Table 1 — Optical particle diameter size rang es for the definition of the ef iciencies, e PM Air filt ers for g ene al ventiation ar widely used in heating, ventiation an air-conditioni
General
The filter element shall be designed or marked for air flow direction in a way that prevents incorrect mounting.
The filter must be engineered to prevent leaks at the sealing edge when properly installed in the ventilation duct If standard testing conditions cannot be met due to dimensional constraints, it is permissible to assemble multiple filters of the same type or model, ensuring that no leaks are present in the final configuration.
Material
The filter element shall be made of suitable material to withstand normal usage and exposures to those temperatures, humidities and corrosive environments that are likely to be encountered.
The filter element shall be designed to withstand mechanical constraints that are likely to occur during normal use.
Nominal air flow rate
The filter element shall be tested at its nominal air flow rate for which the filter has been designed by the manufacturer.
Many national and association bodies classify or rate air filters with a nominal face area of 610 mm × 610 mm (24 inch × 24 inch) using an air flow rate of 0.944 m³/s (2,000 ft³/min or 3,400 m³/h) If a manufacturer does not specify a nominal air flow rate, the filter must be tested at this standard rate, which corresponds to an air flow velocity of 2.54 m/s (500 ft/min).
Resistance to air flow
The resistance to air flow (pressure differential) across the filter element is recorded at the test air flow rate as described in detail in ISO 16890-2.
Fractional efficiency curves (particle size efficiency spectrum)
The initial fractional efficiency curve, denoted as E i, for the unloaded and unconditioned filter element is determined based on particle size at the specified test air flow rate, following the guidelines of ISO 16890-2.
The fractional efficiency curve, E D ,i, of the filter element after an artificial conditioning step defined in ISO 16890-4 is determined as a function of the particle size in accordance with ISO 16890-2.
Arrestance
The initial arrestance, which measures the resistance to airflow in relation to the mass of test dust captured, along with the test dust capacity, is evaluated according to ISO 16890-3 using L2 test dust as outlined in ISO 15957.
The technical specifications and test conditions for the test rig(s) are detailed in ISO 16890-2, ISO 16890-3, and ISO 16890-4 The testing process involves several key steps: measuring air flow resistance as per ISO 16890-2, determining the initial fractional efficiency curve of the unconditioned filter, performing an artificial conditioning step according to ISO 16890-4, and measuring the fractional efficiency curve of the conditioned filter Additionally, ePM efficiencies are calculated as defined in Clause 7, and the filter may be loaded with synthetic L2 test dust to assess initial arrestance and dust capacity, although this step is optional for certain filter groups.
The initial fractional efficiency curve, Ei, of the untreated and unloaded filter element, along with the fractional efficiency curves, ED,i, obtained after an artificial conditioning step, are utilized to compute the average fractional efficiency curve, EA,i, as outlined in Formula (1).
NOTE For further explanations on the test procedure according to ISO 16890-4, please refer to 8.2.
ISO 16890-4 outlines a procedure that quantitatively assesses the impact of electrostatic charge on the initial performance of filter elements without dust load It reveals the efficiency achievable when the charge effect is entirely eliminated, without any compensatory increase in mechanical efficiency Consequently, the fractional efficiencies, ED,i, following an artificial conditioning step may underestimate actual efficiencies in real-world conditions, which are influenced by various uncontrolled parameters The true minimum fractional efficiencies during service are likely to fall unpredictably between the initial and conditioned values To estimate real fractional efficiencies, ISO 16890 suggests using the average of these two values, as indicated in Formula (1) It is important to recognize that fractional efficiencies observed in actual service may significantly differ from those specified in ISO 16890 Furthermore, the chemical treatment of filter media, as described in ISO 16890-4, can alter the fiber matrix structure or chemically damage the fibers, potentially compromising the filter medium Therefore, not all filter types and media are suitable for the mandatory procedures outlined in ISO 16890-4, and such filters cannot be classified under this standard.
7 Classification system based on particulate matter e f ficiency ( e PM)
Definition of a standardized particles size distribution of ambient air
Air filters are evaluated based on their ePM efficiencies using standardized particle size distribution functions that reflect the average ambient air in urban and rural areas In the relevant size range (>0.3 µm), ambient air particles exhibit a bimodal distribution, consisting of fine and coarse modes Fine filters, primarily aimed at capturing PM1 and PM2.5 particles, are assessed using a size distribution typical of urban environments, whereas filters designed for PM10 are evaluated with a distribution representative of rural settings.
The particle size distribution of ambient air is influenced by various factors, including location, seasonal changes, and weather conditions As a result, the actual measured particle size distribution can vary significantly from the standardized values outlined in ISO 16890.
This bimodal distribution is represented by combining lognormal distributions for the coarse and the fine mode as given in Formula (3). f d d d d
In Formula (2), the lognormal distribution function, denoted as \( f_d(d, \sigma_g, d_{50}) \), is utilized to represent particle size distribution for either coarse or fine modes, where \( d \) is the variable particle size, and the scaling parameters include the standard deviation \( \sigma_g \) and the median particle size \( d_{50} \) The bimodal distribution, as expressed in Formula (3), is obtained by combining the lognormal distributions for the coarse (B) and fine (A) modes, weighted by the mixing ratio \( y \).
= d = ⋅ + − ⋅ d ln σ 50 1 σ 50 (3) where the parameters are defined to the values given in Table 2, representing urban and rural areas.
Table 2 — Parameters f or the distribution f unction as given in Formula (3) f or urban and rural environments urban q 3u ( )d i A B rural q 3r ( )d i A B d 50 , u 0,3 μm 10 μm d 50 , r 0,25 μm 11 μm σ g u , 2,2 3,1 σ g , r 2,2 4 yu 0,45 yr 0,18
Figure 1 shows a graphical plot of Formula (3) using the parameters given in Table 2.
Key logarithmic distribution (this part of ISO 16890) logarithmic distribution (cumulative)
Figure 1 — Discrete and cumulative logarithmic particle volume distribution f unctions o f ambient aerosol as typically f ound in urban and rural environments (see Re f erence [7])
As an example, Table 3 gives the values of the standardized proportion by volume, q 3, calculated using Formula (3) for the particle counter channels recommended by ISO 16890-2.
Table 3 — Example o f the standardized urban and rural particle volume distributions, q 3, in ambient air f or the particle size channels recommended by ISO 16890-2
Optical particle diameter in àm Discrete particle volume distribution di di+1 d i = d d i ⋅ i + 1 ∆lnd i =ln(d i + 1 / )d i urban q 3u ( )d i rural q 3r ( )d i
In this section of ISO 16890, the distinctions between aerodynamic and optical particle diameters are overlooked It is also assumed that particle density remains constant, although in real ambient air, this density can vary with particle size.
Calculation of the particulate matter efficiencies ( e PM)
The efficiencies of particulate matter, specifically ePM10, ePM2.5, and ePM1, are determined using the average fractional efficiencies \(E_{A,i}\) as outlined in Formula (1) and the standardized particle size distribution defined in section 7.1, as referenced in Formula (3) These calculations are performed through the application of Formula (4).
(rural size distribution) (4) where d i = d d i ⋅ i + 1 is the geometric mean diameter and ∆lnd i = lnd i + 1 −lnd i =ln(d i + 1 / )d i
In Formula (4), the variable \(i\) represents the channel number corresponding to the particle size range being analyzed, while \(n\) denotes the channel that encompasses the specific particle size \(x\) (where \(d_n < x \leq d_{n+1}\)) For instance, \(x\) is set at 10 µm for ePM 10, 2.5 µm for ePM 2.5, and 1 µm for ePM 1 To accurately determine the efficiency of ePM 1, the upper limit of the largest channel included in Formula (4) must be established.
The efficiency of ePM10 is determined by ensuring that the upper limit of the largest channel in Formula (4) is set to 1 àm (where dn+1 = 1 àm), while for ePM2.5, it must not exceed 3.0 àm (with dn+1 ≤ 3.0 àm).
The particle counter's efficiency values are calculated with a lower size limit of 0.3 µm for the smallest channel (d₁ = 0.3 µm) For accurate calculations, a minimum of three channels is required for ePM₁ (n ≥ 3), six channels for ePM₂.₅ (n ≥ 6), and nine channels for ePM₁₀ (n ≥ 9) It is essential that all channels utilized are adjacent, ensuring no particle sizes are missed or overlapped in the measurements.
Additionally, the minimum efficiencies, ePM2,5, min and ePM1, min are defined by Formula (5). e E q d d q d d n i n
Classification
The classification of a filter into one of the four groups outlined in Table 4 is determined by its initial arrestance and the efficiency values ePM 1, ePM 2.5, and ePM 10, along with the minimum efficiency values ePM 1, min and ePM 2.5, min.
Group designation Requirement Class reporting value ePM1, min ePM2,5, min ePM10
ISO Coarse — —