is the wire itself, rather
than the gas, that is the source of secondary electrons needed to
maintain the current.  Emission from the wire is dependent on the
physical and chemical microstructure of its surface, the presence
of oxides and other films, and similar influences not amenable to
experimental control.  Surface contaminations and irregularities
produce electrical effects that are intensified by the diminution
of lateral diffusion at high gas densities, with a subsequent
increase in space-charge distortion.

     The situation is much different in the positive corona.
Here the cathode is in a reduced-field zone and cannot contribute
significantly to the maintenance of the discharge.  The source of
required seondary electrons — again assuming minimal thermal
ionization effects — is the gas surrounding the anode wire.
Since the corona-starting voltage is now relatively unaffected by
the status of the elctrode surface, it is  less  likely to fluctuate
with time or temperature.

     Irregularities in the corona-starting voltage may influence
the shape of the current-voltage curve well removed from the
point of corona onset.  This behavior can  account for much of the
nonreproducibility exhibited by the experimental results.

Peek's Equation

     A quantitative idea of the experimental variability of the
corona-starting voltage is conveniently obtained by comparing
measured corona-starting fields at the wire surface and comparing
them with values calculated from the Peek  relationship  (17)


                              30

-------
                  = m { A6 + Btfi/r}                       (3)
where :

     E  is corona starting field  (V/m)

     m is surface roughness factor  (ds)

     5 is relative gas density  (ds)

     r. is wire diameter  (m)

For air, recommended values of the constants A and B are 32.2 x
105 V/m and 8.46 x 104 V/m3* respectively  (20).

     The roughness factor m, commonly between 0.5 and 0.7 for
the marred and dusty wires of industrial precipitators , is in
the same approximate range for the heavily oxidized negative
wires of the present tests  (Figure 22) .  In every case, the
negative corona starting field is well below the value for
smooth clean wires.  The positive starting fields (Figure 23)
also show wide variability but, following expectation (19) , are
scattered over a somewhat narrower range closer to Peek's pre-
dicted values.

Current- Voltage Elbows

          The elbow-like bends appearing close to corona onset
in the negative cur rent- volt age curves (e.g. Figures 12 (e) and
(f) and 13 (e) and (f ) ) are also observed in room-temperature air
at elevated pressures (19) .  This phenomenon, at least in the
case of the sharply angular elbows, seems restricted to the
negative discharge, and is more likely for larger diameter wires
(cf. Figures 12 (d) ,  (e)  and (f) and 13 (d) ,  (e) and  (f)) and
higher temperatures  (usually for 1089 K and above) .  It might be
supposed that surface asperities play a central part in initiating
the very low current portion of the discharge.  If these superficial
irregularities consist of fine points of effective diameter less
than that of the wire, a measurable corona current might be
expected to flow at voltages below those anticipated on the
basis of the nominal wire diameter.  Such fine points would have
a more pronounced effect, the larger the diameter of the wire
relative to the size of the points.

     Possibly, however, the explanation of the elbows is to be
sought elsewhere.  Field emission, the next -mechanism to come to
mind, requires fields greater than about  10  V/m, and then only
if a suitably contaminated cathode surface is provided.  The
fact that some of current-voltage curves extend to, or close to,
the origin of coordinates clearly excludes field emission from

                              31

-------
further consideration.  The preceding objection is conceivably
met by postulating thermionic emission from the cathode, a
mechanism which also possesses the advantage of accounting for
the appearance of the elbows — in this study — only at high
temperatures.  Thermionic emission can yield significant currents
at very low fields, at temperatures over 1000 K, provided that a
suitable prepared and preserved emitting surface is available.
That the essential combination of propitious circumstances might
occur together by chance, in air, is altogether unlikely.  The
question of the origin of the elbows remains unanswered.


DISCHARGE ELECTRODE OSCILLATIONS

     Seven discharge electrode geometries have been tested to
determine  (1) tendency to oscillate and  (2) weight required to
dampen oscillations without exceeding the yield stress point.
These results are tabulated in Table 5 and include wire shape,
size, length, weight ratio  (w), temperature, sparking voltage,
oscillation span and change in length.  The parameter w is the
ratio of the weight of electrode plus bottom weight to the
electrode weight alone, and determines the applied tension.  The
oscillation span is the maximum swing of the discharge electrode
toward the pipe.  Peak and average sparking voltages are given
for both polarities.

     The large oscillations that are observed with the smaller
diameter electrodes are successfully damped by increasing the
weight ratio.  The barbed wire, however, exceeds its yield
stress point before being totally damped.  Comparing the electrodes,
the shorter lengths of 3.20 meters are found to exhibit less
oscillation.  The more rigid electrodes of larger diameter have
a smaller oscillation span.  Based on these results, a short
rigid electrode, either twisted square or stranded, with its
higher mechanical strength and low oscillation tendency, would be
preferred to a small-diameter and long electrode.  Oscillations
are minimized by limiting the electrode lengths to 3 or 4 meters
and using a relatively high weight ratio of between 10 and 15
total grams per gram of electrode.

MATERIAL EVALUATION

     The surface roughness of the discharge electrode is a
critical parameter for corona-starting and sparkover voltage.
Oxidation that produces an expanding scale changes the starting
voltage.  Too deep an oxidation produces poor mechanical strength.
Electrodes of 316 stainless steel produce severe oxidation
scaling and are unsuitable for high-temperature operation.
Inconel 600  produces a small increase in wire diameter after
two weeks of cycling between 811 K and 1366 K.  This material
may be suitable at temperature levels of between 1000 K and 1250 K.

                              32

-------
                                                          Table   5.
                                            DISCHARGE ELECTRODE OSCILLATION RESULTS
Electrode
Electrode W Length Ht/Length Temp
Description Ratio m. q/m K
A. 7.94 mm
Twisted Square
(Magnetic)




B. 6. 35 nun
Twisted Square
(Magnetic)


C. 6.35 mm
Round
(Magnetic)


D. 3-2.77 mm
Twisted Round
(Non-Magnetic)

2.8
2.8
2.8
2.8
2.8
3.3
3.3
4.2
4.2
4.2
5.3
5.3
3.3
3.3
3.3
4.2
4.2
8.7
8.7*
11.4
11.4
4.31
4.31
4.31
4.31
4.31
3.20
3.20
4.26
4.26
4.26
3.20
3.20
4.31
4.31
4.31
3.20
3.20
4.30
4.30
3.20
3.20
320.3
320.3
320.3
320.3
320.3
320.3
320.3
177.5
177.5
177.5
177.5
177.5
243.7
243.7
243.7
243.7
243.7
143.3
143.3
143.3
143.3
302
866
1033
1089
1200
1366
303
302
866
1033
1200
302
301
866
1144
1366
302
303
866
L200-1255
305-299
SPARKING VOLTA1
WIRE POLARir
Positive Negat
kVp kVavg kVp 1
58.2
63.5
59.7
60
60
60
55.5
59
65
65
63.5
55
53.1
65
60
60
55
56.8
64
62.9
55
47.1
46.4
43.7
43.3
42.6
42.9
44.1
47.1
46.8
46.0
45.4
44.5
47.5
47.1
42.9
43.5
48.0
47.3
46.2
45.8
45.6
96.5
95
90
93.8
90
90
100
96.5
95
95
95
100
100
90
90
83.3
103.1
100
75
100
97.3
Change In
*ES Electode Electrode Resistance
c Oscillations Length °E Electrode
Lve cm cn ohm/m.
tVavg pos. Neq. poa. tieg. & 300 K
71.4
63.5
60.8
62.3
61.9
60.3
74.6
69.2
64.0
63.2
66.1
74.5
74.5
61.2
60
57.4
83.1
74.1
53.4
68
74.1
0
3.2-6.4
3.2
0
0
0
0
0
1.6-3.2
1.6-3.2
0
0

12.7-19.1
6.4-12.7
0
0

12.7-19.1
0
0
0
3.2-6.4
3.2-4.8
0
0
0
0
0
1.6-3.2
1.6-3.2
0
1.6
3*2-6.4
6.4-12.7
3.2-6.4
0
0

50.8-76.2
0
0
0
50.0
57.2
63.5
76.2
47.6
0
0
44.5
44.5
50.8
0
0
47.6
54
57.2
0
0
57.2
63.5
0
0
57.2
59.1
69.9
76.2
50.8
0
0
44.5
54.0
50.8
0
0
47.6
54
57.2
0
0
57.2
66.7
0
13.1X10"3






36.1X10"3
H M
II M
n •
N N
29.5X10"3
N N
M H
II H
M M
131X10*3
N N
N If
II N
CO
U)

-------
                                     DISCHARGE ELECTRODE OSCILLATION  RESULTS
                                                      (Cont.)
SPA
Electrode W
Electrode W Length Wt/Length Temp Pos
Description Ratio m. g/m K kVp
E. 6.35 nun
Stranded Cable
(Non-Magnetic)





F. 2.77 mm
Round
(Non-Magnetic




G. two-1.59mm
with 19mm barb:
Barbed wire
9.1
7.1
7.1
4.3
7.1
4.3
9.5
9.5
9.5
29.1
25
29.1
38.7
38.7
30.9
30.9
7.5**
14.3***
L4.3
14.3
.31
.31
.31
.31
.31
.31
.20
3.20
3.20
4.30
4.30
4.30
3.20
3.20
3.20
3.20
4.32
4.32
4.32
4.32
172.9
172.9
172.9
172.9
172.9
172.9
172.9
172.9
172.9
49.53
49.53
49.53
49.53
49.53
49.53
49.53
87.31
87.31
87.31
87.31
304
305
866
866
1033
304
300
1200
11-14
302
303
866
304
1061
302
1089
300
300-303
866
922
57.1
65
56
65
56
55
64

66.1
60
59.5
65


45
50
50
55
RKING VOLTA
IRE POLARIT
itive Negat
kVavg kVp :
47.9
47.1
41.4
47.1
47.5
47.6
45.6

50.9
46
•17.7
47.1


34.2
37.4
36.0
39.5

100
85
85

100

99.1

100
60


110.6
90

80
75
75
Change In
GES Electrode Electrode Resistance
Y Oscillations Length °E Electrode
ive cm cm ohm/m.
tVavg Pos. Neg. pos. Meg. 6 300 K

73.7
til. 6
50

74.6

68.8

72.2
43.7


81.8
65

57
52.9
52.4
0
6.4-12.7
38.1-50.8
3.2
3. 2
0
V


50.8
76.2-101.6
9c_ in i
• y L y • J.


0
f\
101.6-127
25.4-38.1
o
19.1-25.4
12.7-19.1
1.6




76.2-127

19.1-25. 4
12.7-19. 1


3.2
3.2
0
57.2
41.3
73.0


73.0

C.
47.6
0
i4


0
0
41.3
41.3
0
0
54
66.7



66.7

0
50.8


0
57.2

0
41.3
41.3
194X10 3
H M
I C4
1 II
M t
II II
•* t
K M
H M
272X10"3
II H
1 1*
1 M
H II
II M


*W ratio = 5.0 for positive polarity
'•Electrode weight not sufficent to hang straight
**'Weight exceeded yield stress point

-------
Inconei 601 produces an oxide coating 0.10 mm deep while being
used at temperatures as high as 1400 K.  The coating does not
appear to change the wire diameter.  Inconei 601 proved itself
acceptable for use in these tests.  Further experience under
actual field conditions is required for long-term evaluation.
                              35

-------
                   40  10  100  IK l«l

                    VOLTAOf.kV
                                                          a  =
                                                         \ VV. I
                                                       120  UO  Ifiu
                                                                              to  to
                                                                           VOUTA«C,kV
                                                • 0 «0  100  IJU

                                               VOLTAIC,W
 •O  IOO  Illl

VOLTAIC. kV
Figure 10-   Current-Voltage  Curves  Taken  in Dry  Air at  Temperatures of 294  K,  533  K,
              and 950  K for  a  2.344 mm Wire Electrode.

-------
Ui
                              1-ISItB..
                              roniivt
                            40  U  10
                             VOLTAM.IV
                    3 '

»
n —
to

AIR ONI
•II" *
ftMtmm
1
40 10 to IOO 110 141
VOUASl.kV
                                                 ft )44u
                                                 NE4AIIVC
                                                 10  40 60  10  IOO  ItO
                                                       VOLTAIC,IV
                                                                              40  40  10  IOO
                                                                                VOLTAtt.tV
<0  40 100 IIO
 VOLTAIC.kV
        Figure  n.  Current-Voltage  Curves  Taken in Air at a  Temperature  of 811  K for
                     Wire Electrodes  of 1.575mm,  2.344mm, and  3.175mm.

-------
U)
CO
                  c
                  5 »l
                         10   40   10   10   no   no
                              VOLTAOC.W

-------
vo
                                                                        VOL1A*(,IV
                                                                              1  I - . I ..
                                                                              ItO  140 UO
 Figure .13. Current-Voltage Curves Taken in Air at a Temperature of 1366 K for Wire
             Electrodes  of 1.575mm, 2.344mm, and 3.175mm.

-------
    COUtUttlUM *.*»
  -    »*   ')$,?'$"
 .2 .  A —Xjkida!!!.
o  eu  «io  io  to  too  (to
      VOLTJtt.hV
                               i.
                               I,
to  «o  to  10  i to  110
   VOlfAOI.kV
                                                               coiuumu* uk

                                                               «•! I4«««
                                                                     «0  M
                                                                     VOLTAIC. IV
                                                                           •0  100  ItO
                                                                                             COUKUftllOH AAI
                                                                                             !}••••
                                                                                             4* 1 !€<••
                                                                                                      •0 100 i/u
                                                                                                 VOLTAOC.kV
                                      •0  10  100  110  140
                                        VOlli4(.tV
                                                                                                             h
                                                                         -.1 - .4	
                                                                 40  ftO  «O 100 1(0  I«O
                                                                  VOIT44I. IV
                                                               •o  too
                                                            ₯OlT«*t,IV
l-'igure   14.  Current-Voltage Curves  Taken in a  Simulated Combustion Gas Mixture for
               Temperatures  of 533 K,  Oil  K,  1089 K, and  1366  K Using a 2.344mm  Wire
               Electrode.

-------
                               HI,	,	
                                40  M W  100  110 I4U
                                  VOLUM.tV
                                                        VOLTAd.kV
40  *0 10
VOL1*g(.lV
Figure .15.   Current-Voltage Curves  Taken With  a  Substitute  Fuel Gas Mixture at
             Temperatures  of 533 K,  811 K, and  1366 K for a  2.344mm Wire  Electrode.

-------
rf*
10
too
1*4,1

1*0

> no
. no
hi
• 100
*j K>
> *O
4U

10
o

r i i r ~f ~T i i
AIM UNI
«»«••
*•* »44»«

HlHAllVt *
«
B
•
• •

» *
• „
V
1 1 1 1 L 1 1 1
' 1 i t 4 > 1 I ,
"ill • r i - 1 r 1 — n r i "r
AIM DNt a
»»•• <*
4- ? I44*»
MIAIIVI *

9 * • • . *
o e
^ • 0 0 0
• °
•5

* u
*l 1 1 1 1-1 11 1 1_ J I 1 II..
DIHIITT
T~r~ iii ii i - 1-
*
•
• ••
•
u
•




• ' ' i i i i-i-i i - 1 .1 a i i i
• IU II u 11 14 1) Ik If II IV to H f| 1} 14













r T i i rT^i
c


• . • •






L 1 1 1 1 II
>ft M IF }• 3« Ju 31 S


IttU

> 140
". 110
bl
.,««
s »
S .0
4O

III
O

1OU
I»U
IfcU
^140
"• IZU
^1
• tou
4
t*
5 *u
a
> »o

4O
2U
U
c
-T — i— r i i i i r"i" T ~i — i T ,i" i
AIM DMV Q

4-1 >44««i
POII1IVC
m
. . ' • '
•
* *
•
..••
* • *
t •
•
_j__i_i. i j i I i. i i - 1— i- * * »
I 1 1 1 4 ft 4 1 • » Kl II U U 1* IS **
DENSITY
I ~\ \ \ III 1 T r"l "1 1 t
AIH UNf J
roiinvi


•
•
•
• •
. •- • •
• •
• ••
m
-JJillllllllJlll
1 > 1 4 ft • r • » IO U U II 14 Ii H,
                                                                   OCNIIT*
    Figure   16.  Sparking and Corona Starting Voltage  in Air as a Function of
                Relative Air Density at 294 K and  533 K.

-------
        • HO
           •It'll
           fll»
           MOAllvl
                    ,140



                    -.00
                        •I* Ml


                        MIIIWC
                                               . I
                                  . J I L I I  I I III I

                                        OfNIItV
ruo
no
I»U
» I4U
m
. tlu
5 ^
J W
O
40
ID
0
0
- i -i t i t i - r i - i -i — i -i • -i — i
AIM BUI
•"'• C

NIGftllVI § *
• ."••••'
•

• .**""*••
. •
t •
I 	 	 .,..,_._

cuo
IMJ
140
> HO
• ItO
5 wo
tj W)
0
40
iu

«
— i--i-n ~i \ r-i t-i i i-r-
. Ml Ml ,
iirii Q
4* 1.14.4BA1
miTivc
m
% * "' *

c * *
J*
• J*

BIMtllV
IUU
wo
ItO
HO
• 110
i-
ti *>
e ..
40
to
U

KTo.'." e
* 4*1 144m *
•(Mil VI •
• . '


* . * "
o
. t
1 1 1 1 1 1 1 1 1 1 1
BIMtlll
IW
110
wo
140
". 110
S»o
5 M
1..
4O
to
0

jivr f
• «*II44M


. " . •

•
• •
* *
. 1

0KNSITV
         rtu

        jMO

        • U4I
       i-
       > M
                 I "I - -|— I" f
fltllMI

M4>ll«l
                   • 100

                   & •«
                DC HSU T
                       10 II 11
                                               h
                          I—I--L I .1.. I. 1 i  l_
                          I4t*rti«iiii
                            MNIITV
Figure  17   Sparking and  Corona Starting Voltages in  Air as a  Function  of the
             Relative Air  Density at  811 K and 950 K for Wire Electrodes of
             3.175mm, 2.344mm, and  1.575mm.

-------
too
wo
1*0
> "u
" no

tf tuu
J •«*
o
4U
»0
o

JOt*
(•U
,140
".UU
JJlOO
^>
-J MU
O
> 40
40
JO
u
1
ItO
IbU
I4U
* KU
u
• IOU
5 •»
2 .«
40
 Hu
".110
M
* IUO
2 10
S.O
.0
^0
«
j i
IIU
, 140
.I2O
• 100
J 10
S .u
4U
to
a
j t
To
I.O
> ""
". ""
S m
i::
4U
IU
(J
0
~"i' ' 11 i" r r"iii — i
AIM UHI |J
lO««*K
f 1 IfftMM
rosinvt .
• •*•.•

a •
• • • •
• •
. «
;
ii i i i .i . i i_ i_
OtNllTV
-ir i- i r i~ i — i — i —
«in uht p
'"'•'*""


• * *f
• m - "
• „ «
. | "
S_j- .1.1 i i t- I .1 -i_
oiNiiri
1 1 - 1 1 1 1 1 - 1 . I
AINtHl ~\
IU.**»
<•»!>. •
fuiitivl


• ". ; : : • : '

! 1

Of Hill >
too
••u
wo
, 140
" no

• no
u «o
o
^ BO
40
to
o
1
>uo
uo
> 140
t| HO
JJ too
ti .0
o •
> 40
4O
IO
k O
0
YUI)
l<0
1*0
>(4U
" »*0
• IOU
a
40
*o
o

""Wo*; •|~t"^c~
ii«ft*K
4* 1 Ifftw* •
NlflAllVK *


• *
. .*
• u
• . • *
• 0
t_l 	 1 	 L—J — 1. L_J 	
DIUIITV
- r i - 1 i-i . i i
>wiwi CJ
«I.",V ' •
•*
•

* •• •

% *
tLl-l ) 4 1 1 1 I
oimiu
til!" ^
* *

•

A
„ „«
• w
• II *
„ *
i! . . ... . .
OhNtlTY

































tot
IM
I«L

M.na
M
• lOfi

> 40
40
to
c

too
uo
14(1
J l«
* 10(1
J «U
* 4O
40
to
u
0
too
I4U
140
120
100
•U
• 0
40
IO
o
1 0
~Wil -i-T-g-r-

4*nr»»*
. POSIIIVI
(«

*
•

• B • •
• *


MMSI1V
«m on* K
wlnJT"

• " . •
a
a
I • "
1 « •
j ? . t j j i i_
l>S4ftft?t
BfHIlII
kin OHI -L
lltft'k
• •*itr»«»
"""""

• * *m .
. "*
1 * "
t 5
^_i_ i i i i i . i -
filHIIIl
Figure 18.
Sparking and Corona Starting Voltages in Air as u Function of
Relative Air Density at 1089 K and 1366 K and for Wire Electrodes
of 3.175 mm, 2.344mm, and 1.575mm.

-------
AMI
4BU
ItU
HO
III*
IO*J
•t)
• U
• U
 140
• 120
-«*
5 •«
9..
40
>0

u 1
-n-r-T-i-1 r^i--
. cawmiNM ui *-•
IIM'I
<•! I44«« .
M4A1IVI •
.
. *
* •_
•:.* "
. i
1 1 i i i i i i i
OtHIITT
          '•U

         > <«<
Ul
             »•>••
             «-i 1*4..
               I  I I I  I 1 I . I I
               1  4 S 4  » ( •  10 II 12 It 14 rt
                   0[NIIT»
1

e



•






_j _
(IM
I.H1
ItU

> 140
u lto
4! IU°
S «°

> to
4U

to
".
"i vri i"i i r • i i i"T~
COHtUSlKIH «*»
•It'll p
PUSltlVI


•
«*•*"•

.
•
1

i.l_L -i i t j i i _j_ i. i a —
no
)M
140

> HO
2««
5 M>o
J §0
o
> to
40

10
U
                                                             COUIIJtIK
                                                             lOtt'lt
                    1 > 4 > « I • » ID II II II
                        DtHIITT
 _J- I J	L-l -I _l- I -I
oit>4itrii
     •CNtlTV

>
jj
5
a
>

a
ton
IM
MO
140
110
•0
to
40
to
«,
- <*t>44»
rotiTivt

! .''.*'
. * .
t * n
LJ — i_i i J_i. a-L.
      Figure  19.
Sparking and Corona Starting Voltages  for a Simulated Combustion Gas
Mixture as a Function of Relative Gas  Density at 533 K, 811  K, 1089 K,
and 1366 K for a  2.344mm Wire Electrode.

-------
      • HM»
      5 •
      §..
          II I i  i
          tMkfltult I ul
                   *ou


                   IBU


                   1*0


                   *UU
i~r i  i i T
Mttlllull I Ult
III* II
4-1 I44-. •
ftCOAllVt
                                                 I I  I
                                                 • 1  10 II I]
100
MO
MO

I4O
*.IK
*»
.1 BO
e>
> *o
« °l
-T~I- i T i r i i r
- tuitluulf run
tun** <-•
ft 144.. **
Mf OftllVI

.
.. • '
• o i* ° u
* 0
• • »
» *
. «
1 -t J i i i i i i i
too
I«O
MO

> ****
•no
M
.,00
in
4O
10
0
*
•~i~i i i i r i
_ ftUlllllhll *U4l _
IUfK (J
- <• I S44M«
HCflAllVI




• «* "
l_V_i 	 1 	 1 1 _i_ I ._
en
                         i i i  i
                                     « IUU

                                       .u
                                         T—r i -i—i -r"i—i -r~i"i
                                         IUIIIIIUII IUII     _
                                         •in          t
                                        _J . I I I  I. J J_ L I l_.
                                        uiii«s»rii^iii
                                              OINSIIT
- r T i - i r i i it
ISTt-'l"1" ""l g
IV" *^ •*
rauiivi


• •::•••
• u
• M
I
* 1 1 1 J I i 1 1 1
1 ' DiNim
too
l«0
IftO
*40
m, ito
• lt>O
i •*
40
III
U

~r r i r i . r
_ luvlfitull run l^
roiifivi


""!"*' :
. . •
- 1 •
LI — t_i 1 1 i i
OfNtlTV
    Figure  20.
Sparking and  Corona Starting  Voltages  in a Substitute  Fuel Gas Mixture
as a  Function of Relative Gas Density  at 533  K, 811  K,  1089V K, and
1366  K for a  2.344mm  Wire Electrode.

-------
B/rt
IflS
       14
       12
       10
                                 -1	r
        0.1
                                                             e294-588°K

                                                             °?5->-H440K

                                                             *I144-1 166°K
0.2     0.4  O.f> 0.8 1.0      2

                        r«
                                                        6  « 1.0
                                                                     20
                                                                             40    60 BO 100
       Figure 21  Corona Starting Field in Air  for Negative Polarity and
                   Compared with Peek's  Equation for m =  1.0, 0.8,  and 0.6.

-------
CD
        14
        III
    1 0° V  <>
          u. I


                        I).',   II. 0 0.8 1.0
                                                      b   H  10
                                                                   . i

                                                                   20
40   60  80  100
           Figure 22  Corona Starting Field in Air for  Positive Polarity
                      Compared with  Peek's Equation for m = 1.0.

-------
                             SECTION 6

                     COMMERCIAL COST ESTIMATE


PRECIPITATOR SIZE FOR FLOIDIZED-BED COMBUSTION

     The estimation of a practical size for a high-temperature,
high-pressure electrostatic precipitator to clean the gases from
a commercial fluidized-bed combustion unit follows seven steps.

     1.   Determination of the relationship between the gas flow
          rate (m /s) and power rating (MW).

     2.   Determination of the size of a commercial unit and its
          division into boiler modules.

     3.   Determination of the number of precipitator pipes
          required to serve each boiler module.

     4.   Determination of the vessel size for the given number
          of pipes and sub-multiples of that number.

     5.   Determination of the vessel thickness and estimated
          weight for each size.

     6.   Determination of the transformer-rectifier capacity
          required to serve each size.

     7.   Selection of the most practical size or sizes.

     The details of these steps follow.

     1.  The relationship between the power rating and gas flow
rate is derived from data for two pilot plants and one demonstra-
tion unit.  These three sets of figures provide an estimated re-
quirement of 0.661 m /s or 560 ACFM per each megawatt of generating
capacity.  See Table 6.

     2.  Two sizes are estabished for commercial fluidized bed
combustion units, 300 MW and 600 MW.  Since the 600 MW unit is
double the size of the 300 MW unit, the 300 MW unit is considered.


                                49

-------
Any requirements for this size can be doubled for the 600-MW
unit.  The 300-MW unit consists of four boiler modules and two
gas turbines.  Each boiler module consists of four stages of
fluidized beds with embedded boiler tubes.

     The logical arrangement is to have one or more precipitators
serve each boiler module.  The capacity per module is 75 MW.
The estimated gas flow is:

     75 x 0.264 =19.8 m3/sec.  (Actual)

or   75 x 560 = 42,000 ACFM

     3.  The gas flow capacity of each precipitator pipe is
based on a treatment velocity of 1.52 m/sec.  The pipe size is
assumed to be 0.229 inside diameter x 3.05 m.  The cross-section
of 0.229 m pipe is 0.041 m.  Multiplying by 1.524 m/sec equals
0.0625 m3/sec per pipe.  A gas flow rate of 19.81 m3/sec requires
317 pipes.  The relationship between the number of precipitators
per boiler module and number of pipes per precipitator is shown
in Table 7.

     The expected efficiency of the precipitator(s) is 98 to 99
percent, based on a migration velocity of 0.122 m/sec.  This
migration velocity is estimated from the results of this report.

using the equation

                   E d E
               w =  o p av
                      y

where:

     w = migration velocity  (m/sec)

    e  = permittivity of free  space  (F/m)

    d  = mean particle diameter  (m)

   E   = average electric field in pipe  (V/m)

     y = gas viscosity (decapoise)

At conditions of 1089°K and 1000 kPa, the gas viscosity is ap-
proximately 4.1 x 10~5 decapoise, and from Figure 19f, the clean
gas maximum field strength can be calculated to be about
1.5 x 106 V/m.  If it is assumed that in the operating precipitator
only half of this field strength can actually be realized, and
that the mean particle size is only one micron, the indicated
value of w = .122 results.  It is felt that this estimated
value of w is, therefore, conservative.

                              50

-------
     4.  In order to determine the size of the vessel required
for the various numbers of pipes, it is necessary to establish
specific parameters.  First, it is decided that 0.229-m  diameter
pipes require 0.305 m spacing, center to center.  A minimum of
0.178 m is established as the clearance from the center of a
pipe to the inside of the refractory lining of the vessel.  This
distance is needed for electrical clearance.  The thickness of
the refractory is estimed at 0.229 m.  Preliminary design calculate
determine that the steel thickness of the pressure vessel is
approximately 5.21 mm for each meter of vessel diameter.

     Using these parameters, a simple computer program is used
to find the relationship between vessel diameter and number of
pipes.  See Table 7.

     5.  The relationship of 5.21 mm of steel thickness to one
meter of vessel diameter is based on an allowable steel stress
87,220 kPa (1).  An internal pressure of 10 atm or 1010 kPa is
used for design.  The steel temperature is assumed to be no more
than a few hundred degrees.

     The estimated area of the vessel is based on a cylindrical
shell, 4.88 m high.  The 4.88 figure allows for the 3.05- m
pipes plus 0.914 m top and bottom.  The top is assumed to be a
dished head.   The bottom is conical with a slope of approximately
56 degrees with the horizontal.

     6.  The size of transformer-rectifier units plays a large
part in establishing the practical-size precipitator.  Most
precipitators can be divided into small bus-sections and can be
energized by any number of transformer-rectifiers.  However, be-
cause of the high temperatures, these precipitators present
special problems.

     The high-voltage system is supported from insulators which
are, in turn, supported by the pressure vessel.  The collecting
pipes are supported by a header plate inside the precipitator.
When the precipitator is in operation at high temperature the
header plate and wire support frame expand.  The pressure
vessel does not expand.  Since the header plate and the wire
support frame are intentionally made of the same material, it
can be assumed that they will expand a like amount.  However, if
the wire support frame is supported from the insulators in an
unsymmetrical manner, it will shift as it expands.  This shift
will throw the precipitator out of alignment.  The discharge
electrodes will no longer be in the center of the collecting
pipes and the collection efficiency of the precipitator will be
seriously impaired.  As long as the wire support frame is sym-
metrically supported, it should not shift.  It the wire support
frame is divided into two or more parts, it must be unsymmetrical.


                               51

-------
     Based on the above, it is necessary for the precipitator to
have a single wire support frame and, thus, be energized by a
single transformer-rectifier unit.  It is estimated that the
current requirement per unit length of discharge electrode is
approximately 3.28 mA/ra.  A precipitator with 317 3.05-m long
pipes requires 3170 mA.  The voltage is 150 kV peak or approximately
100 kV rms.  With 3170 mA and 100 kV the power is 317 kW.

     7.  Almost any selection process is a process of elimination.
First, the impossible choices are eliminated to obtain a list of
possible alternatives.  Table 7 represents this list.  Next, the
more undesirable choices are eliminated.  This elimination is
based on economics.

     It can be assumed that 317 collecting pipes and discharge
electrodes will cost the same, whether they are all in one pre-
cipitator or divided into many precipitators.  Some elements,
such as the header plate, are more costly in a large precipitator
than in several smaller ones.  Other items, such as the support
insulators and insulator compartments, have a fixed cost per
precipitator and thus increase as the number of precipitators
increases.  Up to about 3.6 or 4.0 m, a pressure vessel can be
shop-fabricated.  Beyond this size, shipping would become a pro-
blem.  This factor works in favor of many small precipitators.
While not a precipitator cost, the cost of the gas conveying
systems affects the choice.  In the absence of detailed informa-
tion, it is not possible to be specific about these factors.  It
is estimated that the pluses and minuses tend to balance each
other, with the fewer large precipitators having some advantage.

     The one factor that can be estimated with some degree of
accuracy is the weight of the pressure vessel.  See section 5
above.  The weight of each vessel is multiplied by the number of
precipitators required.  This figure is shown in the fourth
column of Table 7.  This determination combined with other
higher costs eliminates the last four choices.

     The choices are than narrowed down to two.  The costs of
these two sizes are estimated.  The recommended size is one
precipitator with 320 pipes.  The two 160-pipe precipitators are
in alternative selection.  The details of these two selections
are shown in Table 8.

MATERIAL-TEMPERATURE  DEPENDENCE

     At elevated temperatures, a small change in temperature can
make a large difference in material selection.  Characteristics
that have no importance at moderate temperatures become critically
important at high temperatures.  These factors are also time-
related factors which might not be sigificant on a short-term
                              52

-------
pilot plant but would be very significant on a long-term com-
mercial precipitator.  Strength is also very drastically affected
by temperature.  Metals have a fairly uniform strength curve
until a certain temperature.  Above this temperature, the strength
drops sharply.  Thus, a slight increase in temperature could
require much heavier structural sections.  Thermal expansion
also becomes very significant at high temperatures.  At ambient
temperature, it is frequently desirable to use different materials
for different components.  However, at high temperatures,
slight differences in expansion rates could cause high stresses
in materials that are welded together.  If components are bolted
together, different materials can be used.

     The components are divided into two categories.  The first
category are those components where structural strength is less
important but where scaling is the prime consideration.  These
components include the collecting pipes, the discharge wires,
and the liner for the hopper bottom.  The second category are
those components where scaling is less significant but where
structural strength and creep rupture are the prime consideration.
These components include the header plate, the high voltage
support system, and the pipe rapping frames.

     The basic categories of materials are considered:  stainless
steels, andnickel and chromium alloys such as Inconel.  The
stainless steels are less expensive than the Inconels and, thus,
are preferred where technically acceptable.  However, the strength
of these materials begins to drop sharply at 1089 K.  Above 1144
K their strength is so poor that the savings per pound would be
offset by the added quantity required.  For this reason, stainless
steels are recommended only up to 1144 K.  Type 310 stainless is
selected for the first category of components as it has better
resistance to scaling.  Type 316 stainless is selected for the
second category of components because of its better creep rupture
strength.

     Above 1144 K up to 1311 K the use of Inconel is recommended.
Inconel 601 is selected for the first category of components be-
cause it has good resistance to scaling.  Inconel 617 is selected
for the second category of components because it has the best
creep rupture strength.  Above 1311 K, the selction of materials
becomes a very difficult problem.

     The costs of precipitators in Table 8 is divided into
Alternate A and Alternate B.  Alternate A comprises precipitators
with stainless-steel internals for operation up to 1144 K.
Alternate B consists of precipitators with Inconel internals for
operation from 1144 K to 1311 K.  These costs are for equipment
only and do not include erection.
                         53

-------
Table 6
Unit
Power Rating (MW)
Temperature (K)
Pressure (kPa)
Gas Flow (m /s)
(SCFM)
(ACFM)
Gas Flow per MW
Cm3/s)
(SCFM)
(ACFM)

Precipitate rs
per Module
1
2
3
4
5
6
Combustion
Power Exxon
1 0.63
1089 1089-1200
410 1000
1 - 1.5 0.3 - 0.4
2120-3180 636 - 848
1960-2940 235 - 346

1 - 1.5 0.48 - 0.63
2120-3180 1009-1345
1960-2940 373 - 549
* Assumed
Table 7
Pipes per Vessel
Precipitator Diam.
317 7.0
159 5.2
106 4.4
80 4.0
64 3.7
53 3.4
American
Electric Estimated
Power Commercial
60
1172* 1172*
1000 1000
40
84,744
33,738

0.67 0.66
1412 1400
560 560

Weight of
(m) Vessels (10 kg)
52.8
51.4
53.3
55.0
57.4
56.6
      54

-------
                           Table 8
Number of Precipitators
per Boiler Module
Pipes per Precipitator
Pipe Size
Vessel Diameter
Vessel Height
Steel Thickness
Vessel Weight
Capacity m^/sec
         ACFM
         SCFM
  MW of plant
     capacity
Collecting surface
area
Discharge
electrode length
SCA -^7i5o'
    tag	,
     10" ACFM
Expected efficiency
Power Supply
Voltage
Current
Recommended
1

320
 0.229  x  3.05m

7.0
10.7
0.038
52.8 x lO^g

19.82
42,000
106,000

75

700.5m2

975.4 lin. m.
35.3

180
98-99%


150 kV
3000 mA
Alternate
2

160
0.229 x 3.05m

5.2
9.1
0.029
25.7 x lO^-g

9.91
21,000
53,000

37.5

350.2m2

487.7 lin. m.

35.3

180
98-99%


150 k.V
1500 mA
                                 55

-------
                              Recommended         Alternate


mA/m Discharge                3.08                3.08
Electrode

    2
mA/m / Collectincr

Surface                       4.31                4.31


Watts/m3/s                    1.462 x 104         1.462 x 104


WattS/ACFM                    6.9                 6.9


Costs


Per precipitator - less engineering


     Alternate A              $1,945,000          $1,020,000


     Alternate B              $2,183,000          $1,170,000



Per 300 MW Plant - includina enaineerina

     Alternate A              $7.980.000          $8,360,000


     Alternate B              $8,940,000          $9,560,000



Per KW of plant capacity


     Alternate A              $26.60              $27.90


     Alternate B              $29.80              $31.90
                                 56

-------
                             SECTION 7

                            REFERENCES
 1.   American Society of  Mechanical  Engineers,  Boiler and ,
     Pressure Code,"  Section VIII, p.  102,  19

 2.   Blum,  F.  L.,  "Estimated Size of a Commercial Unit for
     Fluidized-bed Combustion,"  Research-Cottrell, Inc.,
     Bound  Brook,  NJ,  1977.

 3.   Brown, R.  F.  and Walker. A.  B.. "Feasibility Demonstration
     of Electrostatic Precipitation  at 1700 F"  J. Air Pollution
     Control Assoc.  21, 617-620  (1971).

 4.   Clavert,  S.  and Parker, R.,  "Collection Mechanisms at
     High Temperature and Pressure," Symp.  Particulate Control
     Energy Proc., Document  EPA-600/7-76-010, NTIS,  1976.

 5.   Calvert,  S.  and Parker  B.,  "Effects of Temperature and
     Pressure on  Particulate Collection Mechanisms:   Theoretical
     Review,"  Document PB-264 203, NTIS, 1977.

 6.   Cooperman, P.,  "Spontaneous lonization of  Gases at High
     Temperature,  "Conference Paper  63-173, Amer. Inst. Elec.
     Engrs., 1963.

 7.   Cooperman, P.,  "Positive Polarity Operation of Electrical
     Precipitators,"  Conference  Paper 63-172,  Inst.  Elec.
     Electron.  Engrs., 1963.

 8.   Cooperman, P.,  "Spontaneous lonization of  Gases at High
     Temperature," Paper  ES-MON-6,  Inst. Electron. Engrs., 1971.

 9.   Cooperman, P.,  "Positive Polarity Revisited," Symposium on
     Changing Technology  in  Electrostatic Precipitation,
     Inst.  Fuel (Australian  Membership), Adelaide, Nov. 1974.

10.   Ekert, E.  A.  "Conditions Which Influence  Spark-Potential
     Values," Sibley J. Eng. 18, 391ff. (1904).

11.   Hall,  H. J.,  Brown,  R.  F.,  Eaton, J. B.,  and Brown,  C.  W.,
     "Removal of  Lube Fume Raises Line Efficiency," Oil Gas
     J. 66   (37),  109-117  (1968).

                              57

-------
12.  Howell, A. H., "Breakdown Studies in Compressed Gases,"
     Trans, Am. Inst. Elec. Engrs. 58, 193-204  (1939).

13.  Hunting-ton Alloys, Inc., "High Temperature Alloys - Hot
     Performance Guide, " Publication JN-159, 1975.

14.  Roller, L. R. and Fremont, H. A. "Negative Wire Corona at
     High Temperature and Pressure," J. Appl. Phys. 21,
     741-744 (1950).

15.  Loeb, L. B. University of California, Private Communication,
     1967.

16.  Lowe, H. J. and Lucas, D. H., "The Physics of Electrostatic
     Precipitation," Brit, J. Appl. Phys. Suppl. No. 2,
     S40-S47 (1953).

17.  Peek, F. W. Jr., Dielectric  Phenomena in High-Voltage
     Engineering, 3rd ed., McGraw-Hill, New York, 1929, pp. 78,81.

18.  Rao, A. K., Schrag, M. P. and Shannon, L. J., "Particulate
     Removal from Gas Streams at  High Temperature/High Pressure,"
     Document PB-245-858, NTIS, 1975.

19.  Robinson, M., "The Corona Threshold for Coaxial Cylinders
     in Air at High Pressures," IEEE Trans. Power App. Systems
     86, 185-189  (1967).

20.  Robinson, M., "Electrostatic Precipitation" in Air Pollution
     Control, W. Strauss, ed., Vol. 1, Wiley-Interscience,
     New York, 1971, pp.  227-335.

21.  Robinson, M., "Critical Pressures of the Positive Corona
     Between Concentric Cylinders in Air," J. Appl. Phys. 40,
     5107-5112  (1969).

22.  Robinson, M., Unpublished data.

23.  Robinson, M. and Brown, R. F., "Electrically Supported
     Liquid Columns in High-Pressure Electrostatic Precipitators."
     Atmos. Environ. 5. 895-896  (1971).

24.  Rvan, H. J., "Conductivity of the Atmosphere," Siblev J.
     Enq. 18, 267ff. (1904).

25.  Shale, C. C., "Electrostatic Precipitation at Hioh Tem-
     peratures and Pressures," Paper 64-8, Air Pollution Control
     Assocl, 1964.
                                56

-------
26.  Shale, C. C., "The Physical Phenomena Underlying the Negative
     and Positive Coronas in Air at High Temperatures and
     Pressures," Int.  Conv.  Record, Inst. Elec. Electron. Engrs.,
     Part "I, 1965.

27.  Shale, C. C., "New Concept of Electron Detachment for Air
     in Negative Corona at High Temperature," U.S. Bureau of
     Mines Information Circular 8385, 1967.

28.  Shale, C. C., "Progress in High-Temperature Electrostatic
     Precipitation, J. Air Pollution Control Assoc. 17, 159-160
     (1967).

29.  Shale, C. C., Bowie, W. S., Holden, J. H. and Strimbeck, G. R.
     "Feasibility of Electrical Precipitation at High Temperatures
     and Pressures," U.S. Bureau of Mines Report of Investigations
     6325, 1963.

30.  Shale, C. C., Bowie, W. S., Holden, J. H. and Strimbeck, G. R.
     "Characteristics of Positive Corona for Electrical Pre-
     cipitation at High Temperatures and Pressures," U.S. Bureau
     of Mines Report of Investigation 6397, 1964.

31.  Shale, C. C., and Fasching, G. E., "Operating Characteristics
     of a High-Temperature Electrostatic Precipitator," U. S.
     Bureau of Mines Report of Investiaation 7276, 1969.

32.  Shale, C. C. and Holden, J. H., "The Role of Wire Size in
     Negative Electrical Discharge at Hiah Temperature,"
     IGA Group, Inst.  Elec.  Electron. Engrs., 1968.

33.  Thomas, J. B. and Wong, E., "Experimental Study of  dc Corona
     at High Temperatures and Pressures, " J. Appl. Phys, 29,
     1226-1230 (1958).

34.  United States Steel Corp., "Steels for Elevated Temperature
     Service," Dec., 1974.

35.  Walker, A. B, and Frisch, N. W., "U.S. Design and Operation
     Experienced with Hot Electrostatic Precipitators on Coal-
     Fired Utility Boilers," Joint US/USSR Symposium on Control
     of Particulate Emission and Related Problems, Moscow, 1976.

36.  Winkel, A. and Schuetz, A., "Electrical Separation of Finely
     Dispersed Iron Oxide Dust at High Temperatures with Staub 22,
     343-359  (1962).

37.  Zabolotny, E. R.  McCarthy, D. J., Korosi, A., Kurh, R. and
     Trueblood R. C.,  "Purification of Hot Fuel Gases from Coal
     or Heavy Oil," Document EPRI 243-1, NTIS, 1974.
                               59

-------
                             SECTION 8

                            APPENDICES


A.   Calculation of Thermal Conductivity

     The thermal conductivity of a gas mixture can be estimated
from the conductivity of each pure component, its molecular
weight, and its mole fraction in the mixture according to the
equation:
Values of the thermal conductivity for the pure components at
300 K have been obtained from the Handbook of Chemistry and
Physics, section E2.  Higher temperature values can be estimated
using the reduced properties chart available in Perry's Handbook
of Chemical Engineering, page 3-245.  Values of thermal con-
ductivity for the gas mixtures used are thus tabulated as follows:
                                   60

-------
                             TABLE A-l

                EVALUATION OF THERMAL CONDUCTIVITY

                       FOR TEST GAS MIXTURES
c uco. ^uiuuua <
Component M. y. y .
C02
°2
N2
He
H20
Air
44
32
28
4
18
(28.8)
0.230
-
0.0535
0.185
0.05
-
0.
0.
0.

0.

092
028
830
-
05
—
L^UJJ
k300
19.
30.
30.
173.
20.
29.
30
6
0
4
5
9
k530
41.
50.
46.
253.
44.
-
4
0
2
0
7

k800
58.2
67.3
62.0
353.0
78.4
—
knoo
74.6
79.8
77.4
447.3
-
_
k!350
87.1
95.2
89.0
523.3
-
_
Fuel Gas
 Mix     (26.8) (1.00)   -        41.1    65.9   91.4  115.9   134.7

Comb. Gas
Mix      (29.1)   -    (1.00)      28.4    45.8   62.5    77.0    89.0

A low Btu fuel gas would have a value calculated as below:


                             TABLE A-2

            THERMAL CONDUCTIVITY OF A LOW-BTU FUEL GAS
Component
co2
CO
H2
CH4
N2
H20
Fuel Gas
*i
44
28
2
16
28
18
(24.8)
Ii
0.087
0.174
0.124
0.033
0.482
0.10
(1.00)
k300
19.1
28.8
214.7
39.4
30.0
20.5
38.3
                                 61

-------
B.  Calculation of Ionic Mobility

     Estimated values of ionic mobility b can be calculated
using the equation:
Values for the pure gas components have been taken from table 3 of
Strauss :  Air Pollution Control Part I, page 240.  The following
table summarizes the calculations for the gas mixtures tested
and for a low Btu fuel gas.
                             TABLE B-l

                         IONIC MOBILITIES
Actual
Component Fuel
He
N2 0.482
°2
C02 0.087
H2 0.124
CO 0.174
H20 1.10
CH4 0.033
Air
Fuel Gas (1.0)
Substitute
Fuel Mix
Com. Gas
Mix
Substitute 2
Fuel Combustion (m /(s-V)^
y. y. b ~ b
Ji •* 3. O O
0.185 - - 16.7xlO~4
0.185 0.83 2.27xlO"4
0.028 2.46xlO"4
0.230 0.092 0.98xlO"4
ISxlO*4
1.14xlO"4
0.05 0.05 0.95xlO"4
0.8xlO~4
2.5xlO~4
1.64X10"4
(1.0) - 1.87xlO~4
(1.0) 1.91xlO~4
10.4xlO"4
1.8xlO~4
2.2xlO~4
0.84xlO~4
12.3xlO"4
l.lOxlO"4
1.10xlO~4
0.8xlO~4
1.8xlO~4
1.49xlO~4
1.58xlO"4
1.59xlO~4
                                 62

-------
C.  Gas -Mixture Composition as Determined Prom Partial Pressure
    and Mixing Procedure


Air

     Compressed dry air containing less than 0.01% R^O as ob-
tained from the Matheson Gas  Co.  has  been used  diredtly  to fill the
pressure vessel from atmospheric  conditions to  3550  RPa.  The
composition thus remains constant:

     02 = 21%

     N2 = 79%

Combustion Gas

     For combustion gas the vessel is first purged with  nitrogen.
Using a 77%-CO2f 23% -O, gas mixture the vessel is filled from
101 kPa to 550 kPa, with the gas passing through the humidifier.
The remainder of the gas — N2 — . fills, the pressure vessel through
the humidifier to a pressure of 3550~kPa.  The humidifier saturates
the gas at 388 K resulting in a 5% moisture content.  The final
composition of the gas can then be determined based on each
components partial pressure:  %(i) =  100    Pi _
                                ft i                        ftTT
where Pi = Pressure increment x    r  in gas added x  (1  -
                                          P total
                                                          ft
                                                        -  — =-=-
                             TABLE  C-l

               COMBUSTION GAS MIXTURE COMPOSITION
    CO    =    (0.55-0.10) x 106   (0.77)  (1-0.0  5)

      2                   3.55 x  106 x 100
    0        (0.55-0.10) x 106  (0.23)(1-0.05)
     2                  3.55 x 10* x  100~  2'8


         -   (3.55-0.55 + 0.10) x 1Q6   (1.0)(1-0.05)
     2   ~               3.55 x 10° x 100~ 83-°
    H2°  *                                                5.0
                                 63

-------
Substitute Fuel Gas

     The fuel gas mixture is made by  first  purging with  a
30% CO, - 70% N2 mixture.  All components are  added  through  the
humidifier - set to give a 5% moisture  content.  A three-step
procedure is used for pressurizing:   (1) Using 30% CO.,  70%  N2
mixture, the vessel is pressurized to 1140  kPa;  (2)  using helium
the vessel is filled from 1140 kPa to 1830  kPa;  (3)  Using the
30% CO,, 70% N  the gas is filled to  a  final pressure  of 3550 kPa.
Using the same^calculation procedure  as for combustion gas,  a
final composition of the substitute fuel gas can be  determined:


                             TABLE C-2

              SUBSTITUTE FUEL GAS MIXTURE COMPOSITION
                                                             5
     _  (1.83-1.14) x 106   (1.0)  (1-0.05)
     - - -
     -                      Z -          _          i o  c
                  3.55 x 10° x  100                 ~          •LB>:>


C02  =   (3.55-1.83 x 1.14)  x 106   (0.03)  (1-0.05)  =          23  Q

                  3.55 x 106 x  100


          (3.55-1.83 x 1.14) x 106  (0.70)  (1-0.05)            „  ,
JJ    =   - % - —          DO.D
 ^                3.55 x 10  x  100
                                 64

-------
                           TABLE D
C1-C18
C19
C20
C21fC22'C24
C29
C23fC26,C27
CR1 - CR4
  COMPONENTS OF POWER-SUPPLY CIRCUIT
            capacitors
            capacitors
            capacitors
/C25,C28,
MrM3
Prpi2
"6
«7
"11, "13
R12
R14
R15
SP1-SP2

SG,
            capacitors
            capacitors
            silicon-rectifiers
            oscilliscope jack plug
            inductor coil - oil
             immersed
            inductor coil - air
            G.E.  metal  oxide
            variator
            Variacs in  rack
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            resistor
            High Voltage Surge
            Protector - Carbon
            needle point spark
            gap
            high voltage trans-
            former
   0.05
   100
     2

    10
   0.1
yF
   1000 mA,  260KV peak
                                               0.56 mH
                                               4.7 mH
   14 V
   280V,  ISA,  each
   20 k ohm
  120 k ohm
  207 M ohm
    2 k ohm
    5 k ohm
   50 k ohm
    5 k ohm
   50 ohm
  200 ohm
  2.5 M ohm
   10 k ohm
51.75 k ohm
10.35 k ohm
 1.00 ohm

 400 V

  50 V

 480 V to 106 V peak
                                    65

-------
                            TABLE  E
Drawing No,

104855-L

104856-L

104857-B

104858-B

104859-B

104860-B


104861-B


104862-B


104865-B


104866-C


104868-B

104869-B
      RESEARCH-COTTRELL, INC.

       LIST OF DRAWINGS FOR

  HIGH PRESSURE/HIGH TEMPERATURE

    PRECIPITATOR CES-273-P-001


Rev.                     Title

       Detailed Fabrication Specifications

       List of Drawings (this list)

       High Temp. High Pressure Precipitator Assembly

       Top Insulator Compartment Assembly & Details

       Top Insulator Compartment Details

       High Temp. High Pressure Precipitator
       Compartment Assembly and Details

       High Temp. High Pressure Precipitator
       Compartment details

       Bottom Insulator Compartment Assembly and
       Details

 2     High Temp. High Pressure Precipitator
       Pressure Vessels

       High Temp. High Pressure Precipitator
       Heater Bushings

       Heater Humidifier Assembly

       Heater Humidifier
                                  66

-------
                RESEARCH-COTTRELL, INC.

                  SPECIFICATIONS FOR

           HIGH PRESSURE/HIGH TEMPERATURE

            PRECIPITATOR CES 273-P-001
SCOPE - This specification establishes the minimum require-
        ments for the mechanical design, materials, fabrica-
        tion, testing, inspection, and delivery of a high
        pressure/high temperature precipitator and its attach-
        ments and appurtenances.

RESPONSIBILITY - In applying this specification, and the de-
                 tails pertinent to the high pressure/high
                 temperature precipitator, the seller accepts
                 the full responsibility for the design, fab-
                 rication, and conformance to this document,
                 referenced specifications, regulations, and
                 purchase order.

REFERENCES - The latest edition or revision of the following
             references become part of this specification.

             ASME Boiler and Pressure Vessel Code

                  Section I - Power Boilers
                  Section II - Material Specifications
                  Section VIII - Dnfired Pressure Vessels
                  Section IX - Welding Qualifications
                  ANSI B16.5 - Steel Pipe Flanges and Flanged
                               Fittings

CODE COMPLIANCE - Design, fabricate, and inspect the high
                  pressure/high temperature precipitator in
                  accordance with applicable sections of the
                  ASME Code.  Code stamp the precipitator and
                  comply with all the acts and regulations of
                  the jurisdiction in which the equipment is
                  to be operated  (Bridgewater Township,
                  Somerset County, New Jersey).

DESIGN CRITERIA - Pressure shell code stamped in accordance
                  with current ASME code section VIII for the
                  following conditions:

                    Designed Operating Pressure     500 PSIG
                    Designed Operating Temperature  375°F
                    Hydrostatic Test                1000 PSIG
                    Joint Efficiency                100%
                                 67

-------
 5.  DESIGN CRITERIA (cont.)

     All gaskets corrugated stainless steel, asbestos filled.

     Flanges (ASTM A181-GRI)

     Concentric reducer, 10"  0 pipe and 10" straight TEE (ASTM
     A515-70 or equal)

     10 ga plate and 3/4" 0 std. pipe (ASTM A106-GRB)

     Flange Bolting: ASTM A193GRB7 studs
                     ASTM A194GR2H nuts

     Flange bolts to straddle centerlines.

     Paint outside surfaces only with high temperature aluminum

 6.  CALCULATIONS - Prepare and submit all design and analysis
                    calculations for review before fabrication
                    starts.  Include all references and assumptions.
                    Calculations shall demonstrate equipment
                    applicability for all design and operating
                    conditions as delineated in the process diagram.

 7.  ATTACHMENTS - All attachments, internal and external, shall be
                   designed and installed to leave no trapped void
                   space behind the attachment.

 8.  CORROSION ALLOWANCE - Provide a 1/8 inch corrosion allowance
                           for the shell, nozzles, and covers.

 9.  MATERIAL - All pressure sustaining parts shall be fabricated
                of carbon or low alloy steel.  Internal parts shall
                be fabricated of stainless steel type 316 or
                INCONEL 600.

10.  WELDING - Shall be in accordance with the American Welding
               Society latest revisions.  Welding Stainless Steel
               to mild steel use A.W.S. A5.ll class E Ni Cr Fe-2
               electrode.  Welding Stainless Steel to Stainless
               Steel use A.W.S. ASTM E316L-16 electrode.  Welding
               inconel 600 to inconel 600 use inconel coated rod
               #182 electrode.
                                       68

-------
11.  QUANTITIES REQUIRED - The quantities required are as follows;

       Qty.

        1     Top Insulator Compartment            104865-B1
        1     Middle Precipitator Compartment      104865-B2
        1     Bottom Insulator Compartment         104865-B3
        1     Humidifer Heater Compartment         104865-B4
                                    69

-------
                                TECHNICAL REPORT DATA
                          (Please read Inunctions on the reverse before completing!
  REPORT NO.
 EPA-600/7-77-132
2.
                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Development of a High-Temperature/High-Pressure
   Electrostatic Precipitator
                            5. REPORT DATE
                             November 1977
                            6. PERFORMING ORGANIZATION CODE
 . AUTHORIS)
J.R.  Bush, P. L. Feldman, and M. Robinson
                                                       3. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Cottrell Environmental Systems
Research-Cottrell, Inc.
Bound Brook, New Jersey  08805
                            10. PROGRAM ELEMENT NO.
                            EHE623A
                            11. CONTRACT/GRANT NO.
                             68-02-2104
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; 8/75-8/77
                            14. SPONSORING AGENCY CODE
                             EPA/600/13
is. SUPPLEMENTARY NOTES jjERL-RTP project officer for this report is  Leslie E. Sparks,
Man Drop 61, 919/541-2925.
i6. ABSTRACT
               report gives results of B. laboratory test demonstrating the feasibility
of electrostatic precipitation at high temperatures (to 1366 K) and pressures (to 3550
kPa): corona currents were stable at all temperatures.  Detailed current/voltage
characteristics under negative and positive polarity were obtained in dry air , in
a simulated combus'tion gas ,  and in a substitute fuel gas  for each temperature and
pressure level. Major effects of temperature, pressure, and electrode geometry on
precipitator design characteristics (sparking voltages , corona starting voltages ,
and current levels) were evaluated.  This work has established that no practical
temperature or pressure limitation exists to the levels tested. A preliminary cost
estimate is presented for a commercial-size precipitator design applied to a com-
mercial fluidized-bed combustor and turbine system.
17.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                           b.lOENTIFIERS/OPEN ENDED TERMS
                                         c.  COSATI Field/Croup
Air Pollution
Electrostatic Precipitators
High Temperature Tests
High Pressure Tests
                Air Pollution Control
                Stationary Sources
13B

14B
18. DISTRIBUTION STATEMENT

 Unlimited
                19. SECURITY CLASS (THiJ Report)
                Unclassified
                                                                    21. NO. OF PAGES
                20. SECURITY CLASS (Tills page)
                Unclassified
                                         22. PRICE
EPA Form 2220-1 (9-73)
                                     70

-------