Electrostatic Percipitation of Fly Ash
Electrostatic Percipitation of Fly Ash
Electrostatic Percipitation of Fly Ash
To cite this article: Harry J. White (1977) Electrostatic Precipitation Of Fly Ash, Journal of the Air Pollution Control
Association, 27:1, 15-22, DOI: 10.1080/00022470.1977.10470386
To link to this article: http://dx.doi.org/10.1080/00022470.1977.10470386
Harry J. White
Rapid growth of the power industry and the need for ever
increasing power generation efficiency led in about 1920 to the
successful development of pulverized coal as a basic fuel for
large steam generating units in the United States. The pulverized coal process for power generation was pioneered in
Milwaukee, Wisconsin, and soon gained recognition as a major
technical advance over the old stoker-fired power boilers.1 The
new firing method brought in its wake, however, severe and
complex gas cleaning problems. In this form of firing, the residual ash from the coal is carried in the furnace gas stream
in the form of suspended finely divided particles which are
emitted from the stack unless effective collection means are
provided. These minute fly ash particles may comprise 10%
or more of the mass of the coal burned, and therefore would
constitute an intolerable nuisance unless removed from the
gas stream before emission from the stack.
In the power generation industry, application of electrostatic precipitation to the fly ash removal problem is almost
as old as the use of pulverized coal, with the first precipitator
going into service in 1923 at the Trenton Channel plant of
Detroit Edison.2'3 Various inertial and mechanical collectors,
such as wet scrubbers and cyclone collectors, had been tried
but proved ineffectual in solving the severe air pollution
problem inherent in the pulverized coal firing process. The
first precipitator installation at the Trenton Channel plant
consisted of three units specified for 90% collection efficiency
and treating a total gas flow of 800,000 acfm. A period of several years was required to overcome the many problems encountered before achieving successful operation, but with this
practical achievement wide acceptance of the precipitator for
high efficiency removal of fly ash was assured in the power
industry. Today a total of more than 1300 fly ash precipitator
installations having a rated gas flow of over 500 million acfm
have been made in the United States.
The extent of the fly ash problem is shown strikingly by the
amount of coal being burned by the utilities, which in 1975
totaled over 400 million tons. The corresponding quantity of
fly ash which could be emitted would be about 30 million tons
annually, assuming an average ash content of 10% and an
average retention of ash in the furnace of 30%. The magnitude
of the problem of fly ash emission control for a large modern
power plant may be illustrated by considering a coal-fired
generating unit producing, for example, 600 Mw electric
output. Such a unit typically will exhaust flue gas at the rate
Dr. White, recipient of the Frank A. Chambers Award for "Outstanding achievement in the science
and art of pollution control," is now a consultant in Carmel, CA. He is the author of several definitive
books on air pollution control.
This paper will be published in four parts in JAPCA in the coming months. Sections 1 and 2, as
identified above, are in this issue. Section 3 will appear in February, Sections 4 and 5 in March, and
Sections 6, 7, and 8'will conclude the presentation in April.
January 1977
15
2000
1750 -
mandated by clean air legislation. New levels of scientific understanding and technical sophistication in the precipitation field will be
necessary to successfully meet the performance and reliability
=1500 -
2 1250 -
1000 -
H 750 500 -
250 -
1920
1930
1940
1950
Year
1960
1970
1980
1920
1930
1940
1950
Year
1960
1970
1980
99.9
99.8
i
_ 99.7
99.5
S-99.3 -
Maximurn,/
8 98. "97 95
93
90
1920
100 -
1920
1980
coal for the same period. The amount produced from coal has
consistently comprised about one-half of the total power
generated. Coal consumption per annum by the public utilities
is shown in Figure 2. Reasonable projections of these historic
growth patterns support the expectation for continued high
coal consumption for power production in the United
States.
Data on the installed fly ash precipitation capacity in this
country and the trend in collection efficiencies, beginning with
the first application in the early 1920's, are plotted in Figures
3 and 4. Particularly notable are the very rapid increase in cfm
capacity and the trend toward ultra-high efficiencies, which
have occurred in the past few years. Collection efficiencies for
the earliest precipitators were 90%, reaching about 99% average value in 1970, and about 99.8% maximum value currently. The rapid increase in collection efficiency in recent
years is directly attributable to the air pollution control requirements stemming from the federal clean air legislation
of 1967 and 1970.4
The overall growth of fly ash precipitation in the U. S. is
even more remarkable when both the increases in the cfm
capacity and in collection efficiency are taken into account.
The higher efficiency requirements are especially important
since the increase in precipitator size is nonlinear with efficiency. For example, a fly ash precipitator for 99% efficiency
typically will be three or four times larger than one of similar
capacity for 90% efficiency. Taking both size and efficiency
factors into account, the per annum growth rate for new fly
ash precipitators turns out to be some 20 times greater for the
1965-70 era than for the 1945-50 era. This rapid growth is
evident in Figure 5 which shows the annual sales of precipitators in the United States for the period 1966-1975, of which
about three-fourths is accounted for by fly ash precipitators.
Precipitator development has been influenced in several
important ways beyond increased use and higher collection
efficiencies by the federal clean air legislation and emission
standards.5* First, the requirements for meeting emission
standards are backed up by enforcement provisions which can
curtail plant output or even force complete shutdown of
generating units if such is necessary for compliance with
emission standards. The practical result of these enforcement
January 1977
/j
^r
/ Average
Figure 4. Trends in the average and maximum collection efficiencies of fly ash precipitators in the United States.
provisions has been to require much more conservatively designed precipitators and to compel higher design and construction standards than were previously acceptable in the
precipitator industry.
Second, the emission limits on sulfur oxides5 have caused
the power industry to turn toward much greater use of low
sulfur coals as a means of meeting these stringent limitations.
This in turn complicates the precipitator problem because of
the greater probability of high resistivity ash with these coals.
Although effective methods are known for dealing with high
resistivity, these methods require a higher level of technical
sophistication and result in increased costs as compared with
fly ash collection from higher sulfur coals.
A third complication can arise from the need to provide
effective removal of both the fly ash and the SO2 from the flue
gas. This introduces several new factors in the overall air
pollution control system for a power plant, and also several
different control strategies may need to be considered. The
technical and economic feasibility of the various possible
strategies have yet to be firmly established. However, it now
appears that dry collection of the fly ash in a precipitator, either before or after the air preheaters, followed by removal of
the SO2 in a wet scrubber may turn out to be the preferred
method.
500
< 100 -
1966
* Particulate emissions for new generating units are limited under the standards to a maximum of 0.1 lb/million Btu heat input, equivalent to about 0.08 gr/scf for a typical fly ash,
and to visible emissions not exceeding Ringelmann 1. Collection efficiencies of 99+% are
usually necessary to meet these standards, with some allowance for safety margins.
/K
1968
1970
1972
Year
1974
1976
17
Purpose of this section is to set forth the basic principles and theory
underlying the precipitation process, which are necessary for understanding the design and operation of fly ash precipitators. Included
are the migration velocities of particles, the idealized efficiency
equation, and the effects on performance of nonuniform particle size
and other nonuniformities. Semiempirical and statistical methods
sometimes used for precipitator design are also examined.
Basically, an electrical precipitator must provide three essential functions: (1) the suspended particles must be given
an electrical charge; (2) they must be subjected to an electric
field to remove them from the gas stream to a suitable collecting electrode; and (3) means must be provided for removing the particle layers from the electrode surfaces to an
outside receptacle with as little loss as possible.
In practice, electric charging of the particles is accomplished
by means of ions produced in the high voltage d-c corona.
Under typical precipitator operating conditions the charge
attained on a 1 micron particle, for example, is about 250
elementary charges.
In a Cottrell or single-stage precipitator the collecting field
is also provided by the high voltage d-c corona. Typical
values of the collecting field are 4-5 kv/cm, and the electric
force acting o n a l / i particle, for example, is about 3000
times gravity. It should be noted that the existence of these
very high collection forces, especially for the finer particles
which are the most difficult to collect, is one of the major
factors which set electrostatic precipitators apart from other
methods of particle collection. Furthermore, the forces are
exerted directly on the particles themselves, whereas in
mechanical collecting methods it is necessary to treat the
whole mass of gas in order to exert the necessary separating
forces on the particles. As a result, electrostatic precipitators
are inherently economical in power consumption and are
free of the high draft losses which characterize most mechanical or inertial collection methods.
In a Cottrell precipitator the actual power required to clean
one million cfm of flue gas at 99%, for example, turns out
to be only about 200 kw, as compared to over 2000 kw consumed by a typical scrubber for the same application.
Removal of the collected particle layers is accomplished by
rapping (by impact, or by vibration, of the electrodes). The
rapping process is critical to overall performance, as a balance must be struck between keeping the electrodes sufficiently clean, and overrapping which causes excessive loss
of already collected particles by reentrainment into the gas
stream.
e-(A/V)w
(1)
This may be calculated from theory by equating the electrostatic force on the particle to the drag force, giving the
relation *
6irda
(2)
Efficiency Equations
* The quantity eo appears in the expression for q and in Eq.(3) because of the use of S.I. units,
and has the value 8.854 X 10~12 farad/meter.
t The units in Eq.(2) may be either cgs electrostatic (esu) or Systeme Internationale (SI).
The latter system is used herein in conformity with the present vogue. Thus w is expressed
in meters/sec.
* The exponent (A/V)w in Eq. (1) is dimensionless so that A, V, and w must he expressed
in consistent units. In practice, usage varies between English and metric systems.
January 1977
determined by the average voltage. Therefore, it is advantageous to keep both the maximum and average values of the
voltage as high as possible.
Variations in Field over the Collection Surface.
For pipe precipitators, the field distribution over the collection surface is uniform, but for duct precipitators it varies
widely because of the effects of the plate baffles and the grid
arrangement of the corona wires. These variations in the
collecting field obviously affect the theoretical efficiency
equation. Since there is no feasible approach for calculating
the effects of the baffles and end effects of the plates, it is usual
to neglect these factors and simply use an average field based
on average current density at the plates. This is an approximation which can be justified only qualitatively in a theoretical sense, but is adequate in a practical sense because of the
presence of other uncertainties of equal or greater magnitude.
Nonuniform Particle Size
= l-
(4)
Jo
A epEcEp
v~T-
(5)
T? = 1 C
(6)
Jo
which can be evaluated analytically in certain cases of practical interest. The most important case is that for the lognormal particle size distribution which is treated in the next
section.
Log-normal Distribution.
(7)
19
. e-(A/V)w(x)
,
*
dx
(10)
'o
(v 2TT In ag) x
Similarly, substitution of Eq. (9) in Eq. (6) gives the efficiency
equation for the log-normal case where w is proportional to
the particle diameter
P-t
/2.p-kx
= 1 - 11
Joo ( v
.
2TT
dx
(11)
In Gg) x
In practice, both the particle size and the gas flow are
nonuniform, and need to be taken into account in formulating
efficiency equations. It is not difficult to develop the theo2
77 = 1 - - i = f " [e-U /2) . e-kxg . et In cA d t ( 1 2 )
retical equation for this case, but the result involves double
integration and requires a great amount of numerical calcuThe lower limit of integration becomes t = <*> which corre- lation to evaluate practical cases. Such calculations can be
performed by computer methods, but otherwise are too time
sponds to x = 0 in Eq. (11). Evaluation of the integral in exconsuming to be justified. An approximate procedure which
pression (12) by analytic means has been studied by Allander
is usually permissible is to calculate the effects of nonuniform
and Matts.15 They were able to relate the integral to the soparticle size and nonuniform gas flow separately. Such a
called aftereffect function which appears in the classical
procedure may be justified in many cases because of other
Jahnke and Emde Tables of Functions, pp. 38-39. Application
approximations and uncertainties which enter into the design
of this method is made in a later section of this paper.
of fly ash precipitators.
It is sometimes convenient to use t instead of x as the
variable of integration in expression (11),
Unbalanced gas flow through a precipitator lowers performance in two fundamental ways. First, uneven treatment of
the gas lowers collection efficiency in the high gas velocity
zones to a degree not compensated for in the low velocity
zones. Second, particles already captured may be blown off
the plate surfaces in high gas velocity regions and be lost from
the precipitator. Although each of these adverse effects is
important, the second or reentrainment loss predominates
where gas flow is markedly bad. Positive measures are needed
to minimize these losses, including gas flow model studies to
insure high quality gas flow, and limiting design gas velocities
to 5 or 6 ft/sec.
The effect of nonuniform gas flow in lowering efficiency can
be treated in a manner similar to that for nonuniform particle
size distribution. Let f{v) du be the fraction of the precipitator
inlet cross-section area which has a gas velocity between v and
v + dv. Then
f(v) dv =
where Ac is the total cross-sectional area for gas flow through
the precipitator. It follows that the fractional loss dQ through
the area dAc is given by
dQ = e-tdA'dV)w f(v)
fo
(13)
where dA is the element of collection surface associated with
dAc, and dV is the gas flux through dAc. But dA/dV can be
expressed by
dA = A(dAc/Ac) = A
dV
vdAr
Arv
20
e-(wkA/V)
(15)
performance data are treated as statistical quantities. Performance data for a group of precipitators are subjected to
regression analysis for the purpose of isolating and evaluating
the significant parameters such as gas velocity, gas temperature, and sulfur content of the coal burned. An example of one
such statistical technique is given later in this paper.17 A
similar statistical approach has been applied recently to the
design problem for so-called hot precipitators, that is, those
operating at high temperatures ahead of the air preheaters.18
The utility of regression analysis methods rests mainly on the
improvement gained over simple averaging techniques. The
disadvantages are the absence of a fundamental theoretical
foundation, and the possibility of perpetuating design weaknesses and other deficiencies through gradual deterioration
of the statistical design base.
9.
10.
11.
12.
13.
References
1. G. A. Gaffert, Steam Power Stations. 2nd ed. McGraw-Hill, New
York, 1940. p. 305.
2. H. M. Pier and A. N. Crowder, "Catching pulverized-coal ash at
the Trenton Channel plant," Power 65: 834 (1927).
3. D. Carlton-Jones, "Electrostatic precipitation offlyash reaches
fiftieth anniversary," J. Air Poll. Control Assoc. 24: 1035
(1974).
4. "The Air Quality Act of 1967" and The Clean Air Act of 1970."
5. U. S. Environmental Protection Agency, "Standards of Performance for New Stationary Sources," Federal Register 36: (159)
Part II, (Dec. 23,1971).
6. A more detailed discussion of the capture process and the derivation of the probability efficiency equation is given in the book:
January 1977
7.
8.
14.
15.
16.
17.
18.
21