Investigation of Air Humidity Affecting Filtration Efficiency and Pressure Drop of Vehicle Cabin Air Filters

Vehicle cabin air filters are exposed to humid air more frequently than any other air filters during routine use. The filtration performance of several commercially-available cabin air filters was investigated, along with the humid exposure period, using laboratory-based measurements. The averaged filtration efficiency and pressure drop were measured at ~70% and 75 Pa, respectively. Significant increases in filtration efficiency (up to 15%) and pressure drop (up to 250 Pa) were observed as the humid exposure time increased. Filtration efficiency increased ~15% and pressure drop increased 250 Pa as 140 g water was absorbed, which represents ~60 minutes humid exposure at a relative humidity of 90%. The pressure drop increased significantly at the beginning of the humid exposure due to the greater water absorption capacity of dryer dust in the filter. The dust load had a significant effect on the changes in filtration efficiency and pressure drop. The filtration efficiency and pressure drop of the 12-month used filter increased 2 times faster than that of the new filter at the same exposure conditions. The filtration efficiency and pressure drop were explicitly expressed as functions of the water absorption mass in the filter. Two coefficients were empirically derived and successfully accounted for the effects of humid exposure on filtration efficiency and pressure drop.


INTRODUCTION
In the last decade, many studies have identified vehicle cabin as a microenvironment for human exposure to Particulate Matters (PM) (Zhu et al., 2007;Kaminsky et al., 2009;Geiss et al., 2010;Knibbs and de Dear, 2010;Bigazzi and Figliozzi, 2012;Hudda et al., 2012).Extensive studies have been conducted to investigate the parameters that determine the PM entry to the in-cabins and measures to reduce the in-cabin PM concentrations (Xu and Zhu, 2009;Knibbs et al., 2010;Hudda et al., 2011).Both experimental and numerical studies have reported a positive effect of cabin air filter on reducing the in-cabin PM concentration (Pui et al., 2008;Qi et al., 2008;Xu et al., 2011;Xu and Zhu, 2013).An appropriate cabin filter under good condition leads up to a 40% reduction of in-cabin PM exposure (Xu et al., 2011;Xu and Zhu, 2013).During the vehicles' routine use, vehicle cabin filter are exposed to humid air more frequently than any other air filters.Vapor and droplets in the humid air may be absorbed by the dust loaded in the cabin filter, which negatively affects the filters' performance, leads to the malfunction of cabin filters or even the failure of vehicle ventilation system.Thus, it is essential to investigate the influence of the humid air on the cabin filters' performance along the filter usage periods.
Previously, several studies have been conducted to assess the performance of cabin air filters.Pui et al. (2008) found that the in-cabin PM concentration was reduced significantly with the recirculated air passing through the cabin filter.Qi et al. (2008) experimentally evaluated the vehicle cabin air filters' filtration efficiency that varied from 20% to 70%.Xu et al. (2011) investigated the cabin filters' performance under different air velocities and dust loadings.Up to 10% increase of filtration efficiency and 45 Pa increase of pressure drop were observed as the filter was used for 20 months.Xu et al. (2013) conducted extensive measurements on the performance of air filters that were used in the airliner cabins and reported a much greater filtration efficiency (86%-99%) and pressure drop (150-250 Pa) than vehicle cabin filters'.However, the performance of cabin air filters under humid exposure condition is still unknown.The knowledge of humid air affecting the filtration efficiency and pressure drop is limited.The understanding of the effect of humid exposure on cabin filter performance at different filter usage periods is even less.In addition, existing theoretical studies do not account for the humid exposure effect on particle filtration.
To fill this important knowledge gap, this study was designed to experimentally characterize the PM filtration efficiency and pressure drop in response to various relative humidity or water absorptions at different cabin filter usage periods.The theoretical equations were then revised to extend its application to calculate the filtration efficiency and pressure drop under humid exposure condition.The relationship between any two of the studied parameters (filtration efficiency, pressure drop, relative humidity, and water absorption mass) was explicitly presented.

Sample Filters and Testing System
The filtration efficiency and pressure drop were measured with four 6-month used vehicle cabin air filters, one new filter, one 3-month used filter, one 9-month used filter, and one 12-month used filter.These usage periods (3 months, 6 months, 9 months, 12 months) were chosen to represent the common range of the vehicle cabin air filters' usage period.The 6-month used filters were manufactured for Toyota Prado, Nissan Teana, Volkswagen Passat, respectively.The other filters that were manufactured for the same vehicle model of Toyota Prado SUV were selected from the maintenance workshop.All of the test filters are made of pleated glass-fiber papers.The cabin air filters for Toyota Prado were 190 mm in width, 200 mm in length, and 30 mm in thickness.By multiplying the pleat number and the pleat surface areas, filter surface areas were calculated as 0.5 m 2 .The Solid Volume Fraction (SVF) of the test filters are 5-10%.
An Europe Standard EN-779 classified testing system (European Air Filter Test Standard. EN 779:2002) was used in the experiment.Fig. 1 illustrated the schematic of the testing system setup.The particles in the ambient air were used as the particle source since the cabin air filters were exposed to the ambient air under the practical condition.The ambient air was injected into the testing system through a fan and a nozzle airflow meter that was calibrated according to ISO9300-2005 standard before the measurements.DI water vapor and droplets were generated from a sprayer that was connected with a peristaltic pump (Model BT100-2, Baoding Longer precision pump Co.).The humidity was controlled by the injected water flow rate from the pump.The relative humidity was monitored by an indoor air quality monitor-Qtrak (Model TSI 7565, TSI Inc. USA).The filter upstream duct was designed long enough to uniformly mix the airflow and the vapor on the cross section area of the test tunnel.The test filter was set up until the uniformity of filter face velocity was reached.The entire testing system was maintained with a positive pressure (5-15 Pa) over ambient to prevent outside air and humid leaking into the system.An Optical Particle Sizer (OPS, Model TSI 3330, TSI Inc. USA) was used to measure particle number concentrations (particles/cm -3 ) in the 0.3-5 µm range alternately upstream and downstream of the test filter.The filtration efficiency was calculated as η = 1 -downstream number concentration/upstream number concentration.The pressure drop across the test filter was continuously monitored using a manometer (Manometer 475 Mark III, Dwyer Instruments Inc. USA) in the experiment.

Test Protocols and Data Analysis
The filtration efficiencies and pressure drops across the test filters were measured at the airflow rate of 150 m 3 /h, which was the most commonly used vehicle ventilation airflow rate under the actual driving conditions (Xu et al., 2011).The measurement was conducted at a relatively consistent temperature (20 ± 2°C).To investigate the humid exposure affecting filtration performance, the filtration efficiencies and pressure drops were measured at three Relative Humidity (RH: 35%, 62%, 90%).This RH range covers most of the cabin air filters' RH exposure levels under real driving condition.It was found that small water droplets exist in the air upstream of the filter at the RH of 90%, and this represents the high humid and raining conditions.The filtration efficiencies and pressure drop across the filter were measured continuously as the water was absorbed.On the other hand, in order to investigate the filtration performance change as the dust desorbed the water, the measurement continued by stopping water injection into the system.The ambient RH is 35 ± 3% in the measurement.
It should be noted that the continuous water absorption in this study may be potentially different from the intermittent water absorption characteristics occurred in real-world.To generalize the results of the humid affecting the cabin filters' performance at different humid exposure frequencies, the filtration efficiency and pressure drop were expressed as a function of the water absorption mass at various filter usage periods.The filter was weighted every five minutes to calculate the water absorption mass in the filter.The data were collected after the filter face velocities and relative humidity were observed to be stabilized within 5% and 10% difference, respectively.

Theoretical Calculation
To extend the theoretical calculation of filtration efficiency and pressure drop under humid exposure condition, the frequently used equations (Eqs.( 1) and ( 2), Rao and Faghri, 1988;Brown, 1993) were revised in this study.
where η is the filtration efficiency of the filter, α is the Solid Volume Fraction (SVF), E is the total Single Fiber Efficiency (SFE), can be calculated as the Eq. ( 3) (Hinds, 1999), h is filter thickness, d f is fiber diameter, Δp is the pressure drop, f(α) is dimensionless pressure drop and can be then obtained as Eq. ( 4) (Davis, 1973).µ is the air viscosity, U is filter face air velocity.
where E D , E R , and E I are the single fiber filtration efficiencies due to diffusion, interception and inertia impaction, respectively (Hinds, 1999).
To extend the theoretical calculation for the filtration efficiency and pressure drop under the humid exposure condition, two coefficients "B" and "C", defined as the ratio of the filtration efficiency and pressure drop with water absorption to the filtration efficiency and pressure drop without water absorption, were added in Eqs. ( 1) and (2).The detailed derivations of the coefficient "B" and "C" are discussed in below section.

Filtration Efficiency and Pressure Drop for Test Filters
Fig. 2 shows the filtration efficiencies and pressure drops of the 6-month used filters.All test filters used in this study exhibits a similar filtration efficiency curve as a function of particle size.Since the vehicle cabin air filters were designed as a medium-efficiency filter, the filtration efficiency was measured in the fine particle size range from 0.3 µm to 5 µm.The filtration efficiencies varied significantly from 35% to 100% with the greatest increase observed for larger particles.For particles larger than 5 µm, the filtration efficiency is close to 100%, which implied the fact that the dust collected in the cabin air filter contains substantial large suspended dust that are enriched in metal elements (e.g., Zn, Cu and Cr) and ions (e.g., SO 4 2-) (Zhao et al., 2006).The pressure drops were found consistent at 60-75 Pa for all the test filters.

Effect of RH and Water Absorption on the Filtration Performance
The vapor or droplets were consistently absorbed by the dust in the filter.It leads to the change of filtration performance, e.g., filtration efficiency and pressure drop.The averaged filtration efficiency that was calculated as the mean of the filtration efficiencies at 6 particle size bars from 0.3 µm to 5 µm was used to express the filtration efficiency as a function of water absorption.Due to the limited number of size bars, the relationship between filtration efficiency and water absorption was not derived on a sizesegregated basis.This is a limitation of this study.Also, it should be noted that humid exposure affects filtration efficiency of smaller particles (diameter < 0.3 µm) due to the varying diffusion effect, which is not considered in this study.Fig. 3 illustrates the averaged filtration efficiency and pressure drop as a function of filter's humid exposure time at various RHs.It was observed that, after 60 min humid exposure at RH of 90%, filtration efficiency and pressure drop increased significantly by 20-25% and 170-200 Pa, respectively.This is because the dust in the filter absorbed the vapor or droplets and increased in sizes that led to a narrower and more curving air pathway inside filters.Narrower air pathway that caused larger fiber projected area vertical to the airflow direction results in a greater filtration efficiency.Also, narrower air pathway led to a significant air velocity increase inside filter media that caused the increase of pressure drop across the filter.More curving air pathway increased the possibility of particles leaving air streamlines and attaching on the fiber due to the inertial effect.In addition, air pathway with more curves led more airflow turns that results in an increased pressure drop.
From Fig. 3, it was also seen that the filtration efficiency and pressure drop increased more significantly with greater RH.Larger RH led more vapor absorption and faster dust size increase.Quantitatively for example, it took ~17 min for the pressure drop to increase 100 Pa at the RH of 90%.On the other hand, it took ~40 min for the same pressure drop increase at the RH of 62%.It was noted that the filtration efficiency and pressure drop increased faster at the beginning of the humid exposure.This is because the as the vapor and droplets were absorbed continuously.In absorption capability of the loaded dust in the filter decreased Fig. 3(b), the pressure drop curves at 20 min and 40 min with respect to 90% RH and 62% RH were bended slightly.This might due to the different absorption type before and after the bending point.After a certain amount of water absorption, the vapor might deposit on the surface of the dust instead of dust inside.After that point, the size of the loaded dust did not increase significantly that led to a slighter pressure drop increase.It was also found that the water can be desorbed as the RH reduced, which led to the decreases of filtration efficiency and pressure drop.In the experiment, the vapor injection was stopped to investigate the filtration performance change as the water was desorbed.As shown in Fig. 3, the dust desorbed the water faster than it absorbed the vapor.For example, at the RH of 90%, the pressure drop increased 200 Pa using 60 min; while it only took 40 min for the pressure drop to decrease 200 Pa.
In this study, the water absorption by the dust was faster than the intermitted absorption under the practical driving condition.This experimental approach can be reasonably applied to estimate the filtration performance change caused by humid exposure since the absorption mechanisms are the same.It should be noted that the dust loaded in the test filters was from the particles in the on-road atmosphere in China.Different loaded dust might lead to different filtration performance changes due to different water absorption characteristics.Furthermore, to apply these results at different humid exposure frequencies, the changes of filtration efficiency and pressure drop were expressed as a function of the absorbed water mass.
Fig. 4 showed the water absorption mass as a function of filter's humid exposure time.As expected, the water absorption mass is linearly proportional to the filter's humid exposure time.Larger RH led to more vapor exposure that resulted in a greater water absorption rate.
Substituting the filter's humid exposure time by the mass of water absorbed, the filtration efficiency and pressure drop were expressed as a function of the water absorption mass.Fig. 5 illustrates the relationship between the filtration performance and the water absorption mass in the filter.The result suggests that the filtration efficiency is significantly related to the water absorption mass in the filter instead of RH.The filtration efficiency is directly determined by the characteristic of the air pathway in the filter media that is related to the water absorption.On the other hand, the increase of pressure drop is related to both of the RH and water absorption mass.Quantitatively for example, with the same absorbed water mass, the increase of pressure drop at the RH of 90% was 50 Pa larger than the pressure drop at the RH of 62%.This is because higher RH caused faster droplet absorption in the loaded dust that led to greater dust size increase and narrower air pathways.

Expansion of Theoretical Calculations with Water Absorption
In order to extend the application of the theoretical calculation on the filtration efficiency and pressure drop under water absorption in the filter, two coefficients (B and C) were introduced and added in Eqs. ( 1) and (2).Eqs.
Fig. 6 shows the relationship between "B", "C" and the mass of water absorption in the measurement range (0-140 g)Coefficient B and C were regressed as a function of the water absorption mass.This is because "B" and "C" are directly associated with the increased SVF due to the water absorption.From Fig. 6, it was found that the coefficient "B" as a function of water absorption was derived as y = 0.003x + 1.032, R 2 = 0.98; the coefficient "C" as a function of water absorption was derived as z = 0016x + 1.456, R 2 = 0.89, where z is "C" and x the water absorption mass.As the water was continuously absorbed in the filter, the air volume fraction inside the filter media was reduced.Greater B indicates narrower air pathways inside filter media.
To further verify if the filtration efficiency and pressure drop with experimentally determined coefficient "B" and "C" agree with experimental results under different conditions, Eqs. ( 5) and ( 6) are used to calculate the filtration efficiency and pressure drop for the filters at the RH of 90%.Fig. 7 illustrates the comparison between the theoretical filtration efficiency and pressure drop corrected with coefficients and experimental data at the RH of 90%.No significant difference was observed.
Other than filtration efficiency and pressure drop, a parameter "Filter quality factor = -lnP/Δp", in which "P = exp(-4αEh/πd f )" is the penetration factor, was commonly used to determine the filter quality.By substituting "P" and "Δp" into the filter quality factor calculation, filter quality factor can be calculated as Eq. ( 7).Therefore, a Water absorbed in the filter, g coefficient "D = B/C" can be used to express the filter quality factor under water absorption condition.

Effect of Humid Exposure on the Filtration Performance at Various Dust loads
Filter usage period determined the mass of dust loaded in the filter, which potentially affect the water absorption capacity.According to the literature, 3-month, 6-month, 9month and 12-month usages represent 0.8 g, 1.5 g, 2.4 g and 3 g dust loaded in the filter (Xu et al., 2011).Fig. 8 illustrated the changes of filtration efficiency and pressure drop as the filter was exposed in the humid air at different usage periods.For clarity, only the measurement results of the filters with three filter usage periods (new, 6-month, 12-month) were shown in Fig. 8.It can be seen that, for the new filter, the effect of humid exposure on the filtration performance is negligible.On the other hand, the filtration efficiency and pressure drop increased significantly in response to the increased humid exposure time for used filters.With the same humid exposure time, 12-month filter usage led to 8-12% increase of filtration efficiency and 50-100 Pa increase of pressure drop, respectively.This can be explained by the fact that the absorption capacity of the dust loaded in the filter is much larger than the filter media.
To further investigate the effect of humid exposure on the filtration performance of vehicle cabin air filters and generalize the findings of this study to estimate other cabin air filters' filtration performance, the coefficients "B" and "C" as a function of water absorption mass were linearly regressed at different filter usage periods, as shown in Fig. 8. Humid exposure led to consistent increases of filtration efficiency and pressure drop.The increasing rate for longtime used filter is slightly larger than short-time used filter.For example, 50 g water absorption caused 120 Pa and 150 Pa pressure drop increase for 6-month and 12-month used filters, respectively.Dramatic pressure drop increase potentially led to significant decrease of airflow rate in the vehicle ventilation system.This finding can be applied to facilitate the maximum humid exposure period to prevent from ventilation malfunction in the practical driving condition.

CONCLUSIONS
In summary, the effect of humid exposure on the vehicle cabin air filter's performance was evaluated and investigated at different RHs at different filter usage periods.For test cabin air filters, the averaged filtration efficiency and pressure drop were measured at ~70% and 75 Pa, respectively.Significant increase of filtration efficiency (up to 15%) and pressure drop (up to 250 Pa) were observed as the humid exposure time increased.Filtration efficiency was influenced most by the mass of water absorbed in the filter.On the other hand, pressure drop was affected by the water absorption mass and the RH simultaneously.The pressure drop increased more significantly at the beginning of the humid exposure due to the greater water absorption capacity of dryer dust.It was also found that dust loading posed a significant effect on the change of filtration performance.The filtration efficiency and pressure drop of the 12-month used filter increased 2 times faster than the new filter at the same humid exposure condition.The filtration efficiency and pressure drop was explicitly expressed as a function of humid exposure time and water absorption mass in the filter.Coefficients "B" and "C" that were derived as a function of water absorption mass were introduced to extend the application of the theoretical calculations for filtration efficiency and pressure drop to particles under humid exposure.Theoretical calculations corrected with "B" and "C" agreed well with experimental measurements under other studied conditions.
The findings of this study can be used to facilitate the maximum vehicle cabin air filter's exposure period to avoid malfunction of vehicle ventilation system, or as a reference to further upgrade the vehicle's ventilation airflow setting to prevent from dramatic pressure drop increase.

Fig. 1 .
Fig. 1.Experimental schematic of the experimental test system.

Fig. 2 .
Fig. 2. Measured particle (a) filtration efficiencies and (b) pressure drops of four 6-month used cabin air filters.The airflow rate is 150 m 3 /h.

Fig. 3 .
Fig. 3.The relationship between (a) filtration efficiency, (b) pressure drop and filter's humid exposure time at different RH conditions.6-month used filter for Toyota Prado were tested.

Fig. 4 .
Fig. 4. The relationship between the water absorption mass and the filter's humid exposure time at different RH levels.

Fig. 5 .
Fig. 5. (a) Filtration efficiency and (b) Pressure drop as a function of the water absorption mass in the filter.

Fig. 6 .Fig. 7 .
Fig. 6.Coefficients "B" and "C" as a function of the water absorption mass.The experimental data at the RH of 62% was used.

Fig. 8 .
Fig. 8. (a) Filtration efficiency and (b) pressure drop as a function of humid exposure time for filters at different usage periods.90% RH was used in the measurement.

Fig. 9 .
Fig. 9.The relationship between (a) coefficient "B", (b) coefficient "C" and the water absorption mass in the filter at different filter usage periods.62% RH was used in the measurement.