Operations Manual
Pilat (University of Washington)
Mark 3 and Mark 5
Source Test Cascade Impactor

by

Dr. Michael J. Pilat
email: mpilat@u.washington.edu
Department of Civil Engineering
University of Washington
Seattle, WA 98195 USA

January 1998

Table of Contents

Page

List of Figures

Fig. No. Page
1. Cross Section of Pilat (UW) Mark 3 Cascade Impactor 5
2 Cross Section of Pilat (UW) Mark 5 Cascade Impactor 6
3. Expanded View of Parts of Pilat (UW) Mark 3 Cascade Impactor 7
4. Calculated Aerodynamic Cut Diameters (d50s) for Pilat (UW) Mark 3
Source Test Cascade Impactor (Temp. of 100, 300, & 500oF) 9
5. Calculated Aerodynamic Cut Diameters (d50s) for Pilat (UW) Mark 3
Source Test Cascade Impactor (Temp. of 500, 1000, & 1500oF) 10
6 Mark 5 Jet Stage Arrangement for 0.1 to 0.3 acfm Flow 13
7 Mark 5 Jet Stage Arrangement for 0.3 to 0.5 acfm Flow 14
8 Mark 5 Jet Stage Arrangement for 0.5 to 0.8 acfm Flow 15
9 Mark 5 Jet Stage Arrangement for 0.8 to 1.0 acfm Flow 16
10 Calculated d50s for Pilat Mark 5 Cascade Impactor (0.1 to 0.3 acfm) 18
11 Calculated d50s for Pilat Mark 5 Cascade Impactor (0.3 to 0.5 acfm) 19
12 Calculated d50s for Pilat Mark 5 Cascade Impactor (0.5 to 0.8 acfm) 20
13 Air Pressure Drop versus Air Flow Rate for Mark 5 Impactor 21
14. UW Mark 3 Cascade Impactor Sampling Train 23
15. Data Sheet for Pilat (UW) Cascade Impactor Test 32
16. Example Size Distribution Graph 33

List of Tables

Table No.
1 Jet Dimensions of Pilat Mark 3 Source Test Cascade Impactor 4
2 Jet Dimensions of Pilat Mark 5 Source Test Cascade Impactor 5
3 Size Data for Example Source Test 30

I. Instructions for Pilat (UW) Mark 3 and Mark 5 Cascade Impactors

A. Function of Pilat (UW) Source Test Cascade Impactor
The eleven jet stage Pilat (University of Washington) Mark 5 Source Test Cascade Impactor is essentially an elongated version of the seven jet stage Mark 3 Pilat (University of Washington) Source Test Cascade Impactor. The seven jet stage Pilat (Univ. of Washington) Mark 3 Source Test Cascade Impactor was developed at the University of Washington Department of Civil Engineering for measuring the size distributions (0.3 to 20 micron diameter size range) of particles in stacks and ducts at air pollutant emission sources. This instack cascade impactor was developed because at that time (1968), there were no available instack cascade impactors. This instack cascade impactor was needed by Professor Pilat and his graduate students for their research projects regarding the design and evaluation of particulate air pollutant control equipment and the characterization of the size distribution and chemical composition of particles emitted from industrial plants. The University of Washington (Pilat) cascade impactor is inserted inside the duct or stack to minimize sampling tubing wall losses and water condensation problems. The 11 stage Mark 5 model was developed a few years later to provide more jet stages and was initially used for measurements upstream of particle control equipment (at the higher particle concentrations). The parts of the Mark 3 and Mark 5 models are interchangeable. Size distributions of aerosol particles at air pollutant emission sources are needed to:

1. Design new air pollutant removal equipment;
2. Measure the particle collection efficiency as a function of particle size of existing particulate control systems; and,
3. Characterize the aerosol emissions from the various sources.

B. Description of Pilat (UW) Cascade Impactor Components

The Pilat Mark 3 and Mark 5 cascade impactors consist of a threaded cylindrical casing, impaction jet stages, particle collection plates, threaded inlet and outlet sections, and a filter holder in the outlet section (for a 47 mm diameter final filter). Fig. 1 shows a cross-section of the Mark 3 and Fig. 2 shows the Mark 5, and Fig. 3 illustrates the individual parts. The number of jets in each jet stage and the number of jets per stage are shown for the 7 stage Mark 3 model in Table 1 and the 11 stage Mark 5 in Table 2. Note that the Mark 3 and Mark 5 jet stages 1, 2, 3, 4, and 5 are identical in both impactors. The Mark 3 jet stage number 6 and the Mark 5 jet stage number 7 have the same jet diameters and number of jets.

Figure 1. Cross Section of Pilat (UW) Mark 3 Cascade Impactor	Figure 2. Cross Section of Pilat (UW) Mark 5 Cascade Impactor

Figure 3a. Expanded View of Parts to the Pilat (UW) Mark 3 Cascade Impactor

Figure 3b. Expanded View of Parts to the Pilat (UW) Mark 3 Cascade Impactor

Table 1
Jet Dimensions of Pilat (UW) Mark 3 Source Test Cascade Impactor

The number of jets and jet dimensions for the Mark 5 model are shown below in Table 2. Note that only 11 or less jet stages can be used in the Mark 5 during sampling. However, there are the extra jet stages furnished to enable the Mark 5 model to be arranged in at least three different sampling configurations (ranging from the higher gas volumetric flow rate to lower gas volumetric flow rates).

Table 2
Jet Dimensions of Pilat (UW) Mark 5 Source Test Cascade Impactor

The collection plates for Mark 3 stages 2 through 7 (Mk 5 stages 2 - 11) are donut shaped, allowing for particle impaction upon the annular shaped collection plate and for the sampled gas or air to flow through the center or the "donut hole" and on to the next jet stage. The final particle collection "stage" is a filter holder designed for glass fiber filters (although other filter materials such as Teflon can be used, but care needs to be taken because some other filter materials such as Teflon have much a greater gas pressure drop compared to glass fiber filters). This filter holder is incorporated into the outlet section of the cascade impactor using a collar to hold the filter in place. The filter collar has a hole in it so that the dowel pin in the outlet section will fit into the filter collar hole, preventing it from rotating when the outlet section is tightened (screwed onto) the cylindrical casing. The rotation of the filter collar could cut the filter. The use of a Teflon gasket between the filter collar and the filter and between the filter and the mesh screen is helpful to protect the filter from being cut or damaged from the pressure (also the Teflon gasket helps to prevent the filter fibers from sticking to the mesh screen, especially if some acid condensation occurs). The filter paper is supported by a fine mesh screen and below the fine mesh screen is a perforated support plate. A cross sectional view with an air flow diagram of the Mark 3 Cascade Impactor is illustrated in Fig. 1 and Fig. 2 shows the Mark 5. Fig. 3 shows the parts of a disassembled impactor.

There are a number of different inlet diameters of the sampling nozzles for the impaction inlet section. The nozzle outlet bores are all 0.500 inches diameter which acts as the #1 jet stage (the nozzle opening tapers out to a jet of 0.500 inches diameter). One 3/32 inch by 2-1/8 inch inside diameter O-ring is used on each jet stage and on the filter collar (for a total of seven on the Mark 3 and 11 on the Mark 5). One 1/16 inch by 2-5/8 inch inside diameter O-ring is used on the outlet section. Viton O-rings are recommended for temperatures of 250 - 500F; for temperatures exceeding 500F, eliminate the O-rings and use an extra collection plate as a spacer after jet stage seven. These O-rings are not absolutely needed for operation of the Mark 3 and Mark 5 cascade impactors because of the smooth finish on the 316 stainless steel parts. However, it is a good idea to use them as it helps the jet stages and the particle collection plates to seat or fit nicely.

C. Aerodynamic Cut Diameters (da50s) for Pilat (UW) Mark 3 Cascade Impactor

The cut diameter is the particle diameter collected with 50% collection efficiency by that jet stage. The aerodynamic diameter is the diameter of a spherical particle of unit density (i.e. density of 1.00 grams/cubic centimeter) which behaves aerodynamically just the same as the real particle. It has been conventional for many years to use the cascade impactor jet stage cut diameters and aerodynamic diameters with cascade impactor particle size measurements. More recently, the EPA PM10 criteria air pollutant for particulate matter under 10 microns diameter refers to the aerodynamic cut diameter of 10 microns. And there is the EPA (proposed Nov. 1996; promulgated July 1997) PM2.5 with the aerodynamic cut diameter of 2.5 microns.

The equation for the cascade impactor jet stage aerodynamic cut diameter da50 is given by the equation:

where m is the gas viscosity, Dj is the diameter of the jet holes in that cascade impactor stage, Y50 is the inertial impaction parameter at the 50% particle collection efficiency of that impactor jet stage (Y50 is about 0.145 for cylindrical round jet stages and Y50 does vary a little with the Reynolds number and other variables), C is the Cunningham slip correction factor for the particle of diameter da50, and Vj is the gas velocity at the inlet (upstream inlet to the jet hole) to the gas jet through the impactor jet stage.

The calculated aerodynamic cut diameters or da50s for the Pilat (UW) Mark 3 Cascade Impactor are shown in Fig. 4 for temperatures of 100, 300, and 500oF and in Fig. 5 for temperatures of 500, 1000, and 1500oF. The horizontal axis shows the gas volumetric flow rate through the cascade impactor at stack conditions (actual ft3 gas/minute at the instack cascade impactor conditions). The vertical axis of Fig. 4 and Fig. 5 show the aerodynamic cut diameters for the various jet stages. These Mark 3 Cascade Impactor have been calibrated by experimental measurements performed under EPA funding contracts to Southern Research Institute. Also research funded by EPA concerning the sampling for electrostatically charged particles (such as downstream of elect. precipitators) resulted in experimental calibrations of the Pilat (UW) Mark 3 Cascade Impactor.

Fig. 4 Calculated Aerodynamic Cut Diameters (da50s) for Pilat (UW) Mark 3 Source Test Cascade Impactor (Temp. of 100, 300, & 500oF)

Fig. 5 Calculated Aerodynamic Cut Diameters (da50s) for Pilat (UW) Mark 3 Source Test Cascade Impactor (Temp. of 500, 1000, & 1500oF)

D. Aerodynamic Cut Diameters (da50s) for Pilat (UW) Mark 5 Cascade Impactor
The Mark 5 model is primarily designed for operation with larger particle mass concentrations where it is necessary to sample at lower gas flow rates (0.1 to 0.3 acfm). Thus, the Mark 5 is typically used at the inlet to particulate control devices where the particle mass concentration is higher (see Figures 6, 7, 8, and 9 for arrangement of jet stages at the various gas sampling flow rates). Note that the Mark 5 can be operated at higher gas flow rates by removing the lower jet stages (i.e. Fig. 9 shows the 11 jet stage arrangement for the highest gas sampling flow rates).

The Mark 5 jet stages should not be operated at gas flow rates such that the stage da50 goes below 0.2 microns. Referring to the Figure 12 graph of da versus the gas flow rate, it can be seen that when Stage 11 with 36 jets of 0.0102 inches in diameter goes above approximately 0.65 acfm, the d goes below 0.2 micron diameter. Thus, this jet stage should be removed at flow rates greater than approximately 0.3 acfm and the stage with 12 jet holes of 0.0960 inches in diameter inserted after the inlet jet stage No. 1. (see Figure 7 for the arrangement of jet stages; see Figure 11 for the graph of d50 versus gas flow rate.). With a gas flow rate greater than 0.5 acfm, the stage with 40 jets of 0.0102 inches in diameter should be removed and the jet stage with 6 jets of 0.2280 inches in diameter inserted between the No. 1 inlet jet stage and the stage with 12 jets of 0.0960 inches in diameter. (See Figure 8 for arrangement of jet stages.). At gas flow rates greater than 0.8 acfm, the stage with 56 jets of 0.0102 inches in diameter should be removed and a spacer inserted to take up the space. This spacer should be placed between the last two collection plates (see attached Figure 9 for arrangement of jet stages).

Fig. 6 Mark 5 Jet Stage Arrangement for 0.1 to 0.3 acfm Flow

Fig. 7 Mark 5 Jet Stage Arrangement for 0.3 to 0.5 acfm Flow

Fig. 8 Mark 5 Jet Stage Arrangement for 0.5 to 0.8 acfm Flow

Fig. 9 Mark 5 Jet Stage Arrangement for 0.8 to 1.0 acfm Flow

The aerodynamic cut diameters for the Mark 5 Pilat (UW) Cascade Impactor are presented in Figures 10, 11, and 12. These graphs are for different arrangements of the jet stages, corresponding to the different gas sampling flow rates at the impactor temperature and pressure.

Figure 10
Calculated da50s for Pilat Mark 5 Cascade Impactor (0.1 to 0.3 acfm)

Figure 11
Calculated da50s for Pilat Mark 5 Cascade Impactor (0.3 to 0.5 acfm )

Figure 12
Calculated da50s for Pilat Mark 5 Cascade Impactor (0.5 to 0.8 acfm )

E. Gas pressure drop for Mark 5

Figure 13 presents the gas pressure drop across one configuration of the Mark 5. The right hand curve is for an arrangement with Teflon gaskets between the stage and the collection plate. Note that it is recommended that this configuration with jet stage No. 11 with 36 holes of 0.010 inches not be operated above 0.3 acfm. It is possible to use the Pilat (UW) Mark 5 Cascade Impactor as a "low pressure cascade impactor" but please note that care must be taken because there can be some leakage of gas around the edges of the cascade impactor jet stage (rather than the gas flowing through the jet holes). The majority of the gas pressure drop is across the last jet stage in the Mark 5 (the number 11 jet stage) and therefore, the use of a gasket between this jet stage and the collection plate may be needed. When operated as a "low pressure cascade impactor" it is necessary to measure the absolute gas pressure downstream of the last cascade impactor jet stage and upstream of the final outlet filter. Because the filter holder is included in the Mark 5 cascade impactor outlet section, for low pressure operation it is necessary to omit the filter in the outlet section so that the gas pressure can be measured at the impactor outlet and then a separate filter holder connected downstream of the gas pressure measurement connection. Although we do not recommend using the Pilat (UW) Mark 5 Cascade Impactor in the low pressure cascade impactor mode, it has been done successfully. The jet stage aerodynamic cut diameters (da50s) for the low pressure operation will need to be calculated using equation 1 and the appropriate Cunningham Correction factor used in this calculation.

Fig. 13 Air Pressure Drop versus Air Flow Rate for Mark 5 Impactor

F. Operating Procedure for Pilat (U of W) Mark 3 and Mark 5 Cascade Impactors

1. Pretest Preparations of Cascade Impactor

1. Clean all parts of any dirt and soiling material and inspect the stages for plugged jets (i.e. hold the jet stage up to the light and determine if you can see through the jet stage holes). It is convenient to use an ultrasonic cleaner with some soap or detergent to clean the impactor parts.
2. Check the o-rings and replace the worn ones.
3. Assemble the impactor to check for missing parts.
4. If you are to use a grease on the stainless steel insert substrates, place a thin layer of suitable grease on the seven collection plate inserts (about 0.02 grams of grease per plate or insert). Place the collection plate inserts in a dust-free container such as a petri dish. For higher temperatures use less grease. For extremely high temperatures, bake the plates before weighing to eliminate volatiles in the grease which may escape during the test and affect the weight data. At these higher temperatures ( 500 F) it may be necessary to use a less volatile grease.

As an alternative to grease, 934AH glass fiber or ultra pure quartz microfiber substrates may be mounted in the stainless steel collection plate inserts. Because the greased stainless steel substrates are sort of "messy", it seems more testing is being done using the glass fiber or quartz fiber substrates. However, note that the particle collection characteristics of glass fiber substrates are not the same as for a clean flat greased surface and this can affect the experimentally measured magnitude of Y50 in equation 1.

5. Use a 47mm diameter filter paper to fit into the filter holder in the outlet section. Use Reeve Angel 934AH glass fiber or ultra pure quartz filters which can withstand high temperature conditions. Other glass fiber filters may contain a plastic resin binder which may vaporize during the source test. Place the filter in a dustfree container (petri dish).
6. Weigh the collection plate inserts and backup filter to 0.01 mg and record. Try and keep the collection plate inserts as dustfree as possible. Also, it helps to use a Teflon gasket on both sides of the backup filter to prevent the filter from sticking to the screen.
7. Assemble the impactor, being careful not to alter the weights of the collection plate inserts. Place the filter collar, screen, and perforated plate backing into the outlet section making sure that the dowel pin is properly aligned and seated. Place a "donut" collection plate onto the filter collar and seat the #7 jet stage labeled Collection Plate Insert into the collection plate. Place the jet stage with the smallest jets first (the jet stages should have the number of jets and the jet diameter inscribed on the back or downstream side of the jet stage so one can identify them). Continue to alternately stack the jet stages and collection plates, mounting the correct collection plate insert, with either grease or fiber substrate, in each collection plate. While placing the jet stages, check to see that the o-rings are in place. Then slide the outer cylinder of the impactor over this stack of jet stages and collection plates and screw the cylindrical casing onto the treads of the outlet section (sometimes Teflon tape should be placed on these threads to give a good seal and to prevent the threads from becoming "stuck"). Secure the outlet section to the casing taking care that the o-ring is properly seated. Then one can look at the top collection plate inside the casing and see if it looks OK (i.e. if it sits too high such there is insufficient room for the inlet section to be screwed on, then perhaps the jet stages and collection plates are not seated correctly). Then screw on the inlet section. Hand tightening of both sections should be sufficient but the unit may be additionally tightened with a wrench using the wrench flats on both the inlet and outlet sections. Teflon tape or a thin coating of graphite or other high-temperature lubricant should be applied to the threaded portions of the inlet and outlet sections and the casing to prevent galling or seizing of the threads (note that sometimes these threads are chrome plated to reduce the problem of galling or seizing).
8. Periodically (every five to twenty tests, depending on the source) the parts should be thoroughly cleaned with a suitable solvent such as ethyl alcohol, acetone, or toluene. Avoid getting the cleaning fluid on the o-rings. An ultrasonic cleaner is useful here. Finally, reclean with soap and water.

2. Necessary Source Test Equipment

A schematic diagram of the source test sampling train is presented in Fig. 14. This train is similar to the EPA Method 5 or EPA Mth 17 particulate sampling train The schematic diagram shows the basic parts of the sampling train, but does not show the instack pitot tube (for isokinetic monitoring of the stack gas velocity) or the pressure and temperature gages in the sampling train.

Fig. 14 UW Mark 3 Cascade Impactor Sampling Train

The list of supplies and apparatus shown below is presented to assist one in identifying the items which might be needed for your operation of the Pilat (U. of W) Mark 3 and Mark 5 Cascade Impactors.

1. Electrical cord long enough to reach test site (for the vacuum pump).
2. S-type pitot tube with tubing to connect the two pressure taps to a gas pressure gauge).
3. Manometer or Magnehelix draft or pressure gauge in the 0 - 2.0 or 0 - 5.0 inches water range (to measure pressure drop with pitot tube).
4. Thermocouple assembly (for impactor exit gas temperature).
5. Thermometer (for stack gas temperature).
6. Thermometer (temperature of gas in last impinger in sampling train).
7. Cylinder for taking plates and stages from impactor.
8. Pipe wrenches and/or clamp pliers.
9. Gloves.
10. Rags.
11. Teflon pipe fitting tape.
12. Extra lengths of vacuum hose.
13. Three inch by 1/4 in. bushing with 3/16 in. hose connection for obtaining static pressure reading (depending on diameter of sampling port).
14. Stop watch (to record sampling time).
15. Assembled Pilat (U of W) Source Test Cascade Impactor.
16. Additional prepared collection plate inserts and backup filters with Teflon gaskets for each test. These should be assembled in labeled petri dishes.
17. Impingers and necessary solvent for each test.
18. Ice bath container.
19. Sampling probe (steel pipe connecting cascade impactor outlet to vacuum hose).
20. Dry gas meter (with temperature and pressure gauges).
21. Leakless vacuum pump (check the vacuum pump to see if it is working correctly before leaving for test site).
22. Sampling nozzles of desired diameters. Bring extras in case the inlet edge of the sampling nozzle gets damaged.
23. Weighing balance for weighing impingers to 0.1 gram.

3. Source Test for Particle Size Distribution Measurement

1. Measure the duct diameter and select the gas velocity traverse points per EPA Reference Method 1 "Sample and Velocity Traverse for Stationary Sources".
2. Measure the gas velocity profile inside the duct at the traverse points. Use a type S pitot tube and a differential pressure gauge (inclined manometer or Magnehix gage) and a thermometer or thermocouple (to measure the gas temperature) per EPA Reference Method 2 "Determination of Stack Gas Velocity and Volumetric Flow Rate (Type S Pitot Tube).

A detailed description of procedures is in the Federal Register, August 18, 1977.

3. Record the atmospheric barometric pressure and the gas static pressure inside the duct. The static pressure can be obtained by threading (with Teflon tape applied to the threads) a 3-inch by 1/4 in. bushing to the test port (assuming the test port is 3 inches in diameter). Connect a 3/16 in. hose connection into the bushing and connect a hose from the bushing to the manometer.

4. Select the sampling nozzle diameter and calculate the gas sampling rate or the dry gas meter rate (Rm) for isokinetic sampling.
   

d = diameter of the sampling nozzle (inches)
T_m = absolute temperature at the dry gas meter
T_s = absolute temperature of the gas in the duct
V_s = velocity of the gas in the duct (ft/sec)
P_s = static pressure of the duct gas ("Hg)
P_b = barometric pressure ("Hg)
W.V. = percent water vapor in the stack gases

 

If using a standard EPA Reference Method 5 "Determination of type Particulate Emissions from Stationary Sources" sampling train, the appropriate orifice gas pressure drop should be calculated to provide for isokinetic sampling.

5. Decide on the actual particle sampling test time which is estimated from the particle mass concentration (grain loading) and the desired deposition on the collection plates (typically the source test duration may range from 10 to 120 minutes). The desired particle deposition on each of the collection plate substrates needs to be sufficient to weigh (say at least 0.05 milligrams) and not too much such that the particles overload the substrate (say 20 milligrams). The amount of particles which can be sampled on a collection plate depends on the type of particles and the particle diameter. The larger particles are more likely to bounce or re-entrain.
6. Assemble the impactor sampling train starting with the Mark 3 or Mark 5 Cascade Impactor. Use Teflon tape on the sampling probe (typically 3/8" or 1/2" stainless steel pipe) threads. Following the probe are four glass impingers. The first two impingers contain 100 ml. of water or suitable solvent, the third impinger dry, the fourth impinger with silica gel. The first and third impingers are modified bubblers while the second is a standard Greenburg-Smith with nozzle and impaction plate. A dry gas meter and a vacuum air pump follow the impingers. The impingers should be clean. Make sure the connecting hoses are clean and dry. See Fig. 14 for a diagram of the sampling train. Essentially, the sampling train is similar to the EPA Reference Method 5 sampling train. Check the sampling train for gas leaks by plugging the sample probe inlet (i.e. the steel sampling pipe which connects to the Mark 3 Cascade Impactor to the connecting hose) and operating the vacuum pump for a short time and noting the change of the dry gas meter vacuum gauge reading. A properly fitted sampling train should hold a vacuum of 15 in Hg.
7. Obtain a particle sample with the Pilat (UW) Cascade Impactor.

1. Preheat the Cascade Impactor to stack temperature (this prevents water vapor or acid condensation in the impactor). One way to preheat the impactor is to cover the sampling nozzle with aluminum foil and heat resistant tape, and place the impactor inside the stack with the nozzle pointed downstream (in the flow direction) for say 15-30 min., depending on the stack gas temp. Then remove the impactor and take off the nozzle cover.
2. Record the initial volume on the gas meter.
3. Insert the heated impactor back in the stack at a representative location if sampling at one point. Be careful when inserting the impactor not to scrape the nozzle on the sides of the sampling port or the sides of the stack or duct (because it is possible to scrape up some particles. Obtain the particle sample by facing the impactor nozzle upstream and turning the vacuum pump to the desired meter rate.
4. Start the stop watch.
5. Periodically check the gas flow rate throughout the test period to ensure isokinetic flow. To do this, record the gas meter volume and divide by the test time. Note any variations in the meter rate during the test. The gas sampling flow rate through the cascade impactor should remain constant during the test.
6. Record the gas meter temperature and pressure several times during the test period (some record this data every 2 or 3 minutes).
7. After taking the sample, face the impactor downstream and turn off the vacuum pump. If there is a large negative static pressure in the stack (i.e. greater than about 4.0" water), one may wish to have a valve between the impactor and the impingers to prevent the water in the impingers from being sucked into towards the stack because this will mess up the test.
8. Stop the watch and record the elapsed time.
9. Note the final gas volumetric reading on the dry gas meter.
10. Remove the impactor from the stack, again being careful not to bump the instrument. Allow the impactor to cool before disassembly. For sure be careful not to burn yourself because these impactors are made of 316 stainless steel and hold the heat for some time. Note any variations which might pertain to the test including the process operation during the test.
    
    

8.Impactor disassembly and storage of sample for transportation to laboratory.

a. Drain any condensed moisture from the connecting hoses into the impingers. Save the impinger water to:

1. Calculate the quantity of water vapor in the stack by weighing the impingers.

2. Evaporate the water to measure any condensable material or particles small enough to go through the filter (i.e. the "back-half particulate catch").

b. If the collection plates or inserts need to be changed in the field for more tests, disassemble the impactor in a clean location with no wind. Place the collection plates or inserts and filter into the petri dishes to be re weighed. Take care not to contaminate the collection plates or inserts. The use of forceps or tweezers may be helpful.

c. Note which set of collection plates or inserts is used for each test.

4. Sample Analysis

1. Weigh the collection plate inserts (to 0.01 milligram).
2. Weigh the backup filter (47 mm diameter filter on impactor outlet section).
3. Weigh the impinger residue (after evaporation of the water).

5. Analyses of Data

1. Calculate the change in weight for each collection plate insert and the outlet filter.
2. Add up the differences to get the total particulate weight collected by the Mark 3 or Mark 5 Pilat (U. of W.) Source Test Cascade Impactor.
3. Divide the amount collected on each plate insert by the total amount collected to find what percent of the total particle weight was collected on each plate insert.
4. Handbook values of the specific gravity of the material may be used for the particle density (if you wish to use Stokes diameter which requires the particle density). For unknown materials, the particle density may be assumed to be 1.0 gm/cm and the particle sizes reported as the equivalent aerodynamic diameter.
5. For the Mark 3 use Fig. 4 or 5 and for the Mark 5 use Fig. 10, 11, or 12, with the gas flow rate at stack conditions and stack temperature, determine the d for each stage. Or one can calculate the jet stage da50s using equation 1. There are computer programs for doing these calculations (be careful when using computer programs written by others because they might have some "strange" Y50 values).
6. Graph the results on log probability paper with the particle diameter (d) as the ordinate and cumulative percent by weight as the abscissa. We recommend that the particle size distribution line be drawn through each of the data points (rather than trying to draw a straight line or curved line) because this presents the data in an "unedited" format.

G Example Particle Size Measurement Source Test

1. Calculate gas velocity at sampling location

Start by calculating the gas velocity at the location you intend to place the Source Test Cascade Impactor. The data sheet for this example is shown in Fig. 15. Only one gas velocity pitot tube and gas temperature measurement is shown on this data sheet. This example data was taken from a source test on the gases exhausting from a pulp mill Kraft liquor recovery boiler exhaust gas stack test (the 30% water vapor concentration is typical of this source). The gas velocity data measured in this example is a gas temperature T of 290 F (750 R) and a pitot tube pressure differential H of 0.11 inches water with the type S pitot tube. Calculate the gas velocity with a pitot tube calibration factor Kp of 0.83.

2. Calculate the Dry Gas Meter Flow Rate

Calculate the dry gas meter flow rate for isokinetic sampling. Select a nozzle diameter which will give a dry gas meter rate between 0.25 and 0.75 cubic feet per minute. For this example a nozzle diameter of 3/8 inch is selected. The following date is measured at the source test site:

Calculating the gas volumetric flow rate at the dry gas meter, Rm ,

Record the test conditions on the data sheet. Run the source test. The example test started with a dry gas meter reading of 106.34 cubic feet at Time = 0 and ended with a meter reading of 118.95 cubic feet at Time = 20 minutes.

Fig. 15 Data Sheet for Pilat (UW) Cascade Impactor Test

After weighing the collection plate substrates plus particles, the initial weight is subtracted from the final weight for each plate. Divide the change in weight (the weight of particles collected on a certain plate) by the total weight of particles collected on each plate, as shown in Table 3 page . To obtain the cumulative percentages, add the incremental percentages from the filter end.

To obtain the ds From Fig. 4 or 5, the cascade impactor gas sampling flow rate at stack conditions Rs and instack gas temperature Ts are needed. The cascade impactor sampling gas flow rate at stack conditions is calculated from the volume of gas recorded by the dry gas meter and from the volume of water condensed (to account for the water vapor in the sampled gas) corrected to the stack gas temperature and pressure, as follows:

1. Calculation of water vapor volume at dry gas meter conditions

2. Calculation of total volume of gas sampled

3. Cascade Impactor Sampling Gas Flow at Stack Conditions

A comparison of the cascade impactor actual gas-sampling flow rate with the isokinetic sampling rate can be done by dividing the flow Rs by the nozzle cross-sectional area.

Therefore, the actual sampling rate was 6% below the isokinetic rate of 22 ft/sec.

Using the cascade impactor sampling gas flow rate at stack conditions (0.956 acfm) and the stack gas temperature of 290 F (about 300 F) the d's can be obtained from Fig. 4 (page 8) and are listed in Table 3 in the right hand column.

TABLE 3 Size Data for Example Source Test

Plate
Number

Delta Weight
(grams)

Percent
Weight

Cumulative Percent
Weight Less than d50

d50
(microns)

1

0.01742

   20.0

80.0

    13.00

2

0.00734

    8.5

78.0

    10.00

3

0.00821

    9.5

68.5

     3.80

4

0.00967

   11.0

57.5

     1.80

5

0.01245

   14.0

43.5

     0.51

6

0.01506

   17.5

26.0

     0.24

7

0.01328

   15.0

11.0

Filter

0.00937

   11.0

Total

0.08709

  100.0

Plot the cumulative % of particle weight less than the particle d50 size distribution on a log-probability graph (logarithm of particle diameter versus the percent by weight of particles less than a given diameter), such as is shown in Fig. 16. The mass median particle diameter is the particle diameter at which 50% by weight of the particles are smaller.

Fig. 16 Example Size Distribution Graph

Assuming a log-normal particle size distribution, the particle size geometric standard deviation sg is given by:

For this example the particle mass median diameter is about 1.30 microns and the geometric standard deviation can be calculated by:

Of course these two geometric standard deviations (one calculated by the 84.13% diameter / the 50% diameter and the other calculated with the 50% diameter / the 15.87% diameter)are different because the particle size distribution is not a straight line (not really log-normal). One approach to obtain a more representative standard deviation is to calculate it as follows:

In general, it is better to present the particle size information in graphical form rather than merely reporting the mass mean diameter and size geometric standard deviation.

II. Impactor Loading Procedure for the Pilat (UW) Mark 3 and 5 Cascade Impactors

Note: All parts must be cleaned prior to assembly

A. Outlet Section of the Pilat (U. of W.) Mark 3 or 5 Cascade Impactor

1. Secure outlet section of impactor in vise or on a lab bench or table.
2. Place o-ring in groove at base of threads.
3. Teflon tape threads, approximately 1 1/2 wraps.
4. Place filter support plate and fine screen in the outlet section, support plate first, fine screen second.
5. When using a second (blank) filter for quality assurance purposes (this blank filter can be used to determine if there are chemical reactions of a stack gas constituent with the filter material which changes the filter weight; for example gaseous HF may react with the glass resulting in a weight loss), one Teflon ring or gasket (marked BF) and blank filter are placed on top of the fine screen. The second Teflon ring or gasket marked BF is then placed over the filter. The filter assembly is then placed on top of the blank assembly as follows: Teflon ring, filter, Teflon ring. If stack temperatures exceed 425 F, Kapton may be substituted for the Teflon.
6. Check to insure that the Teflon rings and filters are lying flat. Place the filter collar onto the outlet section and turn gently until the alignment pin on the top of the outlet section matches up with the hole in the filter collar. The Filter collar should now be properly seated.
7. The inside edge of the top Teflon ring should be visible along the inside edge of the collar. If not, it should be replaced since the collar will cut the filter when the impactor is tightened.
8. Place an o-ring in the groove at the top edge of the filter collar.

B. Impactor Substrates

1. If a blank collection plate is to be included in the run, place the proper foil substrate in a collection plate. Place the collection plate on the filter collar with the substrate facing the impactor outlet (upside down).
2. Starting with the foil designated as the last collection substrate, place the foil in a collection plate and put it on top of the blank with the substrate surface facing up. If a blank is not used, this plate is placed directly on the filter collar with the substrate surface up.
3. The last jet plate (smallest flow area) is then placed on top of the collection plate. The jet plate should be oriented so that the jets are at the bottom, closest to the collection plate.
4. Place o-ring in groove at top of jet plate. The remainder of the donut-shaped substrates should be loaded in the same way and added to the stack, alternating collection plates with jet plates.
5. When the last of the donut-shaped substrates and the corresponding jet stages have been placed in the stack, the first jet stage collection plate should be loaded with the disk-shaped substrate and placed on top of the stack.
6. Align the stack of jet stages and collection plates, then slide the impactor cylinder (outer shell) over it. Tighten the cylinder onto the outlet section until it seals against the outlet section o-ring.

C. Inlet Section of the Pilat (U. of W.) Mark 3 Cascade Impactor

1. Wrap Teflon tape approximately 1-1/2 times around the threads of the inlet section.
2. Screw the inlet section (with connecting tube attached) into the impactor shell. Hand tighten only. Excessive tightening of the inlet section into the shell can cut the back up filter.

D. Precollector (Either PCSC Precollector or the Right Angle Sampling Attachment)

Note that it is not necessary to use the Precollector or the Right Angle Sampling Attachment (i.e. one can use a sampling nozzle connected to the inlet section)

1. Wrap threads of top and bottom sections and the nozzle 1-1/2 times with Teflon tape.
2. Screw bottom section and nozzle into precollector body.
3. Remove foil from petri dish and curl slightly.
4. Insert foil into precollector body, greased side facing nozzle inlet.
5. Screw top onto precollector body.
6. Tighten precollector onto connecting tube (impactor body). Make sure the precollector is aligned such that the bend in the connector tube offsets the nozzle.

E. Leak Check

1. Connect the inlet of the precollector/impactor assembly to the suction end of a pump by attaching a hose to the nozzle. Cap the outlet.
2. Pull a vacuum of approximately 10 inches of mercury on the assembly and observe the vacuum pressure for about a minute.
3. After this observation period is over, release the vacuum at the inlet, not at the outlet. Opening the outlet to ambient can rupture the filter.
4. Pressure losses of approximately four to five inches should be expected. Drastic leaks indicate loose fittings or missing o-rings. Attempt to correct any leaks. Use of a slightly positive pressure ( 6 inches of water) and a soap solution may help to locate the source(s) of the leak(s).
5. If small leaks are present which can not be corrected, a leak check should be performed without the precollector.
6. Enter leak check data in appropriate space on impactor lab sheet.

F. Wrapping

1. On two small pieces of high temperature tape, write the impactor run code. Place one on the impactor and the other on the precollector (if a precollector is used).
2. Wrap impactor body and precollector with aluminum foil and secure with tape. This protects the cascade impactor from getting particles and dirt on the outside surface and makes it easier to clean up after the test.
3. Rewrite impactor run code on the wrapped impactor body.

III. Impactor Unloading Procedure for Pilat (UW) Mark 3 and 5 Cascade Impactors

A. Preliminary

1. Hold impactor upright at all times (this helps to prevent particles from falling off the collection substrates).
2. Remove foil wrapping and blow off any loose dust from the impactor/precollector assembly with compressed air or gas (Effaduster) or wipe clean with a cloth or paper towel. Cover nozzle with thumb (or tape) to prevent blowing into precollector. Exterior surfaces should be clean to prevent contamination during unloading.
3. Secure outlet section in vise for disassembly.
4. Remove impactor lab sheet from run sheet notebook. As substrates are unloaded, observations such as broken particle peaks or loose particulate should be noted on the lab sheet. Sometimes it is helpful to take photographs of the particle collection substrates to illustrate the colors or other characteristics of the sampled particles.

B. Sampling Nozzle and Impactor Inlet Section (or Precollector)

1. Unscrew the inlet section with sampling nozzle (if a precollector was not used).
2. If a precollector was used, separate the precollector from the impactor inlet section where the precollector connector tube is screwed into the inlet section.
3. Unscrew top of precollector and remove foil from body (assuming a foil was used inside the precollector body) or pour particles from the body, placing the particles in a pre-weighed glass petri dish or into a preweighed foil "dish".
4. Remove nozzle from precollector body and using clean dry brush, brush any loose particulate on the inside of the nozzle or the top section of the precollector onto the foil. Place nozzle to the side so that it can be washed in Step 7.
5. Separate body from bottom section and (using the same brush) brush any loose particulate in either section onto the foil. Place the brush to the side so that it can be washed in Step 7. Note: particulate on the inside of the precollector exit tube (bottom section) should be transferred to the first substrate of the impactor.
6. Carefully fold foil in half twice and then loosely fold a small ridge on each side to prevent loss of particulate. The fold must be loose to permit drying during desiccation.
7. The tube connecting the precollector to the impactor should remain connected to the inlet section of the impactor. Any particulate in this tube should be brushed (using a second clean brush) onto the first substrate in the impactor. This is best done by tapping the sides of the tube over the substrate, then brushing the interior of the tube with a small nylon bristle brush. The interior surface of the precollector exit tube (bottom section) should also be brushed onto the first substrate in the impactor. Set tube/inlet assembly, precollector exit tube, and brush, to the side so that they can be washed in Step 8.
8. Wash down techniques as described in Section 4.2 of EPA Reference Method 17 may be used to rinse the nozzle and brush with acetone. The collected rinse must then be evaporated, desiccated, and weighed on a precision balance. Note: It is important that the brushes used were previously cleaned by an acetone rinse and allowed to dry before being used to brush the particulate. Because short straight nozzles are used, it may not be necessary to perform this nozzle wash down.
9. The precollector exit tube and the connecting tube between the precollector and impactor should also be washed into a second sample bottle. The evaporated dry weight gain from this wash down is assigned to the impactor's first jet stage.

C. Impactor Substrates

1. Loosen and remove the cylindrical casing shell of the impactor from the outlet section (fastened in the vise). The impactor inlet section with attached connecting tube was removed in Step B-8 above.
2. Inspect the interior of the cylindrical casing shell for any evidence of internal leakage's. If any such evidence is found, make a notation and try to identify the stage(s) with which it was associated. Leakage of particles from the collection plates and jet stages to the casing shell inside surfaces is very rare (i.e. there is really no way the particles should get there) and so if there are particles seen on the inside of the casing, something went wrong.
3. Remove the #1 jet stage collection plate from the stack (this collection plate will be on top of the stack). If the o-ring of the jet plate directly beneath the collection plate sticks, remove both plates from the stack.
4. Remove the disk-shaped substrate from the #1 collection plate and place in its labeled petri dish. This is best accomplished by grasping the edge of the substrate with tweezers and rotating the disk gently.
5. If the jet and collection plates are stuck together, gently push the collection plate horizontally until the o-ring seal releases.
6. Each donut-shaped substrate should be removed in the same way and placed in its respective labeled petri dish (use glass petri dishes if the samples are to be dried in an oven).
7. Any particulate present on the surface of a jet plate should be brushed onto the substrate directly beneath it unless it is obvious that the material was removed or re entrained from the preceding substrate.

D. Outlet Section

1. Gently lift the filter collar and brush any part of the filter adhering to it into the foil envelope. If Teflon gaskets are used to "sandwich" on both sides of the outlet filter, the outlet filter should be protected against adhering to either the filter collar or the filter support screen.
2. Removing the outlet section from the vise, insert the handle of the brush into the outlet neck and gently lift the filter support plates.
3. Remove filters and Teflon rings from the plates and place respectively, dirty filter and two Teflon rings into the foil envelope and clean filter with two Teflon rings (labeled "BF") into its labeled petri dish.

E. Reloading Preparation

1. All parts of impactor and precollector should be blown off with compressed air or gas (Effaduster) before being reloaded as described in Table 4-2. Or wipe the parts clean with a cloth or paper towel.
2. Wiping the impactor, plates, etc. with a clean cloth or clean tissue is helpful to clean off any remaining soiling or particles.

IV. Reference Literature

A. "Source Test Cascade Impactor" by M. J. Pilat, D. S. Ensor, and J. C. Bosch. Atmospheric Environment, 4 671-679, (1970).

B. "Size Distribution of Particulates Emitted from a Horizontal Spike Soderberg Aluminum Reduction Cell." by T. R. Hanna and M. J. Pilat. Air Poll. Control Assoc. Journal, 22 533- 536. (1970).

C. "Size Distribution of Aerosols from a Kraft Recovery Furnace." by J. C. Bosch, M. J. Pilat, and B. F. Hrutfiord. Technical Association of Pulp & Paper Journal, 54 1871-1875. (1971).

D. "Relationship of Plume Opacity to the Properties of Particulates Emitted from Kraft Recovery Furnaces." by S. Larssen, D. S. Ensor, and M. J. Pilat. Technical Association of Pulp & Paper Journal, 55. 88-92. (1972).

E. "Size Distribution of Particulate Emissions from a Pressurized Fluidized Bed Coal Combustion Facility." by M. J. Pilat and T. W. Steig. Atmospheric Environment, 17 .2429-2433. (1983).

F. "Airborne Particulate Emissions from a Chromic Acid Anodizing Process Tank" by R. C. Pegnam and M. J. Pilat, J. of Air & Waste Management Assoc. 42 303 - 308 (March 1992).

HTML Revision 2/24/1999