Key 1 Pre-filter for the test air 2 Fan with speed regulator 3 Air heater 4 Aerosol inlet in the duct 5 Aerosol generator with conditioning of supply air and aerosol flow regulator tempe
Set-up of the test apparatus
The test apparatus setup, illustrated in Figure 1, is applicable for both monodisperse and polydisperse aerosols The primary distinctions between these two types are found in the measurement techniques employed and the methods used for aerosol generation.
1 Pre-filter for the test air
4 Aerosol inlet in the duct
5 Aerosol generator with conditioning of supply air and aerosol flow regulator
6 Measurement of atmospheric pressure, temperature and relative humidity
8 Sampling point for upstream particle counting
13 Sampling point and partial flow extraction, downstream
17 Computer for control and data storage
18 Measuring system to check the test aerosol
Figure 1 — Diagram of test apparatus
An example of a test rig is shown in Figure 2 (without particle measuring equipment)
5 Dampers to adjust test and sheath air
6 High efficiency air filter for the test air
7 Aerosol inlet in the duct
15 Sampling point for particle size analysis
17 High efficiency air filter for the sheath air
19 Measurement of sheath air speed
21 Flow equalizer for the sheath air flow
23 Screening (linked to the filter mounting assembly during the testing)
24 Traversing probe arm with downstream sampling probe
Figure 2 — Test duct for scan testing
EN 1822-2 provides essential information on the generation and neutralization of aerosols, including the types of equipment required and detailed descriptions of the measuring instruments necessary for testing.
Test duct
Test air conditioning
The test air conditioning unit contains the equipment needed to condition the test air flow (see Clause 7)
The test air flow must be prepared to meet the specifications outlined in Clause 7, ensuring that it remains within the specified limit values throughout the efficiency testing process.
Adjustment of the volume flow rate
A suitable provision, such as adjusting the fan speed or using dampers, will enable the production of a volume flow rate with a reproducibility of ± 3% The nominal volume flow rate must consistently stay within this range during testing.
Measurement of the volume flow rate
The volume flow rate must be determined using a standardized or calibrated method, such as measuring the pressure drop with equipment like orifice plates, nozzles, or Venturi tubes, in compliance with EN ISO 5167-1.
The limit error of measurement shall not exceed 5 % (of the measured value).
Aerosol mixing duct
The design of the aerosol input and mixing duct must ensure that the aerosol concentration at specific points directly in front of the test filter does not vary by more than 10% from the average concentration measured at a minimum of nine evenly distributed points across the duct's cross section.
Test filter mounting assembly
The test filter mounting assembly must securely seal the test filter and allow for proper flow as per specifications, while ensuring that it does not block any part of the filter's passage area.
It is advisable to scan filters for leaks in the same mounting position and air flow direction as they are installed on site.
Measuring points for the pressure difference
To accurately measure pressure, the measuring points must be strategically arranged to capture the average difference between the static pressure in the upstream flow and the surrounding air pressure Additionally, the pressure measurement plane should be located in an area with uniform flow conditions.
In rectangular or square test ducts, pressure measurement holes with diameters ranging from 1 mm to 2 mm should be drilled in the center of the duct walls, perpendicular to the flow direction These four measurement holes must be connected by a circular pipe.
Sampling, upstream
Samples are collected upstream using one or more sampling probes positioned before the test filter The diameter of the probe must be selected to ensure isokinetic conditions at the specified volume flow rate during average flow velocity Sampling errors caused by varying flow velocities in the duct can be disregarded due to the small particle size in the test aerosol Additionally, connections to the particle counter should be minimized in length.
The sampling must be representative, defined as when the aerosol concentration measured from the sample does not differ by more than 10% from the mean value established in accordance with section 6.2.4.
The mean aerosol concentrations determined at the upstream and downstream sampling points without the test filter in position shall not differ from each other by more than 5 %.
Screening
To ensure accurate detection and localization of leaks in the filter's edges, gasket, frame, or sealant, the downstream side of the test filter must be completely shielded from surrounding air impurities This can be accomplished by directing a flow of particle-free air around the outer sides of the filter frame, effectively leading away any particles in the areas being scanned.
The scanning tracks must encompass the filter frame, corners, and ideally the space between the filter frame and gasket to identify potential leaks A validation of the test rig is essential to ensure that leaks in these areas are detected with the same probability and sensitivity as those occurring in the center of the filter.
Scanning assembly
General
In addition to the automated testing for leaks, manual scanning is also permitted, provided that the most important parameters for the test procedure are adhered to
Manually moving the probe leads to irregularities in movement across the filter surface, making smooth and even assessments difficult Consequently, quantitative evaluations are often limited or impossible Additionally, tracking the coordinates of leaks and analyzing particle counts is a highly time-consuming process.
In the following, an automatic scanning apparatus is described.
Sampling, downstream
The local resolution for determining particle flow distribution on the downstream side is influenced by sampling conditions To ensure comparability of measurements for local penetration values, it is essential to conduct sampling under standardized conditions.
The probe aperture can be either rectangular or circular, with rectangular probes having a side ratio not exceeding 15 to 1 The probe's area is specified to be (9 ± 1) cm², and the volume flow rate must be selected to ensure that the speed at the probe aperture remains consistent within specified limits.
25 % from the face velocity of the filter (see B.6)
If the probes have a rectangular aperture, then the measuring time can be shortened by using several probes next to each other (for several particle counters)
The probe shall be positioned at a distance of 10 mm to 50 mm from the downstream face of the filter element
In cases of specialized filter designs and exceptionally high face velocities, deviations from the specified dimensional requirements are allowed However, this leads to a conditional assessment of local efficiency as defined by the standard.
Probe arm
The downstream partial flow probe must be securely attached to a movable probe arm, which is engineered to ensure that neither the arm nor its movement mechanisms disrupt the flow near the filter.
Aerosol transport lines
To ensure accurate measurements, aerosol transport lines must efficiently direct particles to the particle counter's measuring chamber with minimal delay and no losses These lines should be kept short and free of sharp bends, constructed from conductive materials with smooth surfaces that do not release particles.
Provisions to move the probe
These provisions include drive, guidance and control to move the probe arm at right angles to the direction of flow with a constant probe speed
The speed of the probe can be selected, and shall not exceed a maximum of 10 cm/s (see B.7) During a run it shall not deviate from the set value by more than 10 %
Provisions must be established to accurately measure the probe's position in the X, Y, and Z coordinates during operation, as well as to reposition the probe over any identified leaks The repositioning accuracy within the downstream cross-section of the test filter should be no less than 1 mm.
Aerosol generation and measurement techniques
General
The aerosol generator's operating parameters must be fine-tuned to create a test aerosol with a median diameter that aligns with the most penetrating particle size (MPPS) for the specific filter medium For a monodisperse test aerosol, the median diameter should not vary by more than 10% from the MPPS, while a polydisperse test aerosol can have a deviation of up to 50%.
It shall be possible to set the median value of the number distribution of the test aerosol within ± 10 %
The aerosol generator's particle flow rate must be calibrated based on the test volume flow rate and filter efficiency, ensuring that the counting rates on both the upstream and downstream sides remain within the coincidence limits of the counters and are well above the instruments' zero count rate.
The distribution of test aerosol particles can be accurately assessed using a differential mobility particle sizer (DMPS) or a laser particle counter designed for this purpose It is essential that the measurement method employed to determine the median value maintains a limit error of no more than ± 10% relative to the measured value.
To ensure statistically significant results, the number of particles counted both upstream and downstream must be sufficiently large, while also remaining within the measuring range of the upstream particle counter If the concentration of particles upstream surpasses the counter's range in counting mode, a dilution system should be implemented between the sampling point and the counter.
The maximum measurable concentration is constrained by the processing speed of the evaluation electronics in the test apparatus Additionally, measuring uncertainties related to the sample volume flow rate and measurement duration can affect concentration readings It is essential that the particle concentration result, accounting for all error sources at the apparatus interface, does not deviate by more than 10% from the true value.
The particle flow rate must be recorded at specific time intervals (\( \Delta t_i \)) that are at least equal to the time required for the probe to pass through the width of its aperture (\( a_p \)) Additionally, the transmission characteristics of the particle counter and the evaluation electronics must meet these criteria It is essential that the uncertainty in measuring the counting interval duration remains below 10%.
To determine the mean efficiency, the counting signal processing unit must register the total number of particles counted as the probe moves through the passage area and record the total time taken.
Set-up for testing with a monodisperse test aerosol
For technical reasons, the particle size distribution produced by the aerosol generator is usually quasi- monodisperse
When using a monodisperse aerosol for the leakage testing of the filter element either optical particle counters or condensation nucleus counters may be used to determine the particle number concentration
When utilizing a condensation nucleus counter, it is crucial to ensure that the test aerosol does not generate significant quantities of particles smaller than the most penetrating particle size (MPPS) Particles produced by malfunctioning aerosol generators can lead to substantial errors in local efficiency measurements, as they are also counted by the condensation nucleus counter Therefore, it is essential to assess the number distribution of the test aerosol using a method that spans from the lower limit of the condensation nucleus counter up to approximately 1 µm in particle size, ensuring that the resulting distribution is quasi-monodisperse.
Set-up for testing with a polydisperse test aerosol
When testing a filter element for leaks using a polydisperse test aerosol, the particle concentration and size distribution by number shall be determined using an optical particle counter (e.g laser particle counters)
The measuring range of the optical particle counter used in testing efficiency shall cover the following particle sizes (in accordance with Figure 4 of EN 1822-5:2009):
MPPS/1,5 to MPPS x 1,5 (Range I, Figure 4 of EN 1822-5:2009)
The distribution of the size classes shall be such that one class limit meets the following condition:
MPPS/2 < Class limit ≤ MPPS / 1,5 (Range IIa, Figure 4 of EN 1822-5:2009) and a further class limit meets the following condition:
MPPS x 1,5 ≤ Class limit < MPPS x 2 (Range IIb, Figure 4 of EN 1822-5:2009)
The efficiency is assessed for all classes within the specified limits, with no minimum number of classes required in this range Consequently, it is possible for the conditions to be satisfied by just a single size class in extreme cases.
Before mixing with the test aerosol, the test air must be prepared, ensuring its purity with a particle number concentration of less than 350,000 m\(^{-3}\) This can be achieved through appropriate pre-filtering methods, such as using commercially available coarse and fine dust filters, along with high-efficiency particulate air (HEPA) filters.
The temperature and relative humidity of the test air in the duct must be measured on the upstream side and can be adjusted to meet specific requirements through the use of an air heating system.
General
Before beginning the scan test, the test parameters shall be determined or calculated, if this has not already been done for earlier tests, and the appropriate adjustments made
On the basis of the dimensions of the filter and the probe, the parameters for the probe tracking shall be determined These are:
the distance between the probe aperture and the filter element (10 mm to 50 mm; see 6.3.2);
the speed of the probe (to be determined in accordance with B.7);
the number and position of the probe tracks
Test parameters will be established based on the nominal air volume flow rate and the expected penetration of the test filter Key parameters include aerosol concentration on the upstream side, probe volume flow rate, probe speed, and signal value for the counting rate These parameters must be determined following Annex B, along with necessary adjustments to the test apparatus.
Before starting a test with newly established parameters, it is essential to verify the interaction of these parameters and the system's capability to identify leakage limit values Reference filters, which have predefined leakage values, can be utilized for this verification process.
Testing shall not commence until it has been shown that leaks can be detected adequately.
Preparatory checks
After switching on the test apparatus the following parameters shall be checked:
Operational readiness of the measuring instruments
The warming-up times specified by the instrument makers shall be observed and the condensation nucleus counters shall be filled with operating liquid
If the instrument makers recommend further regular checks before taking measurements then these checks shall also be carried out
Zero count rate of the particle counter
The measurement of the zero count rate may be carried out using filtered flushing air
Zero value of the test apparatus
The test shall be carried out using a reference filter with the aerosol generator switched off
If the particle flow rate measured downstream is significantly higher than the long-term zero value of the apparatus, it is essential to address and eliminate the causes before starting the actual test.
Temperature, relative humidity and purity of the test air
These parameters shall be checked to ensure that they comply with the specifications in Clause 7 If this is not the case then appropriate corrections shall be made.
Starting up the aerosol generator
When starting up the aerosol generator, a stand by filter element shall be installed in the test filter mounting assembly in place of the test filter
After modifying the aerosol generator's operating parameters and allowing for a suitable warm-up period, it is essential to verify the particle concentration and size distribution of the test aerosol to ensure compliance with the specifications outlined in section 6.4.
Preparing the test filter
Installing the test filter
To ensure the integrity of the test filter, it must be handled carefully to prevent damage Proper installation is crucial, requiring the filter to be oriented correctly and free from bypass leaks in the mounting assembly.
To accurately identify any leaks after testing, it is essential to document the position of the test filter within the mounting assembly For optimal results, filters should be scanned for leaks while using their original gasket and maintaining the same mounting position and airflow direction as they will have when installed on-site.
Flushing the test filter
To minimize particle emissions from the test filter and ensure temperature uniformity between the test filter and the test air, it is essential to flush the test filter with test air at the nominal volume flow rate for an adequate duration.
The self-emission of the test filter must be assessed through scan testing at the nominal volume flow rate, without adding test aerosol If the particle counting rate downstream is significantly elevated or the mean concentration of downstream air exceeds the zero value, the test filter should be flushed for an extended period before re-evaluating particle emissions.
The testing shall not commence until the particle emissions do not significantly exceed the zero value for the apparatus.
Testing
Measuring the pressure drop
The pressure drop across the test filter must be measured in the unloaded (pre-test) condition at the nominal volume flow rate using pure test air This flow rate should align with the nominal air volume flow rate, maintaining a reproducibility of ± 3% Measurements should only be taken once a stable operating state has been achieved.
Testing with monodisperse test aerosol
In the mixing duct the test air is mixed with test aerosol, the median diameter of which corresponds to the most penetrating particle size (deviation 10 %; see 6.4)
The volume flow rate is calculated based on the contribution from the aerosol generator and is fine-tuned to the nominal flow rate within a ± 3% range Measurements should commence once the system achieves a stable operating condition.
The probe operates based on a tracking program, recording the coordinates where the signal value meets or exceeds a specified threshold on the test filter Additionally, it calculates the total number of particles counted within the passage area and measures the counting duration for this segment of the program.
The aerosol concentration on the upstream side can be measured continuously or intermittently using a dedicated counter or by alternating with the counter for the downstream side It is important to ensure that the testing duration is not excessive to prevent the test filter from becoming overloaded with aerosol.
Testing with polydisperse test aerosol
The test shall be carried out in analogy with 8.5.2, using a polydisperse test aerosol with a median diameter which shall not deviate by more than ± 50 % from the MPPS (see 6.4)
In tests using polydisperse aerosols, both the total number and size distribution must be measured with an optical particle counter To assess efficiency (penetration), upstream and downstream concentrations for all size classes within the range of MPPS/1.5 to MPPS x 1.5 should be utilized.
Leak testing (local penetration)
If the signal value remains within limits during the probe run, the filter is confirmed to be leak-free However, if the signal value is exceeded, it suggests that the local penetration limit may be surpassed at that location To verify local penetration, the probe is repositioned to the coordinates where the signal value was detected during the scan test, aiming to identify the point with the highest count rate At this location, the count rate is measured using a stationary probe, while the aerosol concentration on the upstream side is monitored continuously or intermittently.
The statistical scattering of particle numbers on both the upstream and downstream sides leads to the determination of the maximum local penetration value, as outlined in Clause 9 If this maximum exceeds the limit for the filter class specified in EN 1822-1, the test filter is deemed to have leaks Conversely, if all maximum local penetration values remain below the limit, the filter is classified as leak-free.
A filter may be repaired if necessary and shall then be retested
All repairs, including those performed by the filter manufacturer, must not obstruct more than 0.5% of the filter's face area (excluding the frame), and no single repair should exceed a length of 3.0 cm Alternative repair criteria may be negotiated between the buyer and seller.
Determining the mean efficiency of the filter element
To calculate mean efficiency, the number of particles is counted during the traverse area run, and the total duration of this probe segment is measured The mean particle concentration in the passage area is determined by dividing the particle count by the analyzed air volume, which is the product of the sampling volume flow rate and the sampling duration.
The mean efficiency of the test filter is determined by analyzing the average particle concentration on both the upstream and downstream sides Considering the anticipated statistical variation, a suitable maximum penetration or minimum efficiency will be established, as outlined in Clause 9.
Calculating the penetration and the efficiency
The penetration and the efficiency are calculated from the count data as follows:
P is the penetration (usually as a percentage);
E is the efficiency (usually as a percentage);
N u is the number of particles counted upstream;
N d is the number of particles counted downstream; k D is the dilution factor; c N,u is the number concentration upstream; c N,d is the number concentration downstream;
V & s,u is the sampling volume flow rate upstream;
The sampling volume flow rate downstream is denoted as V & s,d, while the sampling duration upstream and downstream are represented by t u and t d, respectively Calculations must consider the particle counting statistics outlined in Clause 7 of EN 1822-2:2009 For a 95% confidence level, values should be derived solely from pure counting data, without any adjustments for the dilution factor.
P 95%max is the maximum penetration taking into account the particle counting statistics (usually given as a percentage);
95%min is the minimum efficiency taking into account the particle counting statistics (usually given as a percentage);
N u,95%min is the lower limit of the 95 % confidence level of the particle count upstream;
The parameter \$N_{d,95\%max}\$ represents the upper limit of the 95% confidence level for particle counts downstream, while \$c_{N,d,95\%max}\$ indicates the maximum concentration of particle numbers downstream Additionally, \$c_{N,u,95\%min}\$ denotes the minimum concentration of particle numbers upstream.
If the manufacturer's instructions for the particle counter include coincidence corrections for the measured concentrations, then these shall be taken into account in the evaluation
To determine the minimum efficiency, it is essential to consider only the measurement uncertainty arising from low count rates Any additional errors in the measurement must be corrected if they are known.
In order to calculate local penetration values, measurements obtained from scan testing for leaks in accordance with 8.5.4 shall be inserted in the formulae in 9.1
Values for local penetration shall be designated with the coordinates of the position on the downstream filter face detected at which the leakage signal was tested
For the calculation of the mean efficiency or mean penetration, the mean particle concentration downstream of the test filter determined according to Clause 8 shall be used as downstream concentration.
Classification
The minimum efficiency or the maximum penetration is the basis of the classification in accordance with
EN 1822-1 The limit values shall not be exceeded either for the integral value or for the local values
The test report for the leak test of the filter element shall at least contain the following information: a) Test object:
1) Type designation, part number and serial number of the filter;
2) Overall dimensions of the filter;
3) Installation position of the filter (gasket upstream or downstream); b) Test parameters:
1) Temperature and relative humidity of the test air;
2) Nominal air volume flow rate and test air volume flow rate of filter;
3) Most penetrating particle size of filter media (MPPS) at corresponding medium velocity (see
4) Aerosol generator (type designation and part number);
5) Test aerosol (substance, median diameter, geometrical standard deviation);
NOTE In case a solid aerosol (e.g PSL) is used, requirements of EN 1822-5:2009, A.5 have to be met
6) Particle counter(s), upstream and downstream (type designation and part number(s)) and particle size channel(s) used (in case of OPC);
7) Dilution system for upstream particle counter (type designation and part number);
8) Sampling probe downstream side (geometry, sampling air flow);
9) Reference leak penetration and signal value setting (relevant limit value indicating a leak); c) Test results:
1) Mean differential pressure across the filter at test air volume flow;
2) Mean upstream and downstream particle concentration;
3) Confirmation of freedom from leaks (mentioning reference leak penetration);
4) Mean integral efficiency and minimum integral efficiency E 95 %min (in case of combined local and integral efficiency test);
5) Filter class in accordance with EN 1822-1
11 Maintenance and inspection of the test apparatus
Regular maintenance, inspection, and calibration of all components and measuring instruments in the test apparatus are essential As outlined in Table 1, these tasks must be performed at least once within the specified timeframes Additionally, any disturbances requiring maintenance, as well as inspections and calibration following significant alterations or refurbishments, should be conducted immediately.
EN 1822-2 outlines the maintenance and inspection procedures for test apparatus, including the calibration of all components and measuring instruments These maintenance activities are crucial for ensuring that the measurement deviations of the equipment remain within the permitted limit values.
According to EN 1822-2, the maximum limit errors for measuring equipment are applicable at the interface of the measuring chain in the test apparatus, which affects the recorded results To prevent unacceptable measurement deviations between testing sessions, it is essential to use reference filters, which should be replaced periodically to avoid aerosol loading effects Test results obtained with these reference filters must be documented Additionally, corrective measures should be implemented if the penetration result deviates by more than 30% or the pressure drop result deviates by more than 10% from the arithmetic means of comparative tests.
The maintenance, inspection, and calibration intervals are affected by the specific characteristics and operation of the test rig, which should be considered when determining or reviewing these intervals.
Table 1 — Maintenance and inspection intervals for components of the test apparatus
Component Type and frequency of maintenance/inspection
Test air preparation system; test air duct entire system test air filter
– Annually, or – When maximum pressure drop is reached, or – In the event of leaks
Lines taking aerosol to the measuring instruments Cleaning annually or before every change of the aerosol substance
Volume flow rate meter Annually
Repeatability of the adjustment of the test volume flow rate with reference resistances
Ensuring the air-tightness of apparatus components at low pressure is crucial, particularly if the zero count rate of the particle counter is inadequate It is essential to conduct annual checks on the air-tightness of both the apparatus parts and the pressure measurement lines to maintain optimal performance.
Air-tightness of the aerosol transport lines Annually
Measuring equipment for the volume flow rates in the probe
Particle concentration profile over the passage area Annually
Aerosol transport losses on the upstream and downstream sides
Coordinate measurement of the scanning system Annually
Probe speed of the scanning system Annually
Checking the apparatus with reference filters Annually
The leak test is essential for ensuring that filter elements do not exceed permissible local penetration values, as outlined in EN 1822-1:2009, Table 1 For filters in group H (classes H13 and H14), the Oil Thread Leak Test serves as an alternative method This test is based on the particle count scan method specified in the standard Additionally, the Oil Thread Leak Test is suitable for filter shapes where the scan method is not applicable, such as V-bank media panels or cylindrical filters.
The Oil Thread Leak Test is a qualitative method that visually demonstrates the absence of leaks Regular training for test personnel is crucial, along with periodic verification of the procedure's sensitivity using reference filter elements that have well-defined leaks, as characterized by the reference scan test method The local penetration of leaks in these reference filter elements must fall within the limit values specified for the filter class in EN 1822-1:2009, Table 1, and should not exceed double the corresponding limit value.
The filter will be tested with a polydisperse oil-drop aerosol flow at approximately 1.3 cm/s (42 m³/m²/h), adjustable for optimization It must be positioned horizontally on a diffuser or box, with a mounting assembly that ensures a proper seal and compliance with flow requirements, while not obstructing any part of the filter's cross-sectional area.
The polydisperse test aerosol is produced by nebulizing a liquid aerosol substance, following the guidelines of EN 1822-2:2009 The median particle diameter must range from 0.3 µm to 1.0 µm, with a mass concentration of 1.5 g/m³, as determined by gravimetric methods.
The filter's downstream side must be illuminated from above with a white fluorescent or halogen lamp, ensuring a color temperature of at least 4,000 K and a brightness exceeding 1,000 Lux at the working plane Additionally, the area surrounding the filter should be darkened, with a black observational background to enhance visibility It is also essential to eliminate any uncontrolled air currents from the environment.
Leaks can be identified by the presence of a visible oil thread resulting from the leakage If no oil threads are observed, filters up to class H14 are considered leak-free according to the leak limit values specified in EN 1822-1:2009, Table 1.
The lamp's position and brightness can be adjusted based on the examiner's perception by utilizing reference filter elements with clearly defined leaks, as determined by the scan test method It is advisable to employ reference filters with well-defined leaks both in the medium and at the frame corners, particularly near the sealant.
A test report for the oil thread test shall contain at least:
details of the filter tested (type, dimensions, identification number, nominal technical data);
details of the test parameters (flow velocity, test aerosol, median particle diameter and mass concentration of test aerosol);
identification of tester and date of test; and
the test result (confirmation of absence of leaks)
On the test report it shall be clearly stated that the filter was tested using the test method as per EN 1822-
General
Before starting the test, the parameters must be calculated based on the defined boundary conditions and the test filter data If the calculations yield parameters that are unattainable, an iterative process will be employed to adjust the input data as needed.
All particle numbers and concentrations mentioned pertain to the size range of the monodisperse test aerosol or the size range utilized for assessing filter efficiency with a polydisperse aerosol (refer to section 8.5.3).
Boundary conditions
The following boundary conditions shall be complied with:
Probe aperture cross section: A p = (9 ± 1) cm 2
Minimum particle number for a leak signal:
(lower limit of the 95 % confidence interval) N min, 95 % = 5
Value for particle number to be expected traversing a leak: N min, leak = 10
Minimum particle number on the downstream side for determining the efficiency: N min, abs = 100
Test filter data
The following data for the test filter are to be taken into consideration when determining the test parameters:
The filter class which is to be established, characterized by the limit values of the penetration:
Data for the apparatus
Particle counters
The following data is relevant for the particle counters employed:
Number of counters operating in parallel M
The zero count rate for the entire system on the downstream side must be established, rather than just the zero count rate of the counter (refer to EN 1822-2) This rate is measured with the test filter installed and the aerosol generator turned off, accounting for impurities in the test air and potential particle release from the measuring lines.
The minimum counting (particle flow) rate of the counter on the downstream side is determined from the zero count rate of the apparatus as follows:
N & min, is the minimum counting rate of the downstream particle counter;
N & zero is the zero count rate of the system on the downstream side.
Downstream sampling probes
Probes can feature either a circular or rectangular cross-section, with the selected diameter or side lengths determining the required cross-sectional area For rectangular probes, the ratio of side lengths must not exceed 15 to 1.
Using circular cross-section probes presents several challenges, particularly in leak detection The time required to traverse a leak varies based on its location relative to the probe, making consistent leak detection difficult without overlapping passage runs Typically, a 20% overlap of the probe diameter is necessary to achieve reliable results for aP.
The following considerations refer to a probe with a rectangular cross-section The calculations may, however, be applied by analogy for use with circular probes
internal side-length in scan direction a p ;
internal side-length at right-angles to scan direction b p
Loss factor
The minimum counting rate for a leak, as outlined in section B.2, must be met even if the leak occurs at the edge of the probe's coverage area Consequently, it is anticipated that the average counting rate for a leak located at the center of the path will be greater.
N min is the minimum counting rate for a leak in middle of the probe;
N min ,leak is the expected minimum particle number for a leak; k b is the loss factor for a leak at the edge of the probe path
For probe paths that touch without overlapping, the loss factor is set at \( k_b = 0.5 \), resulting in a minimum leak detection rate of \( N_{min} = 20 \) However, when overlapping occurs, the loss factor can be increased In uncertain situations, it is recommended to experimentally determine the loss factor using a stationary probe.
Sequence of calculation steps
Figure B.1 illustrates the flow diagram for calculating test parameters It indicates that if the parameters do not meet the required standards or if the signal difference is inadequate (refer to B.10.2), the initial parameters must be adjusted until the results permit the test to proceed.
Figure B.1 — Flow diagram of determination of test parameters
Checking the isokinetic sampling
The mean air speed in the probe is calculated as follows from the volume flow rate in the probe and its cross-sectional area:
EN 1822-4:2009 (E) w p is the mean air speed in the probe;
V & p is the volume flow rate in the probe;
A p is the probe intake cross-section
The calculated value of w p shall be compared with the mean air speed w d for the passage area downstream The deviation between the two speeds shall not exceed 25 % (see 6.3.2)
If the volume flow rate of the probe is variable, then the speed in the probe can be adjusted to the speed in the passage area.
Choosing the probe speed
Any traversing speed can be chosen for the probe up to the limit value of 10 cm/s
The time required for the probe to traverse a leak can be determined using the formula: \$t_{leak} = \frac{a_p}{u_p}\$, where \$t_{leak}\$ represents the duration spent crossing the leak, \$a_p\$ denotes the width of the probe aperture in the scanning direction, and \$u_p\$ indicates the speed of the probe.
It is also possible to determine the total scanning time t p,tot during the scan test
The counting rate must be measured at intervals that align with the time required for the probe to move across its aperture width, denoted as \( a_p \) Additionally, the particle counter's transmission characteristics and the evaluation electronics must meet these specifications It is essential that the uncertainty in the duration of the counting interval remains below 10%.
In the event of a leak occurring at the leading edge of the probe during the initial counting interval, all particles passing through the leak will be recorded Conversely, if the leak is located in the middle of the probe's path during the time interval, the counts associated with the leak will be distributed across two counting intervals To ensure accurate evaluation, it is recommended to combine the two adjacent counting intervals.
To effectively localize leaks, it is essential to understand the delay time of the aerosol within the transport line This delay time, denoted as \( t_{del} \), is influenced by the probe speed \( u_p \) and the aperture width \( a_p \) in the direction of movement, following the relationship \( u \cdot t_{app} \leq t_{del} \).
Minimum aerosol concentration
The minimum aerosol concentration is the maximum value permitted by the four boundary conditions or limiting parameters specified below
The minimum aerosol concentration for identifying limit leakages shall satisfy the condition: t V p c N
& s l leak class, min min u, ≥ × × (B.6) where c u,min is the minimum aerosol concentration for the identification of limit leaks;
N min is the minimum counting rate for a leak in middle of the probe;
P class,l is the limit value for the local penetration of the filter class; tl eak is the time spent by the probe above a leak;
V & S is the sampling volume flow rate
The minimum aerosol concentration necessary to ensure the required minimum counting rate in the downstream particle counters shall satisfy the condition:
& s c min, i eff, min u, ≥ × × (B.7) where c u,min is the minimum aerosol concentration for the particle counter on the downstream side;
P eff ,i is the effective value of the integral penetration;
N & min,c is the minimum counting rate for the particle counter;
V & s is the sampling volume flow rate
The effective value of the test filter penetration, denoted as \$P_{\text{eff},i\$, may be significantly lower than the local penetration limit, \$P_{\text{class},i\$ For accurate calculations, it is essential to use the effective value If the effective value is not available from previous measurements, it should be either estimated or determined through measurement.
Further boundary conditions for the minimum aerosol concentration needed for the determination of the
N P c l tot s p, abs min, i eff, min u, ≥ × ×
& (B.8) where c u,min is the minimum aerosol concentration to reach N min,abs on the downstream side;
P eff,i is the effective value of the integral penetration;
N min,abs is 100 (= min particle number (see B.2));
V & S is the sampling volume flow rate; t p,tot is the total path time of probe
For the upstream particle counter the condition is: t l V k N c u s p, abs min, D min u, ≥ × ×
& (B.9) where c u,min is the minimum aerosol concentration to reach N min,abson the upstream side; k D is the dilution factor, upstream;
N min , abs is 100 (= min particle number (see B.2));
V & S is the sampling volume flow rate; t p,u is the duration of sampling on the upstream side.
Maximum aerosol concentration
There are three boundary conditions for the maximum aerosol concentration, and these also have to be examined individually In this case the lowest resultant concentration gives the maximum concentration
In order to avoid an alteration of the size distribution of the test aerosol due to coagulation, the following maximum concentration shall not be exceeded:
, ≤ cu (B.10) where c u,max is the maximum aerosol concentration to avoid aerosol losses
The maximum concentration measurable by the particle counters provides the other two boundary
For the counters on the downstream side the condition is:
P c c l max, c max, max u, ≤ (B.11) where c u,max is the maximum aerosol concentration for the downstream counter; c max , c is the maximum concentration measurable with the particle counter on the downstream side;
The maximum measurable local penetration, denoted as \$P_{max,l}\$, must be specified and is greater than or equal to \$P_{class,l}\$ For the upstream counter, the relationship is defined by the inequality \$k_{c} \cdot c_{u,max} \leq max_{c} \times D\$, where \$c_{u,max}\$ represents the maximum aerosol concentration for the upstream counter, \$c_{max,c}\$ is the maximum concentration measurable with the upstream particle counter, and \$k_{D}\$ is the dilution factor on the upstream side.
Leak signal
Effective value
The minimum expected value for the counting rate when the probe crosses a leak in the middle of the probe path is given by:
N min,em= u× class, l× &s× leak (B.13) where
N min,em is the expected minimum particle number for a leak in the middle of the probe path; c u is the measured number concentration on the upstream side;
P class,l is the class limit value for the local penetration;
V & S is the sampling volume flow rate; t leak is the time spent by the probe over the leak
For a leak at the edge of the path, the equation is:
N min, eb = min, em × b (B.14) where
N min,eb is the expected minimum particle number for a leak at the edge of the probe path;
N min,em is the expected minimum particle number for a leak in the middle of the path; k b is the loss factor for a leak at the edge of the probe path
The statistical minimum value for the 95 % confidence level of N min,eb is determined in accordance with
EN 1822-2, and designated N min,eb,95 % When this value is reached the apparatus shall report a leak (leak signal value).
Signal difference
The signal difference is defined as the discrepancy between the leak signal value and the signal produced by the particle flow rate in a section of the filter that is unaffected by leaks.
The mean expected value for the particle number in a probe passing through a filter section is determined by the penetration that aligns precisely with the class limit value.
N em is the mean expected value of the particle number; c u is the number concentration on the upstream side of the test filter;
P class,i is the limit integral penetration value;
V & S is the sampling volume flow rate; t leak is the time spent by the probe over the leak
The statistical maximum value for the 95 % confidence level of N em is determined in accordance with EN 1822-2, and designated N em,95 % 1)
The signal difference is then defined as:
1) Since the counting rate calculated from the particle concentration is the actual expected value, the so-called error band should
N min,eb,95 % is the lower limit value of the 95 % confidence level for the minimum expected counting rate when passing over a leak at the edge of the probe path;
The upper limit of the 95% confidence level for the expected counting rate is 95% when traversing a section of a filter that is free from leaks and has a penetration value precisely at the class limit.
A positive value for S indicates a sufficient signal difference, while a negative value suggests a higher likelihood of false leak signals during scan tests.
Example of an application with evaluation
Typical test parameters for a filter of class H14 are summarized in Table C.1
Table C.1 — Typical test parameters for a filter of class H14
Limit value for the integral penetration
Limit value for the local penetration
Particle concentrations: upstream downstream, integral downstream, local c u 1,73 x 10 4 cm -3
Volume flow rate in the probe
Mean air speed in the probe
Probe time spent above site of leak
Expected particle number per time interval ∆∆∆∆ t i: without leak with leak with leak; loss factor k b = 0,7
Limit value from Poisson statistics:
Max particle number without leak
Min particle number with leak
The relationship between the individual test parameters and the determination of signal value and signal difference is presented graphically in Figure C.1
Figure C.1 — Determining the signal value and the signal difference from the test parameters for a filter of class H14
In Table C.2 the most important test parameters for filters of the classes H13 up to class U17 are compared
Table C.2 — Examples of important test parameters for the filter classes H13 to U17
Term Symbol Unit Filter class
Limit value for the integral penetration
Limit value for the local penetration
Expected particle number: without leak with leak with leak; k b = 0.7 a
Max particle number without leak
Min particle number with leak
Min aerosol concentration c u,min cm -3 1,55 x 10² 1,98 x 10² 1,98 x 10³ 8,48 x 10³ 8,48 x 10 4
Max aerosol concentration c u,max cm -3 5,30 x 10³ 2,12 x 10 4 2,12 x 10 5 4,55 x 10 5 4,55 x 10 5 a For leak at the edge of the probe path (kb = 0,7)
Leak Test with solid PSL Aerosol
Background
In the semiconductor and space industries, the presence of a liquid oil-like substance poses a significant risk, making it unsuitable for testing HEPA and ULPA filters used in Cleanrooms During testing, these liquid particles can accumulate in the filter and may outgas during operation, potentially disrupting the production process Additionally, using liquid particles in leak tests for filters with PTFE-membrane media is inappropriate due to the unique properties of this filter material.
All standardized methods for leak and efficiency testing, as well as classification to EN 1822, rely on liquid particles such as DEHS, PAO, and paraffin oil as test aerosols The use of liquid particles like DEHS is advantageous due to its ease of use and reproducibility of results The choice of test aerosol significantly impacts every aspect of EN 1822, including instruments, test rigs, statistics, test results, and classification Consequently, substituting liquid test aerosols with solid alternatives can lead to substantial changes in test outcomes and filter classification.
Annex D introduces an alternative leak test and classification method for filters tested with solid particles, specifically utilizing a scanning method with solid PSL aerosol Despite this alternative approach, the efficiency determination and classification continue to follow the established EN 1822-1 guidelines, which employ the reference test method using liquid DEHS aerosol.
General Remarks
When using a solid test aerosol like PSL for scanning, the efficiency derived from average upstream and downstream particle concentrations cannot be used to classify the filter per EN 1822-1 This integral efficiency value will differ from that obtained with the liquid DEHS reference aerosol because of electrostatic effects.
The solid testing aerosol scanning procedure is exclusively employed to verify the absence of leaks in filters This method aligns with the boundary values specified for maximum leak penetration according to EN 1822-1:2009, Table 1, which is applicable to each filter class.
To classify filters, a representative sample from the same production batch undergoes an efficiency test according to EN 1822-5 using DEHS aerosol These tested filters serve as a reference for efficiency and classification per EN 1822-1 for the entire batch Subsequently, all other filters are subjected to PSL leak testing as outlined in Annex D It is crucial that the specifications and test data, including filter size, design, and test air flow, of both the DEHS tested reference filters and the PSL tested filters are identical.
Test Procedure
For the PSL leak test outlined in this annex, the test equipment and procedures specified for DEHS aerosol in EN 1822-4 can be utilized However, it is important to note that the aerosol generator type and usage must differ due to the specific requirements of the PSL aerosol The primary objective is to ensure adequate testing results.
EN 1822-4:2009 (E) concentration levels for PSL particles in the upstream air which, in case of PSL particles, needs special generating equipment
Presently, there is only one high output PSL aerosol generator commercially available from:
MSP Corporation, Shoreview, MN 55126, USA (www.mspcorp.com), PSL generator model No 2045 2)
Figures D.1 and D.2 show an example of a specific design of a high output PSL particle generator, operating with PSL-water emulsion, with spray nozzles and with a corresponding drying section
Figure D.2 — PSL Generator Design Design description
Nozzle 1 utilizes clean compressed air at pressure P to spray an aqueous solution of PSL particles mixed with clean water into a chamber This chamber is heated with HEPA filtered hot air at temperature T1, facilitating rapid distribution and evaporation of the water The heated air, generated by an adjustable heater and a fan with an airflow rate of 40 m³/h to 50 m³/h, then moves through a cooling/condensation section where the temperature drops to T2 and relative humidity RH A water trap container effectively collects excess water from the cooling section, minimizing the risk of water entering the test system.
P = 1 bar to 5 bar q = 5 ml/min to 25 ml/min
T2 = 20 °C to 23 °C (preferably at or below test air temperature)
Test Protocol
The test protocol, in addition to the requirements mentioned under EN 1822-4:2009, Clause 10, shall contain the following additional information:
Statement, that the filter was leak tested using the test method as per EN 1822-4:2009, Annex D and efficiency tested on statistical bases;
Test aerosol used (e.g solid PSL);
Statement that the average PSL particle concentrations cannot be used for the classification of the filter
0,3 àm – 0,5 àm Particle Efficiency Leak Test
The "Oil Thread Leak Test" (Annex A) is a visual assessment, leading to potential variations in leak detection results among different operators or throughout an operator's shift In contrast, the "0.3 µm – 0.5 µm Particle Efficiency Leak Test" (Annex E) aims to automatically identify leaks by measuring efficiency within the particle size range of 0.3 µm to 0.5 µm.
This method of measuring efficiency utilizes a particle counter in the 0.3 µm – 0.5 µm particle size channel to assess H13 class filters for leaks, serving as an alternative to the oil thread test (Annex A) The 0.3 µm – 0.5 µm Particle Efficiency Leak Test can be employed as a reference procedure for H13 class filters with turbulent airflow that cannot undergo scan testing due to their construction type, such as V-bank or cylindrical filters.
Based on both practical experience and theoretical calculations involving a defined leak, it is established that a class H13 filter, which has a local MPPS efficiency of 99.75%, must achieve a minimum global efficiency exceeding 99.9996% at particle sizes of 0.3 µm to 0.5 µm.
According to EN 1822-1, filters are tested on a bench to measure their integral MPPS efficiency, as outlined in EN 1822-5 Simultaneously, the efficiency test for particles sized 0.3 µm to 0.5 µm can be conducted using the appropriate channel of the particle counter For accurate results, it is crucial to ensure proper aerosol distribution upstream of the filter and effective air mixing downstream.
For accurate measurements of the 0.3 µm – 0.5 µm Particle Efficiency Leak Test, it is crucial to use a polydisperse aerosol, as a monodisperse aerosol is inadequate To ensure reliable results, more than ten particles within the 0.3 µm – 0.5 µm size range must be sampled downstream of the filter Consequently, a minimum of 2,500,000 particles in the 0.3 µm – 0.5 µm range must be present upstream of the filter during the sampling time interval.
For the filter class H13 (integral MPPS efficiency > 99,95 %, local MPPS efficiency > 99,75 %), the efficiency at 0,3 àm – 0,5 àm must be > 99,999 6 %
Regular verification of the sensitivity and accuracy of the procedure is essential, ensuring that the penetration of leaks does not exceed twice the limit value specified for the H13 filter class To confirm adequate upstream aerosol distribution and effective downstream air mixing, reference filters with defined leaks should be tested at regular intervals, particularly in the frame corner and near the frame/sealant These filters, ideally square-shaped for 90° rotation and repeated measurements, can be characterized using the oil thread leak test, with leak levels again not surpassing the H13 filter class limit by a factor of more than two Proper aerosol distribution and downstream mixing are crucial for identifying leaks in these filters according to established criteria.
When conducting a leak test on an H13 filter using the “0.3 µm – 0.5 µm Particle Efficiency Leak Test,” it is essential to document this on the filter and in the test report, including a note such as “leak tested as per EN 1822-4:2009, Annex E.” Additionally, the test report must include the actual measured efficiency at the specified particle sizes of 0.3 µm – 0.5 µm.
[1] Wepfer, R (1995): "Characterisation of HEPA and ULPA filters by proposed new European test methods", Filtration & Separation, vol 32, n° 6, pp 545-550
[2] EN ISO 5167-1, Measurement of fluid flow by means of pressure differential devices inserted in circular cross-section conduits running full — Part 1: General principles and requirements (ISO 5167-1:2003)