Why air flow matters more than filter rating
After lots of reading about different filters and how effective they are at filtering different particle sizes I have come to an unexpected conclusion, the rating doesn't matter that much. What really matters is your air changes per hour, that is how many times per hour the air is filtered.
e.g. If your room is 60m3 and your air purifier does 120m3 per hour then its doing 2 Air Changes per Hour (ACH).
The reasons I came to this conclusion:
This shows it would take 11 ACH of a HEPA filter to get under 5. That may explain why I cannot seem to get there. Interestingly there is very little difference between a 99.97% HEPA and a 95% MERV 16. As we get down to a MERV 13 and MERV 11 we are starting to get mostly small particles which is not ideal.
The filters are stating to struggle! Running my Philips Series 2000 Air Purifier which can do up to 79m2 room would only keep a 12m2 under 25 PM2.5.
So halving the houses leaks meant the filters could do double the size.
This simulation shows that the increased airflow of the MERV 13 filter meant it almost matched the HEPA on PM 0.3-1.0, while beating it on PM 1.0
e.g. If your room is 60m3 and your air purifier does 120m3 per hour then its doing 2 Air Changes per Hour (ACH).
The reasons I came to this conclusion:
- Adding a DIY purifier using a less effective filter in combination with my proper air purifier consistently got me under 10 PM2.5 in my rumpus/living area.
- Adding a DIY duct filter using a MERV 16 (might be fake and really MERV 13) has helped my upstairs area have low readings
- SmartAir found in Why a weaker filter actually makes a stronger purifier that using a slightly less effective filter that allowed for a 26% increase in air flow improved the system by 27%.
- SmartAir found air flow is the only way to tell a filer needs replacing and not how full it is. In fact filters capture more dust the dirty they get. Essentially the effectiveness of the system was proportional to the air flow.
- Research into Comparison of MERV 16 and HEPA filters for cab filtration of underground mining equipment found no statistical difference between a HEPA filter (99.97%) and a MERV 16 (95%)
- Reducing patients’ exposures to asthma and allergy triggers in their homes: an evaluation of effectiveness of grades of forced air ventilation filters found via simulations that a MERV 13 reduced outdoor particles by 70%, while a MERV 16 achieved only a little more at 76%. So a 20% increase in filter efficiency resulted in only an 8.5% increase in performance.
Air Changes per Hour ACH
When you think about it if the filter removes say 80% per pass, then after one pass you have 20% left, and on the second pass 4% are left. Considering the recommendation is 5 passes per hour it doesn't take long to clean up the air.
Here is the filters removal rate considering this:
Simulation
With this in mind I ran a few spreadsheet simulations to see how different filters should perform. The conditions were:- PM2.5 count of 60 outside, and initially 60 inside. For Canberra 60 outside is optimistic.
- I actually assumed PM 0.3-1 = 30, PM 1-3=30, PM 3-10 = 30
- Outside air leakage into the house was 1 Air Changes per Hour. This is a bit above the 0.77 average found by CSIRO
- I assumed the pollution was uniformly distributed
The simulation shows that after about 30 minutes the inside level would stabilize and no longer drop.
 |
| 99.97% HEPA filter at 5 ACH |
Simulation 1: 60 PM2.5 outside and house leaks at 1 ACH
Next I decided to see what Air Change per Hour (ACH) is required to get under various PM2.5 readings.
This shows it would take 11 ACH of a HEPA filter to get under 5. That may explain why I cannot seem to get there. Interestingly there is very little difference between a 99.97% HEPA and a 95% MERV 16. As we get down to a MERV 13 and MERV 11 we are starting to get mostly small particles which is not ideal.
Simulation 2: 330 PM2.5 outside and house leaks at 1 ACH
Over the last month Canberra was over 60 PM2.5 for 64% of the time. If you want a solution that will work 90% of the time over the last month in Canberra you need to deal with up to 330 PM2.5.The filters are stating to struggle! Running my Philips Series 2000 Air Purifier which can do up to 79m2 room would only keep a 12m2 under 25 PM2.5.
Simulation 3: 330 PM2.5 outside and house leaks at 0.5 ACH
Next I decided to see how making the house more air tight would help with the terrible air in Canberra. Given the average in a CSIRO study was 0.77 I am using reasonable numbers. Although 0.15 is apparently a good aim for new builds.
So halving the houses leaks meant the filters could do double the size.
Simulation 4: 60 PM2.5 outside and house leaks at 1.0 ACH comparing HEPA to MERV 13
I have 2 HEPA filters I got off eBay I thought I would compare how putting the filters onto the same fan would go. I measured the HEPA with a shroud to blow 307m3/hr while the MERV 13 was at 576 m3.hr. I am using a 55 watt fan.
This simulation shows that the increased airflow of the MERV 13 filter meant it almost matched the HEPA on PM 0.3-1.0, while beating it on PM 1.0
Conclusions
The simulations have helped me to understand what I have observed, that a MERV 13 filter strapped to a fan doing 568m3/hr has helped along side my HEPA filter doing 416m3/hr due to higher air flow.
I would also aim for an ACH of at least 5 10, if not 15-20.
I would also aim for an ACH of at least 5 10, if not 15-20.
Notes:
I have since read that I should drop the efficiency by 40% as explained here: What’s CADR and How is it Different From Airflow? As such I expect I need to almost double the above ACH results.Others Research
I have found a much more advanced simulation Reducing patients’ exposures to asthma and allergy triggers in their homes: an evaluation of effectiveness of grades of forced air ventilation filters that came to a similar conclusion.For outdoor PM2.5, the 70% removal efficiency of Filter D (MERV 13) achieves ∼70% effectiveness, while the MERV 16 achieves only 76% effectiveness on outdoor PM2.5.
The portable HEPA unit offered little effectiveness on a whole-home basis and had the greatest annual operating cost at $140/year, including replacement filters and electricity, but excluding the original purchase price.
Far out, thanks for sharing all this data!
ReplyDelete