Encyclopedia of Environmental Science and Engineering, Volume I and II

828 PARTICULATE EMISSIONS

The ASME power test code,^27 in contrast, is designed to

measure performance of devices such as precipitators and

cyclones, and thus is concerned only with substances which

are particulate at conditions prevailing in the equipment. This

test usually used a filter assembly with the filter very close to

the sampling probe so that the filter may be inserted into the

stack avoiding condensation. No impingers are used.

To some extent filter characteristics are determined by

process conditions. Alundum thimbles and glass wool packed

tubes are used for high temperatures. If liquid droplets are

present at the filter inlet, glass-wool tubes are the only useful

collection devices, because conventional filters will readily

become plugged by droplets. Glass-wool collection greatly

complicates quantitative recovery of particles for chemical

or size analysis.

In sampling a large duct having several traverse points

for flow and particle measurement, particles for all points

on a traverse are usually collected in a single filter impinger

train, thus giving an average dust concentration. Each sample

point is sampled for an equal time but at its own isokinetic

velocity. The probe is then immediately moved to the next

point and the flow rate adjusted accordingly. Sample flow

rate is adjusted by rotameter or orifice readings, but total gas

flow during the entire test is taken from a dry meter.

Minimum sampling time or volume is often set by regu-

lation. Examples are:

Bay area^24 —Sample gas volume = 20 L 0.8 , where volume

is in standard cubic ft, and L is duct equivalent dia. in ft.

A maximum sampling rate of 3 SCFM is specified and a

minimum time of 30 min.

ASME^27 —Minimum of 2hr with at least 10 min at each

traverse point through two complete circuits.

APCD^25 —5–10 min/point for a total run of at least 1 hr.

Industrial gas cleaning institute —At least 2 hr or 150 ft^3

sample gas or until sample weight is greater than 30% of filter

weight.

Emissions are calculated from test volumes of weight of

particulates collected and volume of sample gas through the

gas meter. Care must be taken to include particles deposited on

tubing walls as well as those trapped by the filter. If conden-

sibles are to be included, the liquid from the impinger train is

evaporated to dryness, and the residue is weighed and included

with the particulates. Corrections to the gas volume depend on

sample train operation and on standard conditions for report-

ing emissions, and these are spelled out in detail in the specific

test codes to be used. Results are usually expressed both as

grains per cubic foot (using standard conditions defined in the

code) and as lbs/hr from the whole stack.

Measurement and Representation of Particle Size

A determination of the emitted particle size and size dis-

tribution is a desirable element in most control programs.

Collection efficiency of any given piece of equipment is a

function of particle size, being low for small particles and

high for large ones, and capital and operating costs of equip-

ment required increased steadily as the dust particle size

decreases.

Perhaps the simplest method of particle size measure-

ment, conceptually at least, is by microscope count. The

minimum size that can be counted optically is about 0.5 μ

which is near the wavelength of visible light. Electron micro-

scopes may be used for sizing of smaller particles. Counting

is a laborious procedure, and sample counts are often small

enough to cause statistical errors at the very small and very

large ends of the distribution. This method requires the small-

est sample size and is capable of giving satisfactory results.

Care must be taken in converting from the number distribu-

tion obtained by this method to mass distribution.

A second simple method is sieve analysis. This is com-

monly used for dry freely flowing materials in the size range

above 44 μ, a screen size designated at 325 mesh. Using spe-

cial shaking equipment and very delicate micromesh sieves

particles down to 10 μ can be measured. Error can be caused

by “blinding” of the sieve mesh and sticky or fine particles,

incomplete sieving, and particle fragmentation during siev-

ing. A sample size of at least 5–10 g is usually required.

Another class of measurement techniques is based on the

terminal falling velocity of particles in a gas (air). The quan-

tity measured is proportional to rd^2 , where r is particle density

and d is diameter. Hence a separate determination of density

is needed. One such device is the Sharples Micromerograph

(Sharples-Stokes Division, Penwalt Corporation, Warminster,

Pennsylvania). The device records the time for particles to fall

through a 2 m high column of air onto the pan of a continu-

ously recording balance. Templates are available to convert

fall time to rd^2. The Micromerograph is mechanically and

electrically complex but easy to use. An objection is that a

significant fraction of the injected particles stick to the column

walls and do not reach the balance pan. This effect can some-

times be selective, and it thus gives a biased size distribution.

A second sedimentation device is the Roller elutria-

tor tube, Figure 10. A powder sample is placed in the tube

and air is passed upwards through it for a specified time.

A separation is effected with small particles being carried

overhead and large ones remaining in the tube. Often a series

of tubes of decreasing diameter are connected in a cascade

with each successive tube having a lower air velocity and

retaining finer particles. The Roller method was used quite

widely in the petroleum industry for many years. However, it

is slow, requires a large sample, does not give clean particle

size cuts, and is sensitive to tube orientation. It is therefore

being supplanted by newer methods.

A third sedimentation is centrifugal sedimentation. This

is the standard test method of the Industrial Gas Cleaning

Institute, and use of such devices of the Bahco type has been

standardized by the ASME.^32 The Bahco analyzer consists

of a rapidly spinning rotor and a superimposed radial gas

flow from circumference to center. Larger particles are cen-

trifuged to the outside diameter of the rotor, while small ones

are carried to the center with a cut point determined by gas

velocity and rotor speed.

Still another method is the Coulter Counter. In this

technique the test powder is dispersed in an electrolyte,

which is then pumped through a small orifice. Current flow

between electrodes on each side of the orifice is continuously

C016_001_r03.indd 828C016_001_r03.indd 828 11/18/2005 1:15:35 PM11/18/2005 1:15:35 PM