|<< CTC Home Page|
This procedure describes a method of leak testing cyclone sampler assemblies. It is intended to reveal excessive leaks at the
1.2 ENVIRONMENTAL AND OPERATIONAL LIMITS:
No limitations need to be placed on the unit under test (UUT) or the test system operation regarding facility temperature, pressure, or humidity. No warm-up or initial stabilization time is needed.
1.3 TEST METHOD AND ACCURACY:
1.3.1 Tolerances and Limitations:
This procedure is designed to reveal excessive leaks at the connections between the Cyclone assembly parts. Aside from assuring a good fit between the nylon parts, testing for leaks in the parts between the pre-weighed filter input and the air sampling pump is the most important portion of the assembly to minimize leaks.
A 25% change in pressure gauge reading is used as the pass/fail criteria in the main part of this leak test procedure. When using the equipment setup described in this procedure, this criteria of 25% has been shown to be a good, achievable, and practical number for checking the general leakage in the cyclone assembly connections from the Vortex Finder to the air sampling pump, especially between the pre-weighed filter and the air sampling pump. Although a much higher leakage rate can be tolerated at the Plastic Filter Adaptor, the same 25% pass/fail criteria is applied to the whole chain of parts from the Vortex Finder to the air sampling pump.
In using the 25% pass/fail criteria it is very important that the total volume of air in the test system (including the pressure gauge internal volume) remains constant (approx. 335 cc). That is, if different test equipment is used for this test procedure which causes a change in the air volume of the test configuration, the pass/fail criteria must likewise be changed to a different number.
See Appendix 1 for a detailed discussion of the tolerances and limitations.
1.3.2 Summary of Test Method:
A comprehensive leak test of the Dorr-Oliver Cyclone assemblies can be accomplished by an inspection of the nylon parts for proper fit, applying a partial vacuum to the tubing-filter-adaptor-vortex combination and monitoring the leakage rate, and finally running the pump/cyclone assembly at 1.7 LPM and checking for a pump fault when the cyclone input is sealed closed. Sub-sections 184.108.40.206 through 220.127.116.11 summarize these three tests.
18.104.22.168 Nylon Part Inspection:1.3.3 Description of the Cyclone Assembly Parts:
Figures 1.1 and 1.1a show the MSA and the Sensidyne/Bendix BDX 99R cyclone assemblies. Below them are Figures 1.2 and 1.2a showing the same assemblies disassembled and the parts identified.
The following is a brief description of the cyclone assembly parts shown on the previous page in Figures 1.2 and 1.2a:
1) Cyclone Sampling Line: the tubing that connects between the Cyclone assembly and the air sampling pump. This is the same tubing used with the Cyclone during sampling.1.3.4 Discussion of Test Method:
The pressure difference between the inside of the Cyclone Assembly and atmospheric pressure outside the Cyclone Assembly increases for each stage of the Cyclone Assembly through which the air moves. Each restriction to the flow increases the pressure drop in that stage and in subsequent stages. The Cyclone Body and Vortex Finder offer very little resistance to the air flow, and the pressure decrease is very low. On the other hand, the filter is a large restriction, causing the pressure decrease (vacuum) after the filter to be much larger than in the Cyclone Body. As the filter collects more and more of the sample, the pressure drops even more.
Even though the pressure decrease (vacuum) in the Cyclone Body and Vortex Finder is small, it is important to keep the leaks to a minimum at the connection between the Cyclone Body and the Vortex Finder. Excessive leaks at this connection could disturb the cyclone pattern of the air flow in the Cyclone Body. Excessive leaks in the other connections of the Cyclone Assembly need to be kept to a minimum as well, since the cyclone operation depends on a given air flow rate through the Cyclone Body for proper operation. Leaks in the filter cassette and subsequent parts are important since the pressure drop in these parts can be as much as 37" H2O below that in the sampling environment. Leaks in the filter cassette and subsequent parts not only alter the effective flow rate, but can introduce dusty contaminants into the air sampling pump and possibly damage it.
A nominal pressure (vacuum) in each stage of the Cyclone assembly can be determined by measuring the effect each part of the Cyclone assembly had on the total loading of the air sampling pump. Keeping these different levels in mind helps in understanding the relative impact each leak can have on the functioning of the assembly. For an air flow rate of 1.7 LPM, the approximate pressure (vacuum) levels in a Dorr-Oliver Cyclone Assembly are:
If the Vortex Finder, Cyclone Body, Grit Pot, and Grit Pot
The O-Ring, Tubing, and Filter Leak Test test does not include the Cyclone Body or the Grit Pot which are inspected for proper fit and are included in the final general leak test. This test quantitatively checks the connections between the Sampling Line, Flow Connector Assembly, Pre-weighed Filter, Plastic Filter Adaptor, and the Cyclone Vortex Finder. The 4 - 10" H2O range used in this leak test procedure falls within the normal load range expected in a typical sampling session.
The volume of gas released through an opening is directly proportional to the size of the hole, the gas pressure, and the time lapsed, assuming the supply of gas is unlimited. If the container for the gas is limited by a fixed volume, the pressure will drop as the gas is released. The pressure in the fixed volume will change at a rate directly proportional to the size of the hole and the time that has lapsed, and inversely proportional to the size of the fixed volume supplying the gas. For a fixed volume and time, the leak can be measured by the percent drop in pressure for that given time duration.
This method of quantitizing leaks is limited, since any significant change in the air volume of the test configuration requires changing either the percentage limits or the measurement time. When using this procedure, the test volume must remain constant. That is why it is necessary to close off the test volume at the tubing between the aspirator bulb and the "Tee" fitting. Keeping the aspirator bulb in the measurement volume would significantly increase the test volume and alter the test results.
The final leak test is simple and straight forward, but only gives a general, non-quantitative, check for leaks that are so large that even the fault indicator of the air sampling pump fails to respond because of them. The primary purpose for this test is to provide a final leak test of the entire cyclone assembly, especially the parts upstream from the filter. It is only a general test which does not provide a quantitative measure, and does not adequately check the parts downstream from the filter. Periodic comprehensive leak testing should include all of the steps presented in this leak test procedure.
This section describes the test equipment used to perform the leak tests described in this procedure.
2.1 EQUIPMENT LISTING:
2.2 PICTURE and BLOCK DIAGRAM OF EQUIPMENT SET-UP:
3.1 SAFETY CONSIDERATIONS:
No special safety considerations are required for this procedure. If alcohol is used to help dry the inside of the Cyclone Body, be sure to observe all related safety precautions pertaining to its use.
3.2 EQUIPMENT CONSIDERATIONS:
The Dwyer pressure gauge is designed to take overloads. It is rated at 15 psi ( 415" H2O). Therefore, pegging the meter with a vacuum from the aspirator should not damage the meter.
3.3 PRELIMINARY STEPS:
3.3.1 Cyclone Preparation and Cleaning:
22.214.171.124. Unscrew the grit pot from the Cyclone. Empty the grit pot by turning it upside down and tapping it gently on a solid surface.3.3.2 Test Equipment Preparation:
This test of the pressure gauge (2010) is only necessary if you suspect a leak in the pressure gauge:
126.96.36.199. Connect the test equipment as in Figure 2.2, but leaving the Cyclone assembly disconnected.
188.8.131.52 With your finger over the open end of the "Tee" fitting, squeeze and release the aspirator bulb, and when the pressure gauge is reading a vacuum above 5" H2O, tightly fold the tubing between the "Tee" and the pressure gauge. If the reading is beyond full scale, release some of the vacuum (allow air in) until the reading is below full scale, but above midscale.
184.108.40.206 Observe the pressure gauge reading for about a minute. The needle should hold steady. If the reading drops more than a ½" H2O in a minute, the pressure gauge should be fixed.
4.1 Nylon Parts Inspection:
The Vortex Finder, Cyclone Body, Grit Pot, and Grit Pot
4.1.1 With the Cyclone disassembled and cleaned as in section 3.3.1, inspect the Cyclone Grit Pot, Cyclone Body, and Cyclone Vortex Finder for cracks. Replace cracked parts.
4.1.2 Inspect how well the Cyclone Vortex Finder fits into the Cyclone Body. Compare with a new Cyclone, if possible. Replace worn or damaged parts that would cause excessive leaks.
4.1.3 Install the Cyclone Grit Pot and Grit Pot
Note: Spare replacement4.1.4 Record the results of this check (PASS/FAIL) on the data sheet, if used.
This test looks for leaks at the connections of the Sampling Line, Flow Connector Assembly, Pre-Weighed Filter, Plastic Filter Adaptor, and the Vortex Finder. A partial vacuum is applied to the Cyclone assembly (minus Cyclone Body) with the Vortex Finder sealed closed. The pressure is monitored for 30 seconds, and if the pressure reading changes less than 25%, the UUT passes.
4.2.1 Connect the Cyclone assembly to the pressure gauge and aspirator as shown in Figures 2.1 and 2.2, using the actual Cyclone parts that are to be used later at the survey site, including the same Cyclone Sampling Line.
Note: If the Sampling Line of an MSA Cyclone Assembly ever becomes disconnected from the Flow Connector Assembly, do not reconnect the stretched-out end of the tubing to Flow Connector Assembly. Cut off the stretched end, and then connect the new unstretched end to the metal shaft. When attached, the tubing should not move freely on the metal shaft.MSA Cyclones: Be sure to install the Metal Clamp (or a #2-012 Clamp Substitution/Spacer
Sensidyne/Bendix model BDX 99 or BDX 99R cyclones: A #2-010
4.2.2 Seal the opening to the Cyclone "Vortex Finder" with a rubber or plastic cap, plugged up tubing, or by pressing the soft tissue of your finger over the hole. Whichever method is used, be sure to hold the assembly together with your hand, since the parts may leak more it they are not held in place.
Note: Rubber caps (hollow plug sleeve stoppers), or tubing capped at one end, are provided in the leak test kit accessory package to serve as plugs. Do not use putty or tape on this inside surface of the Vortex Finder; they leave sticky residues which can collect dirt and grit during sampling.4.2.3 Squeeze and slowly release the aspirator bulb until the gauge reads between 4" H2O and 10" H2O vacuum, and then immediately fold (or clamp) the tubing between the "Tee" and the aspirator bulb.
If the vacuum reading is too high, release some of the vacuum at the Vortex Finder or at the fold in the tubing. Be patient. This may take a little practice.
Important Note: Remember to fold (or clamp) the tubing between the aspirator bulb and the "Tee". Do not rely on the Aspirator bulb valve to hold the vacuum. That would introduce the aspirator bulb into the test volume and change the results. Also, the aspirator valves can leak.4.2.4 Observe and remember the pressure gauge reading. It will be recorded when your hands are free.
4.2.5 After 30 seconds, observe and record the pressure gauge reading again. If the vacuum dropped less than 25%, the unit passes the leak test.
If the pressure drops more than 25%, corrective action is necessary. This probably means one of the
4.2.6 Record the results (PASS/FAIL) of this leak test on the data sheet, if used.
4.2.7 Remove the Filter Adaptor, rotate it so the opposite ends of the adaptor will connect at the filter and the Vortex Finder, reinstall the Filter Adaptor, and leak test again as in steps 4.2.2 through 4.2.6.
Note: The test described here in step 4.2.7 is not required if there is no chance the Filter Adaptor will be reversed during sampling (i.e., if one end is marked so it is obvious which end goes where).4.3 Pump-Fault Test:
As a final test, the entire Cyclone assembly (including the Cyclone Body and Grit Pot) is connected in the normal sampling configuration, the air sampling pump is set to the normal sampling flow rate, and then the inlet to the Cyclone is sealed closed. If the pump bears down and goes into a fault mode, the assembly passes the final leak test.
4.3.1 Disconnect the Cyclone assembly sampling line from the "TEE" fitting and assemble the entire Cyclone Assembly (including the Cyclone Body and Grit Pot) in its normal sampling configuration to the air sampling pump.
4.3.2 Set the air sampling pump to RUN at a flow rate of 1.7 LPM ±0.2 LPM.
4.3.3 Place your finger or a piece of tape over the Cyclone inlet.
4.3.4 If the pump responds to the added load and faults, the Cyclone passes this test.
If the pump does not fault, either the Cyclone leaks excessively, or the pump is malfunctioning.
4.3.5 Record the results of this leak test including the serial number of the pump on the data sheet, if used.
4.4 Data Record:
A data sheet is provided with this procedure for optional use. Alternate methods of recording the results may be a record log, and/or on the OSHA -91A.
4.4.1 Record your name, the date, repairs, or any other comments on the data sheet, if used.
5.1 If the Cyclone Assembly fails the main leak test of section 4.2, try leak testing each of the following configurations, watching for which added part causes an increase in the leak rate:
a) Flow Connector Assembly and Cyclone Sampling Line only (MSA). (Seal with moist finger.) This tests for a good seal between the Sampling Line and the Flow Connector Assembly.5.2 If the Plastic Filter Adaptor (coupler) is suspect, try reversing it. If the results change, then one of its
If it fails both ways then both of the
If the seal is not improved by a new
The Plastic Filter Adaptor has two
5.3 A very likely source of leaks, especially with age, is the
If the tests in step 5.1 show the leak is in the MSA Flow Connector Assembly, and replacing the
5.4 If you need an
5.5 To test the Pre-weighed filter for leaks, connect it to the leak tester by itself using a piece of 1/4" ID tubing, seal the other end of the filter either with a moist finger or with the red plug that came with the filter, and check for leaks. If a pre-weighed filter fails the leak test, replace it.
5.6 Sometimes someone will remove the Sampling Line from the MSA Flow Connector Assembly. This should not be removed unless necessary. If the tubing (Sampling Line) has been removed and you are going to reconnect it, first cut off and discard the stretched out end. Then warm the new end of the tubing (can run hot water over the outside of the tubing), and stretch it over the flared end of the Flow Connector Assembly about 2 cm (or 3/4 inch) beyond the flared end. Assure the flared end of the Flow Connector Assembly makes a good seal with the tubing (i.e., the metal tube will not rotate inside the Sampling Line). Leak test this new connection before using.
This sample data sheet shows the link between the data sheet and steps in the testing procedure. If the results are recorded, a blank data sheet is provided with this procedure which may be copied for optional use. Alternate methods of recording the results could be a record log, and/or on the OSHA -91A.
As stated in section 1.3.1, this procedure is designed to reveal excessive leaks at the connections between the parts of a Cyclone Assembly. Aside from a good fit between the nylon parts, the portion of the Cyclone Assembly between the pre-weighed filter input and the air sampling pump is the most important portion of the assembly to check for leaks. When using the test equipment setup described in this procedure, a 25% change in pressure gauge reading has been shown to be a good, achievable, and practical number to use for the pass/fail criteria in checking the general leakage in the cyclone assembly for parts between the Vortex Finder and the air sampling pump, especially between the pre-weighed filter and the air sampling pump. Since the normal operational pressure loading between the Vortex Finder and the Pre-Weighed Filter is independent of the loading of the filter by the collected sample, a much higher leak can be tolerated during leak testing between these parts, but for simplicity's sake, the 25% pass/fail criteria is applied to the whole chain of parts.
In the section of the cyclone assembly between the Pre-Weighed Filter and the air sampling pump, the pressure drop (vacuum) can range from 2.5" H2O when the filter is clean, to about 35" H2O when the filter is heavily loaded. This leak test assures that even under the extreme conditions of 35" H2O loading (very dirty filter), the effective nominal flow of 1.7 LPM would be reduced by less than 1% due to leakage.
When this leak test was developed, practical limits for the leak test were derived by examining new and used Cyclone assemblies. The procedure was tested by both laboratory and field personnel. The trial leak tests showed that new cyclones parts produced pressure gauge reading changes of less than 10% during a 30 second test period. Changes in pressure reading for UUTs with worn
Analysis of the test configuration provides interesting results. With the test volume sealed, a 5 cc change was induced in the test volume using a syringe to observe the resulting pressure gauge reading. The 5 cc volume change produced a change in the pressure gauge reading of approximately 6" H2O, and a 2.5 cc volume change produced a pressure reading change of 3" H2O. Now assume a leak test which begins with a 10" H2O reading and drops by 30% in 30 seconds. That would then represent a leakage of 2.5 cc per half minute at an average pressure of 8.5" H2O (average of the start pressure of 10" H2O and the ending pressure of 7" H2O). Restating this we could say that in using this leak test procedure, a 30% pressure gauge reading change during a period of one half minute from an initial 10" H2O reading represents a 5 cc per minute leak at 8.5" H2O. Proportionally, the same leak at 35" H2O would be 20.6 cc per minute (5 ÷ 8.5 × 35), or 1.21% of a 1.7 LPM air flow rate. To determine what leakage a 25% pressure change would represent, we would proportionally say a 4.16 cc volume change produces a 5" H2O change in pressure gauge reading (a 2.08 cc change produces as 2.5% change) and repeat the above logic. Thus, a 25% change from a 10" H2O reading represents a 16.6 cc per minute leak at 35" H2O (4.16 ÷ 8.75 × 35), or 0.98% of a 1.7 LPM air flow rate. A leakage of less than 1% under the heaviest loading sounds reasonable.
This procedure relies on measurement of pressure differences up to 10" H2O (inches of water). To avoid the mess associated with using a water manometer, a Dwyer Magnehelic pressure gauge is used to monitor the pressure (vacuum). Dwyer model 2010 has a full scale indication of 10" H2O (inches of water). A water manometer or other makes and models of pressure gauges may be used, but if the total volume of air in the test system (including the pressure gauge internal volume) is different from that described in this procedure (approx. 335 cc), different limits for the change rate of the pressure reading (or an adjustment in the air volume of the test configuration) would needed.
Note: The 5 cc analysis of the test configuration was performed when the atmospheric pressure was 999.1 mbar (401" H2O). With the temperature held constant, the volume of the test configuration can be calculated using P1V1 = P2V2 . Thus 401.1 x Vtest = (401.1 - 5.9) × (Vtest- + 5), which yields a value of Vtest = 335 cc, most of which is inside the pressure gauge. If a substitute pressure gauge has much less internal volume, the test configuration will be too sensitive to leaks, unless additional volume is added to the test configuration to compensate. Likewise, a pressure gauge with larger internal volume will be less sensitive to leaks, requiring a smaller percent change in pressure gauge reading for the pass/fail limit.The Dwyer Magnehelic pressure gauge accuracy is guaranteed by Dwyer to be within ±2% of full scale. The OSHA CTC plans to calibrate these gauges on a periodic basis, but in applying this procedure, the absolute accuracy of the pressure gauge is not as important as the relative linearity between pressure levels, since the purpose of the pressure gauge is to compare a second reading relative to an initial reading. As long as the gauge's relative linearity is reliable, an absolute uncertainty of ±10% at the full scale reading is more than adequate.
The final Pump-Fault leak test described in this procedure primarily checks the parts upstream from the filter, i.e., the Filter Adaptor, Vortex Finder, and Cyclone Body. To fail this leak test the total leaks must amount to 95% or more of the 1.7 LPM setting while under near fault loading of the pump and all of the Cyclone parts. Under normal operating conditions all of the parts are not subjected to such extreme loading. Assuming a 37" H2O fault level for the pump, the estimated leakage which would occur under normal operating conditions for a cyclone assembly failing this test would be greater than 6.7-100% (2.5/37 to 37/37) between the filter and the pump, or 0.7% (0.25/34.5) at the Filter Adaptor, or 0.4% (0.15/34.4) at the Vortex Finder and Cyclone Body connection. These percentages were calculated by dividing the normal loading at that location by the minimum loading needed at that point to cause a pump fault, times 100 percent.
Read the entire procedure before using this summary list to leak test the cyclone. This summary is provided for the convenience of those who have already read and are familiar with the procedure. A copy of this page may serve as a checklist for the principle steps of the procedure.
PREPARATION and INSPECTION: